DFID Understanding Afghanistan

Document Sample
DFID Understanding Afghanistan Powered By Docstoc
					DFID Understanding

Growth Diagnostic

3.4.1 Draft Final Report
- Technical

Alfie Ulloa

May 2008
DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Acronyms and Abbreviations ................................................................................................... 3
      Acknowledgements ................................................................................................................ 4
      Abstract ................................................................................................................................ 5
      Executive Summary ................................................................................................................ 6
      1        Introduction and Background ........................................................................................ 11
      2        Purpose, Methodology and Limitations ........................................................................... 11
               2.1      Purpose ......................................................................................................................................... 11
               2.2      Methodology ............................................................................................................................... 12
               2.3      Limitations .................................................................................................................................. 14
               2.4      Reducing the limitations ........................................................................................................... 14
      3        The Afghan Economy, a Literature Review ...................................................................... 15
      4        The Economic Context ................................................................................................. 16
               4.1      Growth ......................................................................................................................................... 16
               4.2      Investment ................................................................................................................................... 18
               4.3      Regional Trading Context ........................................................................................................ 20
               4.4      Sector analysis ............................................................................................................................ 22
               4.5      The Investment Climate ............................................................................................................ 27
      5.       Growth Diagnostic ...................................................................................................... 27
               5.1      Finance ......................................................................................................................................... 28
               5.2       Low Social Returns ................................................................................................................... 31
                        5.2.1          Geography ................................................................................................................... 31
                        5.2.2          Infrastructure ............................................................................................................. 33
                        5.2.3          Human Capital ........................................................................................................... 38
               5.3      Low Appropriability of Returns .............................................................................................. 41
                        5.3.1          Market Failures: Self-Discovery .............................................................................. 41
                        5.3.2          Government Failures ................................................................................................. 50
      6        Conclusions ................................................................................................................ 59
      7        Policy Recommendations .............................................................................................. 61
      References ............................................................................................................................ 62
      Annex1            Table of Key Indicators (Time Event since Peace Agreement) ................................... 66
      Annex 2           Classification of Occupation by Respondents ......................................................... 68
      Annex 3           Methodological Note for Returns to Education (Mincer equation) ............................ 69

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            List of Figures
            Figure 1. The HRV Growth Diagnostic Framework
            List of Tables
            Table 1. Afghanistan, Real GDP and GDP per capita
            Table 2. Total Number of companies registered with AISA
            Table 3. Afghanistan, investment and savings
            Table 4. Afghanistan, Balance of Payment, 2002/03-2006/07
            Table 5. Summary of Import Tariff Schedule (as of January 10, 2008)
            Table 6. Afghanistan, GDP at constant prices, annual percentage change
            Table 7. Summary Statistics on Afghanistan‘s Opium Economy
            Table 8. Afghanistan, occupation profile (% of households reporting the activity as first or
            second most important source of income)
            Table 9. Overall and sector specific ranking on obstacles to business, the World Bank
            Investment Climate Assessment
            Table 10. Afghanistan, the financial sector
            Table 11. Source from which the household got the largest loan last year
            Table 12. Main use given to the loan
            Table 13. First source of income for urban households investing in business
            Table 14. Afghanistan and selected countries, Logistic Performance Index (2007)
            Table 15. Sources of electricity supply (operating capacity in MW)
            Table 16. Average years of education in population over 25, selected countries
            Table 17. Primary enrolment, selected countries
            Table 18. Secondary and tertiary enrolment, selected countries
            Table 19. Afghanistan, top ten exports in 1962 and 2006 (2000 US dollars, millions)
            Table 20. Afghanistan, fiscal sustainability analysis
            Table 21. Afghanistan, revenue portfolio
            List of Graphs
            Graph 1. GDP Growth, Afghanistan and selected post-conflict countries
            Graph 2. Afghanistan, Exports and Imports (real U.S. 2000 dollars)
            Graph 3. Afghanistan, Iran and Pakistan, Yield (wheat, hg/HA)
            Graph 4. Afghanistan, selected exports index (1998=100)
            Graph 5. Total exports per destination
            Graph 6. Export Growth Decomposition 1977-2006
            Graph 7. World Dried Fruit Market, 2006
            Graph 8. Russia Dried Fruit Market, 2006
            Graph 9. Dried Fruit imports in Russia
            Graph 10. Edible Nuts World Market, 2006
            Graph 11. India Edible Nuts Market, 2006

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Acronyms and Abbreviations
          ACD                 Afghan Customs Department
          AISA                Afghanistan Investment Support Agency
          AREU                Afghan Research and Evaluation Unit
          ARTF                Afghan Reconstruction Trust Fund
          CAO                 Control and Audit Office
          DAB                 Da Afghanistan Bank
          EPAA                Export Promotion Agency
          FDI                 Foreign Direct Investment
          GD                  Growth Diagnostic
          GDP                 Gross Domestic Product
          HIPC                Highly Indebted Poor Countries
          IARCSC              Independent    Administrative    Reform   and    Civil   Service
          ICA                 Investment Climate Assessment
          IMF                 International Monetary Fund
          ISAF                International Security Assistance Force
          LOTFA               Law and Order Trust Fund
          MDG                 Millennium Development Goals
          NATO                North Atlantic Treaty Organisation
          NDB                 National Development Budget
          NRVA                National Risk and Vulnerability Assessment
          NSP                 National Solidarity Program
          PIP                 Public Investment Program
          PMU                 Program Management Unit
          PPP                 Purchasing Power Parity
          PRR                 Priority Reform and Restructuring
          UNAMA               United Nations Assistance Mission to Afghanistan

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            We would like to thank the HMG Department for International Development for the
            opportunity to undertake this challenging study, also for the support and valuable
            contribution provided during the process. Miguel Laric, DFID Economist in Kabul, provided
            insightful comments and helped to guide Team members during their time in Afghanistan. A
            long list of key informants, colleagues and anonymous persons contributed to the report while
            the Growth Diagnostic Team was in the field in Kabul in April 2008. Without them our
            understanding of the Afghan economy and reality would have been incomplete. Particular
            thanks must be given to Royce Wiles who provided much needed information from the
            precious AREU library. We would like to thank Ajmal Ahmady for his contribution,
            particularly in relation to the literature review. Sharon Miller, Anna Patterson, Nipa Banerjee,
            Michaela Prokop and the entire UA Team provided helpful and constructive comments.
            Thanks are due to all the entire Understanding Afghanistan team for a friendly and fruitful
            exchange of ideas, comments and contributions. Thank you to Pamela Frenk, Reinier
            Schliesser, Nicholas Khaw, and Andres Zahler for the invaluable research assistance. We also
            would like to recognise the tremendous efforts of the UA Team Management, particularly
            Oliver Mathieson, Maxwell Stamp Project Director, Robbie Gregorowski, Maxwell Stamp Project
            Manager and Peter Middlebrook, Understanding Afghanistan Team Leader who oversaw all
            components of the engagement.           All error and omissions remain author‘s exclusive

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

             This paper revisits economic growth in Afghanistan using the growth diagnostics framework
             proposed by Hausmann, Rodrik and Velasco (2005), with the purpose of identifying the most
             binding constraints for economic growth. To rank public policy priorities the HRV (2005)
             methodological approach is complemented with quantitative and qualitative analysis, aimed
             at exploring the several hypotheses the model present as potential constraints. The challenges
             in applying this methodology in Afghanistan are many and exacerbated by severe data
             limitations. Our analysis shows how fragile is the Afghan financial market and economy and
             the deep impact the many years in war have left for both human and physical capacity in the
             country. Also the challenges the Afghan geography and location impose to the country. We
             identify microeconomic risks, in the form of property rights, taxation and corruption as the
             most constraining issues for private investment in Afghanistan today. Infrastructure, in
             particular associated to the rural economy and the agribusiness sector and to the dynamic
             exporting subsectors appear as the second constraint, and it is probably the most important in
             long term. From a public policy perspective, the most relevant result is the confirmation that
             in Afghanistan, private investment decisions are being negatively affected by the uncertain
             environment, including security and the unpredictable tax structure, and by rampant and
             widespread corruption.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Executive Summary

            Afghanistan has long been troubled by political turmoil and instability. Twenty -three years of
            war and conflict has made the country an extremely unattractive location for investors, both
            domestic and foreign. Indeed, Afghanistan ranks among the poorest nations in the world. The
            2001 overthrow of the Taliban regime is seen as the dawning of a new era for economic growth
            in Afghanistan.
            Boosted by donors‘ funds and the catch-up impulse, recent years have seen a high positive
            growth rate for the Afghanistan economy. How does the Afghan economy continue to grow at
            a sustainable rate and what are the challenges facing the economy? This paper attempts to
            address those issues, using the Growth Diagnostic model to identify and pinpoint the binding
            constraints to growth in the Afghan economy. This approach therefore strengthens the present
            efforts around the ANDS and delivers a cogent approach to linking binding economic
            constraints to public investment outcomes.
            The Growth Diagnostic model was developed by Hausmann, Rodrik and Velasco in 2005. This
            methodology classifies and ranks constraints on investment growth in a given economy. This
            framework is designed to encourage context-driven policy and reform supported by economic
            analysis, rather than the ―laundry list‖ approach, of implementing a generic list of ‗best
            practice‘ policies.
            It first asks if the lack of private investment in an economy is due to a high cost of investment
            in the economy or a low returns to investment. If the problems are due to a high cost of
            finance, then the model poses the next question; is the high cost of finance due to inadequate
            international finance or inadequate domestic finance? Having identified that the most binding
            constraint is not in the high cost of finance section, the problem is within the low returns to
            economic activities branch, which then splits into low social returns or low appropriability,
            and further down into complementary factors that make returns low or failures due to market
            imperfections or government mismanagement. Going down the Growth Diagnostic tree, it is
            crucial to keep in mind that a country can suffer from many different economic constraints,
            but what the researcher is trying to pinpoint is the most binding constraint in the economy.
            Thus, the model prioritizes problems that pose the biggest challenges to the economy, enabling
            policy makers to focus their efforts on one problem at a time, rather than the ‗kitchen sink‘
            approach that was the premise of the Washington Consensus. Resources can therefore be used
            more efficiently and effectively.
            We begin by providing an economic context for Afghanistan. The double -digit growth rates
            posted by Afghanistan since 2001 are recognizable in post-conflict countries as the economy
            bounces back from very low GDP levels thus paying a ―peace-dividend‖. Still, compared against
            other selected post-conflict countries, Afghanistan‘s economic performance outperforms other
            post-conflict economies in terms of GDP growth. But now that the catch up effect is
            dwindling, growth in the medium and long term will depend upon improvements in
            governance and a successful integration into the world economy.
            Driven by donor-funded government investment, investment peaked in 2006/2007. However,
            private investment fell in the same period from its peak in 2005. Only the services and
            construction industries seem to be attracting more investments, particularly domestic
            investment. The IMF projects that private investment will not recover until 2009/2010, and
            projects FDI of 4.1 percent of GDP by 2012/2013, below the level attained in 2005/2006.
            On a more positive note, trade in Afghanistan has been a very active sector. Afghanistan has a
            very open economy, relative to its neighbors. Recently, Afghanistan has joined the South Asian
            Free Trade Agreement area and is pursuing trade agreements with its neighbors, hoping to
            increase the amount of trade opportunities. Full accession into the WTO is expected to begin
            in 2010. Excluding the opium industry, total trade in 2007 was estimated to be over $8.5
            billion, comprising of $6.7 billion in imports and $1.8 billion in exports. Official exports have
            grown over 300 percent over the last 5 years.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            A sector analysis in the post-2001 era shows services as the main sector of the economy at 38
            percent of GDP in 2006, followed by agriculture at 32 percent and industry at 27 percent.
            Services overtook agriculture as the main contributer to GDP in 2004. And since 2005 industry
            is the most dynamic sector in terms of growth rate. Construction and manufacturing explain
            industry performance. Growth in services is driven by aid injection into telecommunications,
            aviation, restaurants and other subsectors in which there is a high demand from the
            international community. Agriculture has not been particularly successful posting negative
            accumulated growth since 2003, but this is due to external factors such as drought and the
            interaction between legal agriculture and opium production. The most recent period of
            drought caused a negative 20 percent growth in the sector in 2006/2007, but desp ite this,
            agriculture still accounts for almost one-half of Afghanistan‘s exports in 2006. On the other
            hand, illegal agriculture, particularly opium cultivation remains the largest and most
            important industry in Afghanistan. Approximately 3.3 million Afgh ans are involved in opium
            cultivation. Once opium is removed from official agriculture figures, legal agriculture is
            reduced dramatically to 19 percent of official GDP and thus becomes the sector with the
            smallest contribution to the economy.
            The underground economy also includes barter trade, manufacturing, commerce and other
            non-recorded services, as well as smuggling and re-exports. Afghanistan informal economy has
            been estimated in around US7 billion, or between 80 to 90 percent of Afghanistan‘s recorde d
            GDP. The main components in the informal economy are: opium production and processing
            activities (35 percent of official GDP), subsistence agriculture and livestock (30 percent of
            official GDP), and illegal trade (8 percent of official GDP). The bulk of the labor force is
            employed by the underground economy and the large majority of Afghanistan‘s population is
            dependent on it. Absence of statistics and unreliability on these estimates make impossible to
            explicitly consider the informal economy into the GD analysis. Yet, this does not invalidate the
            analysis since the main variables in the model; GDP growth and investment (i.e., capital
            accumulation) are mostly dependent on the formal sector.
            Subsistence agriculture is by definition a low technology and low productivity sector, with no
            capital accumulation. Even opium, the largest component of the informal (and in this case
            illegal) economy requires low capital investments. Re-exports is just brokerage and a cash-
            making activity with no relevant investments in terms of capital accumulation. Only smugglers
            (i.e., illegal exports) and some informal commerce and services (e.g., private generators) would
            invest in capital assets, although marginal for total investment. Also, these sectors are a very
            small fraction of the Afghan GDP and therefore have a small contribution. Of course, parts of
            the income generated in all these activities will enter the formal economy either as
            consumption, savings or investments. This is more evident, for example in the constructio n
            sector. In those cases, the multiplier effect of these activities is being recorded and therefore
            considered by the analysis. Focused on explaining constraints for long -term growth in
            Afghanistan and centered on explaining the slowdown on private investment since 2005, our
            ignorance on the informal economy does not limit the GD framework capacity.
            In 2005, electricity, access to land and corruption were identified, by the Investment Climate
            Assessment, as the three most important barriers for business in the country. Since 2005, we
            have observed some progress on electricity and a little progress on land-related issues. But
            since then corruption has worsen and the institutional setting deteriorated as captured by
            governance and transparency indicators, as well as businesspeople perceptions.
            Having drawn out the current economic context in Afghanistan, we begin applying the
            Growth Diagnostic framework to the Afghanistan economy. The first branch which we
            consider is the Finance branch. The cost of finance in Afghanistan is high, but not excessive.
            The relevant lending rate is listed as 18.8 percent in the third quarter of 2007. In comparison
            with other post-conflict nations, this figure is average and thus, relatively expected. Although
            formal financial services remain limited, the World Bank survey, data from the NRVA and
            anecdotal evidence implies that there is no extreme shortage of capital in the business
            community. Even when formal banking is irrelevant and most investment are self -financed. As
            the economy continues to outgrow the catch-up phase and thus requires more private

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            initiative, access to finance might become a potential binding constraint. But at the present,
            with domestic investment in around 5 percent of GDP and savings over six times that num ber,
            finance is not Afghanistan‘s most binding constraint.
            If not due to high cost of finance –the GD sequence goes– low investment should be explain by
            low returns to investment or lack of appropriability. We move first into the low social returns
            analysis, which is categorized into three branches: geography, low human capital and poor
            infrastructure. We first consider geography. Afghanistan is located at the centre of Asia and is
            landlocked and mountainous. Afghanistan‘s location was its main asset for ce nturies, during
            which the country served as a land bridge in the region, connecting south and central Asia‘s
            trade routes. Its geography, location and topography impose serious mid and long-term
            challenges on the country, while also being a potential source for long term benefits as it
            possesses abundant natural resources (much of which remain untapped) and a large variety of
            climate and soils. Mining has the potential for becoming a key sector in the future. Recurrent
            droughts severely damaged agriculture and livestock before 2001, in 2004 and 2006. But, in
            spite of being landlocked and droughts prone, Afghan exports have kept growing at very high
            rates (higher than GDP since 2001) and agribusiness (fruit, nuts, skins and other produces) are
            the start performers. However, this boom in commerce is dependent on neighboring Pakistan
            and India, reflecting Afghanistan limited capacity to reach distant markets. This evidence
            suggests geography, even if challenging, is not constraining country‘s economic activity.
            Infrastructure is the most critical issue for Afghanistan‘s competitiveness and development in
            the long term. With the appropriate investments on infrastructure, country‘s geographic
            characteristics could shift from a liability to an asset. Electricity and irrigation were identified
            as two key sectors likely to be binding constrains for specific business, but we have no
            capacity to test at the microeconomic level. On-site generation (diesel generators) and
            industrial parks seem to be the way those business have overcome electricity limitation. At the
            same time, cereal yields, highly dependent on irrigation, are currently at historically high levels
            and agribusiness is the most dynamic sector in terms of export growth. At the same time,
            infrastructure is one of the sectors that have benefited most from aid and donor-financed
            reconstruction efforts, with steady improvements since 2001. In the GD way of thinking, if a
            complementary factor (like infrastructure) is constraining, those sectors using the factor m ore
            intensively will be the most constrained. We don‘t see this happening in Afghanistan, and on
            the contrary evidence suggests infrastructure is not the most binding constraint to the Afghan
            economy. Nonetheless, we do have evidence of exporters using Pak istan infrastructure and of
            large impacts on business margins due to the energy bill. Also, the potential benefits as land
            corridor and natural resources exporter would only be real if the specific investments are done.
            Given these long-term perspectives for the Afghan economy and the interactions between
            geography and infrastructure, we see infrastructure as the most relevant complementary factor
            in a long-term perspective.
            Another complementary factor the GD considers is human capital. Unsurprisingly,
            Afghanistan‘s stock of human capital ranks among the lowest in the world. But again we have
            no evidence of this as the key constraining issue at present. This is due to a combination of
            returning educated Diaspora, labor mobility within the area and low dema nd for highly-skilled
            workers given the current technological structure.
            With the evidence at hand, we consider Afghanistan is not an economy with ―low returns‖.
            Despite limited human and physical capacity and challenging geography, there are profits to b e
            made and businesspeople taking advantage of them. The question then is why private
            investment has stagnated since 2005. The GD considers another set of explanations on a
            second branch to investigate entrepreneurs‘ capacity to appropriate returns. Under t his
            branch, it can be that market or government failures prevent investors from appropriating the
            fruits of their projects. This is, even if the cost of investment (i.e., the interest rate) is lower
            than the expected returns (i.e., not a low returns economy), still entrepreneurs would not
            invest if the perceive those returns won‘t be fully appropriated.
            We discard ―self-discovery‖ and market failures. Even if the Afghan export package continues

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            to be dominated by a few traditional goods and concentrated in regional countries, it is very
            dynamic. Also, there is space for increasing market share and diversifying destinations while
            constructing a solid exports base. The country is being re-integrated into the world economy
            by its traditional comparative advantages, but the lack of specific infrastructure reduces the
            chances for those exports to locally add-value and expands. In the long run, the tradable sector
            will need to develop new export opportunities and further integrate Afghanistan‘s export
            sector into the international economy.
            The second branch under the low appropriability of returns is government failures. Under this
            branch, there are, again, two branches: macroeconomics and microeconomic risks. We first
            consider macroeconomics. Currently, there are no signs of macroeconomic imbalances or
            expectations of macroeconomic risks and volatility. It is important for the government and for
            the monetary authority to continue its prudent policies and macro -economic reforms,
            particularly in this period of rising food and oil costs. More importantly, in the medium term,
            as donor funds reduce, fiscal sustainability might become a serious issue. But at the present,
            macroeconomic volatility is not a constraining issue for the economy.
            The last branch to consider is microeconomic risks. The GD framework identifies three micro
            issues that might reduce appropriability of returns: property rights (i.e., rule of law), taxation
            and corruption. On this branch we identify the 2004 tax reform and the subsequent
            selectively-aggressive behavior of the Large Taxpayers Office as a possible shock that explains
            the reduction and stagnation in private investment since 2005. At the same time, a perceivably
            increased on corruption has been mentioned. We argue these two events are not ind ependent,
            and the political economy of this patrimonial and tribal society might explain the connection.
            Even if the efforts to isolate large taxpayers from previous corrupt and inefficient tax
            collection agencies move into the right direction, an inherently weak institutional setting
            cannot prevent new sources of informal taxes and corruption for creating new illegal
            fundraising operations. Is being reported administrative fees, permits and licenses at
            ministerial and municipal level have increased. Corruption and inefficiency also limits
            contracts enforceability and property rights in Afghanistan. The Ministry of Interior and the
            judiciary are perceived as the most corrupt entities in the Afghan Government. In addition,
            security issues liked to the conflict (i.e., suicide bombings and armed conflict) and criminal
            activity (i.e., the opium economy, abductions, rubbery) have also worsened since 2005. In
            conclusion, all three factors into the micro risks node are increasingly affecting the expectation
            of returns appropriability. We conclude this (i.e., micro government failures) is the most
            binding constraint to private investment in Afghanistan today.
            We argue this business environment is not conducive to the development of the private sector
            and is not conducive to growth or long-term sustainable fiscal structure, and therefore it may
            explain investors‘ fears. The government is taking misguided steps by overtaxing the formal
            sector and increasing tax uncertainty to large taxpayers instead of broadening the ta x
            structure by increasing investment and formalizing part of the legal but informal economy.
            Over-taxation, corruption, selective law enforcement and under-provision of public goods will
            deter further investment and alienate potential formal investors, thereby leading to greater
            informality and the need for greater taxation. This vicious cycle will be perceived by
            entrepreneurs as a risk and therefore less investment will exist.
            Absence of primary data impeded efforts to conduct more robust empirical tests that would
            more accurately support our findings, and therefore their interpretation and the constraint
            ranking proposed should be taken with caution. Nonetheless, we have used all available
            information to rank the most pressing urgencies in terms of private investment, or in the GD
            jargon, the most binding constraint.
            This first attempt to use the GD in a conflict economy leaves two main lessons in terms of the
            framework performance. On the one hand, the many pressing needs and the limited amount of
            accurate data limit researchers‘ capacity to confidently pinpoint the most binding constraint in
            the given context. But at the same time, GD‘s flexibility and crafted sequencing does provide
            with a roadmap to where attention and further studies should be directed, helping both the

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Afghan Government and the donor community to get a sense of priorities. Other growth
            analysis techniques (i.e., growth accounting) are just impossible to use, and uninformative in
            terms of prioritization.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      1     Introduction and Background
             Since 2001 the Government of Afghanistan and the donor community have sought to develop a
             cohesive national development strategy that creates stability and security while shaping and
             driving high levels of sustained and broad based economic growth. Afgha nistan is not only one
             of the poorest countries according to the Human Development Report, but also one where
             necessity to reconstruct the war torn economy is almost unparalleled at this moment in
             history, from a physical, human and institutional infrastructure point of view. Twenty-three
             years of war and devastation cannot be easily reversed, evidenced by a multi-billion
             reconstruction exercise.
             Whilst the need to deliver development solutions aimed to position the country on a path to
             sustained growth and effective poverty reduction is well known, the means by which this may
             be achieved is less clear. This chapter seeks to reduce that uncertainty by identifying the
             underlying binding constraints to broad based economic growth that continue to inhibit
             productivity, stifle entrepreneurial activity and encourage the burgeoning of illicit economic
             activity. To that purpose, we follow the Hausmann, Rodrik and Velasco (2005) Growth
             Diagnostic methodology to classify and rank constraints on investment and economic growth.
             This supports and strengthens current efforts around the ANDS and will deliver a coherent
             approach to link binding constraints to public investment outcomes.
             In terms of understanding the extent of economic barriers to enhanced production, the Wor ld
             Bank Doing Business report lists Afghanistan 178th out of 178 countries in terms of investor
             protection, 177th in relation to credit access, 174th in terms of cross border trade, and 160th
             with regards to contract enforcement. Put simply, Afghanistan is one of the last countries
             where an investor would consider making substantial investment within the formal economy,
             even if the government has managed to control inflation, sustain the currency and boost
             exports. Afghanistan is known to have a high number of constraints ranging from a precarious
             institutional environment, lack of human capital and infrastructure, low competitiveness,
             microeconomic failures, land management and ownership constraints, insecurity, high levels
             of corruption that increase core transaction costs among many others. It is essential to have
             some sense of priorities and to identify the most constraining areas to growth. The Growth
             Diagnostic has been designed with this challenge in mind, and a sense of priorities will be
             given in terms of most binding constraints and less urgent issues to solve.
            Our findings are not new, as they confirm the relative consensus in existing data on the factors
            that inhibit private sector development. We conclude that ‗micro risks‘ (i.e., taxation on the
            formal sector, corruption and property right issues), in the ―government failures‖ section
            currently represent the most binding constraint for private investment in Afghanistan.
            Infrastructure, in particular to overcome the many geographical challenges and directly
            focused on sectors with export potential appear as the second most constraining issue, and is
            probably the most important in long term perspective.

      2     Purpose, Methodology and Limitations
            2.1        Purpose
            The aim of this study is to identify the most binding constraints to economic growth in
            Afghanistan today. In the next sections we apply the growth diagnostics methodology
            developed by Hausmann, Rodrik and Velasco (2005), which seeks to prioritize the economic
            distortions a country should eliminate or reduce in an effort to enhance and sustain economic
            Paucity of data makes it difficult to conduct a thorough study of country‘s growth barriers, in
            particular when using the Hausmann, Rodrik and Velasco (2005) Growth Diagnostic
            framework. There is a lack of historical information on key macroeconomic variables, and most
            microeconomic or sectoral data is unavailable. A large informal sector is unaccounted for in
            official statistics and its composition and evolution is by necessity based on estimates and
            conjecture. In order to support our growth analysis in the presence of data deficiencies and the

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            underestimation of GDP due to the informal and illicit economy we complemented available
            data with perceptions-based information from the Investment Climate Assessment (World
            Bank, 2005d), the Doing Business Report (www.doingbusiness.org), and other sources, as well
            as from key informant interviews in Kabul, in April 2008. The usual caveats when perception -
            based information is considered apply, and therefore their interpretation should be taken with
            caution. These general caveats and the absence of primary data impeded efforts to conduct
            more robust empirical tests that would more accurately support our findings. Nevertheless,
            the analysis is sufficient to identify the most relevant factors holding back licit (including
            informal) economic growth in Afghanistan, and should therefore facilitate the formulation of
            policy priorities to ignite economic growth. 1

            2.2          Methodology
            This technical report applies Hausmann, Rodrik, and Velasco (2005) Growth Diagnostics‘
            methodology. The framework focuses on what authors call the ―binding constraints to growth‖
            – i.e. the constraints whose removal would have the largest payoff in terms of growth. The
            growth diagnostics framework provides a consistent strategy for identifying the most critical
            or binding constraints to growth, aiming to inform the prioritisation and sequencing of
            policies required to ignite and sustain growth. Interactions between and sequencing of reforms
            are key for the development of policies and strategies. This framework is designed to
            encourage context-driven policy and reform supported by economic analysis, rather than the
            ―laundry list‖ approach, of implementing a generic list of ‗best practice‘ polici es.
            The Growth Diagnostic (GD) framework maps the development process in a decision tree,
            with a series of successive stops (see Figure 1). On each decision node (the tree should be read
            top-down) the researcher assesses the direction of his or her decision until no more nodes are
            left, thus indicating the binding constraint. The model acknowledges developing countries
            suffer from a host of constraining issues, but key to the model is the idea of ranking and
            therefore sequencing. If the most binding constraint is removed, the economy will grow until
            another constraint binds it. However, authors suggest when the economy is constrained; a
            minor economic impact will be recorded as a result of reform unless these reforms are focused
            specifically on the most binding constraints. In fact, Hausman, Rodrik and Velasco consider
            that some reforms not specifically targeted at binding constraints could actually be
            detrimental to growth due to possible ―second-best interactions‖, where the usual benefits of
            reforms are overturned by other factors and the need for a gradual and appropriately
            sequenced approach becomes critical. 2
            The model behind the framework starts by asking why a country exhibits low rates of private
            investment and centres the analysis on the very decision an investor will consider. Starting
            from a standard endogenous growth model, authors derive the equation for balanced growth.
            The balanced-growth path of consumption (c) and capital (k) is given by equation (1).

                                               ct  k
                                                   t   [r (1   )   ]             (1)
                                               ct  kt
            c = per capita consumption
            k = per capita capital
            σ = elasticity of intertemporal substitution in consumption
            r = rate of return on capital
            1-τ = private appropriability of social return
            ρ = cost of financing

                See our analysis on the informal sector in section 4.4 below.
             As examples, partial trade or capital-account liberalization may reduce welfare unless more
            extensive reforms in trade and in financial markets and the institutions surrounding them are
            done at the same time.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Equation 1 summarizes the entire GD analysis (see Figure 1). Or in the authors‘ own words:
            ―An exercise of growth diagnostics simply consists of reviewing and analyzing these factors to
            ascertain which of them are the most binding constraints on growth.‖ Capital accumulation
            and income growth may be limited by an array of factors. First, the rate of return to capital (r)
            may effectively be low due to an absence in the county of complementary inputs like
            infrastructure or human capital, or adverse geography and location vis a vis international
            markets or resources; or may be reduced by appropriability issues. Private returns may be
            reduced by different issues, in the very stylized formulation of equation (1), a high τ. Those
            (subsumed into τ) may be: (A) government failures increasing appropriability risks on the
            microeconomic side of the economy (taxation, corruption, poor property rights and contract
            enforcement, etc.) and the macroeconomic sector (inflation, currency crises and financial
            meltdown, etc.) or (B) market failures such as learning processes, asymmetries or information
            externalities (―self-discovery‖ as labelled in the tree) and coordination failures (i.e., chicken -
            egg problems). Second, it may be that the cost of capital (ρ, interest rate) is higher than the
            expected return and therefore investment is impeded.
                            Figure 1. The HRV Growth Diagnostic framework

            The application of this abstract model to the real life problems faced by Afghanistan is
            achieved by following equation (1) and the GD decision tree (see Figure 1). This research
            analyses cost of financing (ρ), returns and appropriability issues (r and τ), and discards a
            number of potential constraints as the evidence indicates they currently are not the most
            constraining issues preventing private investment in Afghanistan. Discarding a constraint does
            not indicate the issue needs no further attention or reform. This is a central feature of the
            model when applied to a country, like Afghanistan, where almost each of the branches in the
            GD tree (or the coefficients in equation (1)) is identified as constraining. Discarding is the
            elimination method followed by the framework to identify the most binding constraint for
            investment, while relegating other constraints further into the ranking, with the direct
            implication of sequenced policies, by allocating scare resources (i.e., financial and political
            capital and human resources) on the most pressing one.
            Our analysis and the whole ―diagnostic‖ approach is an attempt to slice the most critical issues
            limiting private investment and thus economic growth in Afghanistan. The conclusion and
            recommendations here presented should be read as an effort to narrow down the policy agenda
            and focus efforts and resources on the expectation of the highest social and economic payoff.
            We are certain a comprehensive attempt to simultaneously reform the many imperfections on

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Afghanistan is doomed to fail. The available evidence suggest instead the country should focus
            in four main areas: (i) regain the confidence of large investors and attract formal sector
            investments by reforming tax policy, improving tax administration and ensuring transparent
            and fair enforcement; (ii) reduce the licenses and permits burden by creating a one -shop stop
            for permits (as AISA is for registration) and address the issue of corruption; (iii) focus
            infrastructure investments in the most competitive sectors and products (agribusiness); and
            (iv) secure transit rights and compliance in neighbouring countries, aimed to re ach regional
            markets with Afghan exports.

            2.3        Limitations
            Applying the GD framework to Afghanistan is a challenging task. The post-conflict
            environment creates cross-cutting issues that go beyond the model as they interact with
            several of the model variables in a way difficult to predict. The persistent dominance of
            informality and the importance of the drug economy, exacerbated by the paucity of data add to
            the challenge. The nature and extent of a constraint and its impact on formal and informal
            business is likely to differ. Whilst issues of taxation or bureaucracy would apply directly to
            formal business and indirectly to discourage the formalisation of informal firms, property-
            rights insecurity, corruption and other appropriability issues, or infrastruct ure limitations will
            affect both. Although the diagnostic methodology can be applied in a variety of ways and the
            testing process varies from country to country, it is nonetheless largely dependent on the
            hypotheses being tested, and the scope for testing creativity is severely affected in this case by
            data availability and quality. The methodology explicitly rejects a single approach and favours
            multiple testing at specific decision nodes when possible. The fact that this cannot be done in
            Afghanistan, represents a real limitation as many factors are constraining economic activity;
            complicating the process of ranking constraints.
            What the model does on each variable in equation (1), or its visual representation in the
            decision nodes, is to present the arguments and consider the most salient issues as they apply
            to Afghanistan, according to the available data. When data is available, we try fully testing the
            hypotheses to at least discard with some level of confidence the particular node as a candidate
            for the most important constraint. This process of elimination is in practice the application of
            the model. Following this process we discarded ‗cost of finance‘, ‗low social returns‘, ‗market
            failures‘ and ‗macroeconomic risks‘, concluding that microeconomic risks in the ‗government
            failure‘ branch what is constraining Afghanistan the most. Like all scientific inquiry, each
            hypothesis (i.e., a potential constraining issue) can be rejected, but none can be ―proven‖.

            2.4        Reducing the limitations
            For future similar studies to contribute to a better understanding of the economy and an
            improved policy design, a significant investment in data collection and analysis is required,
            especially to record both formal and informal agribusiness and industry/manufacturing sectors,
            but also to breakdown national accounts (GDP and employment) into sub -national levels.
            Currently only the National Risk and Vulnerability Assessment (NRVA), price and other ad
            hoc surveys are available. While the NRVA was originally designed to improve understandings
            of food security, the NRVA 2005 and 2007 have also provided data on household consumption,
            occupation and other indicators, contributing to a better understanding of the country. A
            business enterprise survey is currently planned with the support of the ADB, yet results are
            not yet available, an important limitation for the application of the growth diagnostic.
            Some of the specific problems we faced when applying the model to Afghanistan are:
                  a)   Perception-based information cannot always be validated or contradicted/contested
                       with the relevant data. This limits our capacity to calibrate and contextualize the
                       qualitative assessments, and to control from private agendas and selection biases in
                       our sources.
                  b)   Statistical series are not available and therefore little econometrics can be done. This
                       reduces the level of confidence in the conclusions presented here, and more

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                       worrisome, it limits our capacity to infer causality. At the same time, cross country
                       comparisons are almost impossible since data is not only scarce but often not
                       comparable internationally.
                 c)    There is little data on the formal economy, but even less for the informal one. For a
                       country where the informal sector represents 80-90 percent of the nondrug GDP,
                       having little information on it limits the scope for research.
            In order to reduce the limitations imposed by the lack of data, a joint effort by the Government
            of Afghanistan and international donors is required. Donors have invested vast amounts of
            money in reports and surveys, but those are rarely used to inform policy making as they are
            shelved and inaccessible. A repository of all work being done in Afghanistan, including most
            data sets should be kept and maintained.
            In a longer perspective, the Statistical Master plan for Afghanistan prepared by a joint Mission
            of the World Bank (WB), the Asian Development Bank (ADB), the International Monetary
            Fund (IMF) and the United Kingdom Department for International Development (DFID)
            should be updated to take into account the pressing needs for data. Data demands arising out
            of the monitoring requirements for the Afghanistan National Development Strategy (ANDS)
            and other requirements should be met or else face the incapacity for government and donors
            fully to understand, calibrate and really contribute via policies to a sustainable rise on income
            and development. Further, sufficient resources need to be made available to conduct regular
            surveys, to update the Master plan, to develop an implementation plan and to support the
            ongoing capacity building efforts of the Central Statistics Office and other agencies.
            The GD does not have a single testing methodology and the data requirements are not
            standards as they vary from country to country and within the country among sectors. In
            general one would require national account data (at national and sub-national levels)
            including economic activity, employment, investment (at sector and provincial level), trade
            and credit. AISA figures should be refined to reflect actual investment, and a panel data sho uld
            be created by surveying those firms registered on an annual basis. Trade data at the most
            disaggregated level, including provincial data export potential should be estimated. A
            provincial level analysis of competitiveness (i.e., a kind of ―Doing Business‖ for each one of the
            34 provinces) would provide valuable information on the specific sectors and its needs at
            provincial level. Also, household and firm-level data is needed. NRVA and especially the
            coming Census are much needed efforts that will hopefully cover most needs at the household
            level. Firm-level data should be produced, explicitly targeting both the formal and informal
            sectors including specific questionaires for both. A credit survey should be conducted for the
            Central Bank, to identify areas, trends and sectors where money is being allocated. Customs
            information on exports, imports and re-exports including as much information as possible on
            destination, volume, price, tariffs, etc., should be made available by Customs. Information on
            migration and remittances, beyond the basic questions included in NRVA should be produced.
            A key effort remains to estimate and identify the large informal (but legal) economy and the
            smuggling economy.

      3     The Afghan Economy, a Literature Review
             This section presents a very brief summary of the rich literature on Afghanistan‘s economy.
             This summary is intended to contextualize our GD exercise and discussion on constrains to
             growth in the Afghan economy. In particular, we will examine the areas of consensus, and
             areas of contestation that derive from the literature review.
             Areas of Consensus : There is broad agreement that Afghanistan economy has been growing
             rapidly since 2001. Whilst important progress has been made in most key areas such as
             infrastructure, health and education, these are far from sufficient in terms of leading Afghanistan
             into a path for a stable and fiscally sustainable future. Basic public services are being provided, but
             are insufficient and totally reliant on foreign aid. Most of these services, particularly in the health
             and education sectors, are being provided by NGOs. Currently, Afghanistan is at the bottom of
             almost every ranking of economic and human development indicators.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

             Double-digit growth rates have been reported, but growth remains inherently fragile as it
             depends—and will continue to depend for the foreseeable future—on international aid and
             military support, illegal trade and drug production. There is agreement on what has been driving
             Afghanistan‘s growth rates subsequent to 2001. The main source of growth in the country is
             international aid. Foreign aid disbursement amounts to almost one half of GDP, and it is the
             main inflow to the country‘s economy. Illegal drug production and trafficking — worth over
             one-third of non-drug GDP—is second to aid as a share of GDP (Byrd, Mansfield, Oldham and
             Ward, 2008). It is also the main basis for income and employment in rural areas, with more
             than three million Afghans employed in cultivation (UNDOC, 2007). Even if the bulk of the
             revenue from this business is assumed to leave the country, large parts of it are again
             reinvested within Afghanistan through construction and consumption activities (Byrd and
             Ward, 2004). The third growth driver in the country and the first in terms of long -term
             sustainability is agriculture (including livestock). Agriculture has been the main sector in
             Afghanistan‘s long-term economic history (Guimbert, 2004).
             Areas of Contention : The three main areas of contention centred on the appropriate policy
             responses towards sustaining economic growth, reducing opium production, and maintaining
             security. The first challenge remains how to achieve sustained rapid growth to reduce poverty,
             create jobs, reduce opium dependence and maintain a functional government active in
             implementing the reforms. Aid dependence, donor-driven assistance, limited state control over
             resources, and both core and external budget assistance are issues of debate, as they present
             issues for long-term state building.
             Fiscal consolidation and continued substantial donor support are required over the medium
             term for security and development (IMF, 2008a). Current government revenues continue to be
             lower than even operating expenditures, raising issues of fiscal sustainability. Government
             spending, especially for development purposes, remained lower than budgeted because of
             fiscal and capacity constraints. Government voracity and corruption are being denounced as
             rampant and rising (IWA, 2008).
             Second, the appropriate policy towards opium is debated. Even it there seem to be agreement that
             alternative livelihoods should be combined with eradication, the proportion of each tactic is in
             disagreement. A few even cite the possibility of legalizing opium production for licit uses (Selenis
             Council, 2007). And although the size of the opium economy is known, the indirect
             mechanisms by which it affects the economy are unmeasured and not properly understood.
             The third area of contention is in the realm of ‗security‘. The security situation is still volatile,
             there is an alarming rise in criminal kidnappings and suicide attacks continued growing in
             2007 (Rubin, 2006a and 2007). This includes not only insurgency, but criminal activity as
             well. Evidence suggests that avoiding further exacerbation of insecurity in rural Af ghanistan
             requires careful balancing of reductions in opium poppy cultivation, security measures,
             governance and economic growth (Mansfield and Pain, 2007).

      4     The Economic Context
            4.1        Growth
            Afghanistan‘s post-2001 growth rates are highly influenced by pre-2001 economic conditions
            and post-conflict recovery. Almost a decade of civil war following the military conflict and
            intervention of the Soviet Union has left a legacy of significant destruction. In spite of some
            fluctuations, it is safe to say that from 1980 to 2001 Afghanistan had no growth or negative
            growth. Guimbert (2004) reports average growth for 1960-1970 of 2 percent, and 3.6 percent
            for 1970-1978. From 1978-1990 Afghanistan real income level decreased 1.8 percent (i.e.,
            negative GDP growth). Based on Maddison (2001) and Penn Word Tables 6.2 (see Output III
            literature review) GDP growth was negative until mid-90s with a small recovering to then
            plummet again with further reductions until 2001. The 1998-2001 periods also included a
            severe drought with a large impact on agriculture, the main economic activity until very

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Post conflict growth rates include a catch up effect as the economy bounces back from
            extremely low GDP levels, posting impressive growth rates. 3 This effect is apparent during the
            first years immediate to the Taliban period, when aid levels and investment was substantially
            below current amounts yet growth rates where over 20 percent. Other post -conflict recovering
            cases suggest that a few years after the end of conflict the catch-up process is still in place, as
            both physical and human capital returns and the government increases its capacity, but their
            impact fades out shortly (World Bank, 2005c; Collier and Hoeffler, 2002). As is to be expected,
            if both physical and human capital are prevented from returning or their contribution is
            limited by a constraining factors, the fading out of the catch-up phase will be faster, an issue
            we believe is happening in Afghanistan at the present time.

            Source: IMF, Art. IV (several years) and World Economic Outlook (see Annex 1 for data
            As Graph 1 suggests, Afghanistan‘s economic performance outperforms other post -conflict
            economies. Since 2002 (t=2 in Graph 1 for Afghanistan), economic activity has expanded
            rapidly as a consequence both of the post conflict recovery, a demand impulse based on donor
            financed spending (especially since 2004/2005, t=4 in Graph 1) and drug money. Now that the
            catch-up effect is fading out, growth in the medium and long term will depend to an ever
            greater extent upon improvements in governance and a successful integration into the world
            economy, while foreign aid will remain a key contributor. Private investment had consistently
            grown since 2002 and reached a peak in 2005/2006, reflecting untapped e conomic
            opportunities of the country and an improved macroeconomic stance (see Section 4.2).
            Legal economic activity has been stimulated by the large influx of international aid,
            accompanied by the return of many skilled people and businesspeople from the Afghan
            Diaspora, crowded in by the large reconstruction budget. 4 A limited number of large foreign
            direct investments for example in the telecoms sector and other services have also contributed
            to growth and productivity. Exports (official) have been growing rapidly, more so than the real
            GDP (see Table 1), even if the country has developed no new exports sectors outside the

              This effect is also known as the ―peace dividend‖, but since no peace has being achie ved in
             Afghanistan, we will resist such concept. Instead, the catch up should be understood as
             produced by the combination of extremely low GDP levels and large inflows of aid.
          One-fifth of the firms surveyed by the World Bank Investment Climate Asses sment in 2005 were
        owned by investors who returned to the country after the fall of the Taliban regime, most of them
        focused on the more secured donor-funded projects than in long term investments, proving the
        ―crowding in‖ effect of donor inflows.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

               traditional sectors of carpets, minerals and horticultural products (see Section 5.3.1). 5
               Vulnerability to drought was responsible for lower GDP growth in 2006/07, although growth
               is expected to rise based on a post-drought rebound in agricultural production (see Table 1).
               Moreover, a large proportion of GDP is accounted for by foreign aid and assistance. Large
               inflows of aid, coupled with inflows of hard currency from the opium economy has led to a
               relative appreciation of the Afghani, and higher costs of labour and other inputs for Afghan
               producers, undermining the competitiveness of the manufacturing sector. However, the
               manufacturing sector has been able to keep up with the rest of the licit economy in terms of
               growth. Non tradable sectors in general have been growing more than the rest of the economy.
               However we do not believe this reflects an exchange rate misalignment but mostly is a
               consequence of reconstruction efforts that given its nature are biased towards services
               (construction, transport and telecommunication, electricity and power, and general public
               administration services).
               Table 1 shows the GDP growth and per capita evolution since 2001. Except for the 2004/2005
               and 2006/2007 drought-induced slowdown in agriculture the transitional economy has grown
               rapidly posting double-digit growth rates. The ADB reports growth in the 2007/2008 reached
               13.9 percent (ADB, 2008). Assuming the externally financed development efforts are sustained
               and that climatic factors do not affect agricultural production, the economy is expected to
               grow between 8.4 and 9 percent during 2008/2009 period (IMF, 2008a and ADB, 2008). But
               given the size of the informal economy these numbers appear to underestimate real economic
               performance (World Bank, 2005c). Accurate data is lacking but estimates of Afghanistan‘s
               GDP per capita is $335 (IMF, 2008a) with an estimated 70% of the population living below the
               $2 a day poverty line (DFID, 2006).
                     Table 1. Afghanistan, Real GDP (annual percentage change), Official exports
                     (annual percentage change) and GDP per capita (US dollars)
                                                           Estimates            Projections
                                   2002/   2003/   2004/    2005/   2006/    2007/   2008/    2009/   2010/
                                   2003    2004    2005     2006    2007     2008    2009     2010    2011
                     Real GDP      28.6    15.1    9.4      16.4    6.1      13.5    9.0      9.0     8.0
                                          43.6    112   26.4        7.9      13.8    25.7     22.1    20.1
                   Per     Capita
                                  182     199     252   300         335      383     433      …       …
                  Source: IMF (2008a, 2008c) and WEO (2008)

               4.2        Investment
               Licit investment peaked in 2006/2007 driven by donor-funded government investment. But
               private investment declined in the same period, with 2005 marking the inflexion point (see
               Table 3). Also, the Afghanistan Investment Support Agency (AISA) reported a decline in the
               number of new businesses registrations during the same period (see Table 2). Notice further,
               in 2005 the number of registered companies (with AISA) decline d for both domestic and
               foreign, and in 2007 the number of foreign companies declined further almost 50 percent in one
               year. Similarly, the IMF (2008c) reports foreign direct investment reaching in 2005 a peak at
               4.2 percent of GDP, reduced to 3.4 percent in 2006/07 and projected at 3.3 percent until 2009.
               Based on AISA registrations, only services and construction seem to be attracting more
               investments, especially from within the country and by 2007, most sectors have not yet
               reached the 2005 level (see Table 2). Because this variable is the key observable for the model,
               we take this decline and stagnation in private investment very seriously, and we seek to
               explain what is behind this decline.
                 Table 2. Total Number of companies registered with AISA (as of December

            Some new products are now exported, but all related to the agribusiness sector.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

             Sector                            2003 2004 2005 2006 2007                      Total
             Total Number of companies         3.243 1.386 2.503 2.046 2.096                 11.274
                Domestic                       1.474 1.137 2.123 1.746 1.925                 8.405
                   Construction                 596     554     1.309 1.048 1.249            4.756
                   Industry                      611    401      546      422      291        2.271
                   Agriculture                   118     33       48       45       35         279
                   Service                      149     149      220      231      350        1.099
                Foreign                         101     249      380      300      171        1.201
                   Construction                  53      94       111      93       65         416
                   Industry                      24      34       68       47      22          195
                   Agriculture                    0       3        6        3        5          17
                   Service                       24     118      195      157       79         573
            Source: AISA
            *Fifty percent of the companies are currently in the inactive list (AISA)
            Projections from the IMF suggest that private investment will not recover until 2009/2010 and
            coupled with foreign direct investment they will replace a decline o f donors‘ financed public
            investment of 16 percentage points of GDP by 2010/2011 (see Table 3) 6. Yet, under these
            projections, the highest level of FDI projected is 4.1 percent of GDP by 2012/2013, still below
            the 2005/06 level. Also, assumptions behind this projected improvement in private investment
            may be overly ambitious, as they presume that lack of electricity in Kabul 7 and the ―precarious
            security environment‖ will be improved, and rely on licit agriculture (to benefit from the
            ―donor-supported strategy‖ in favour of opium-alternative crops), FDI-driven mining projects,
            trade and services as drivers of growth. Based on our analysis, there is little space for such
            optimism in the short-term.
            As will be elaborated later on the issues of appropriability in the form of uncertain and
            discretionary taxation (especially from the Large Taxpayer Office on firms with foreign
            capital) on the formal sector, increased corruption and property rights issues should be
            addressed for private investment to grow. Even more, we consider the environment for
            investment may further deteriorate due to increasing levels of political and security risk, and
            the uncertainty surrounding the Presidential election in 2009 may exacerbate these factors.
             Table 3. Afghanistan, investment and savings (over GDP)
                                                                      Estimates               Projections
                                      2002     2003         2004    2005 2006       2007    2008 2009         2010
                                      /        /            /       /      /        /       /        /        /
                                      2003     2004         2005    2006 2007       2008    2009 2010         2011
             Gross domestic
                                      28.8     33.0         46.1    44.6    46.0    40.3    38.6      33.6    31.8
             Non-government           1.2      1.3          9.4     9.6     8.9     8.9     9.0       9.5     10.6
             Government               27.6     31.7         36.7    35.0    37.1    31.4    29.6      24.1    21.2
             Gross domestic
                                      25.1     22.7         41.2    41.8    39.7    38.9    38.1      32.4    28.8
             Non-government           30.4     28.3         46.6    45.4    43.9    43.0    41.7      35.8    30.7
             Government               (5.3)    (5.6)        (5.4)   (3.6)   (4.2)   (4.1)   (3.6)     (3.4)   (1.9)

              Since GDP is expected to grow faster than AID, AID/GDP is expected to decline (IMF,
             A transmission line from Uzbekistan to Kabul is likely to be completed in 2009/2010, yet the
            power purchasing agreement with Uzbekistan has yet to be concluded and the distribution
            system in Kabul is not yet in place.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

             Source: IMF (2008a)

            4.3         Regional Trading Context
            Since 2001, trade has been a very active sector. Total trade in 2007 (exports plus imports) was
            estimated over $8.5 billion (excluding opium), comprised of $6.7 billion in imports and $1.8
            billion in exports, of which $1.2 billion were re-exports and $150 million were smuggling (IMF,
            2008a). In five years, licit trade (official exports) has grown over 300 percent, with volume
            increasing four-fold. Official imports are equally impressive, growing by a factor of three
            during the same period, mostly due to reconstruction efforts (see Table 4).
            Table 4. Afghanistan: Balance of Payment, 2002/03-2006/07 (In millions of U.S.
                                                    2002/03     2003/04     2004/05    2005/06     2006/07*
            Current account (including grants)         -149.8     -456.2       -265       -182.5         -444
            Current account (excluding grants)        -1371.6    -2932.1      -3894     -4880.3        -5435.5
            Trade balance                             -1217.6    -2485.7    -3443.3     -4335.3        -4941.6
            Exports of goods (f.o.b.)**               1290.6      1893.6       1643      1794.8         1801.3
            Official exports                            100.1      143.7       305.3      385.9          416.5
            Unofficial exports                         1190.5     1749.9      1337.7     1408.9         1384.8
            Smuggling                                   176.6      232.8       152.7       173.7         149.9
            Transit trade                              1013.9      1517.1     1185.1     1235.2         1234.9
            Imports of goods (f.o.b.)                -2508.2    -4379.3     -5086.3     -6130.1        -6742.9
            Official imports                          -1983.1    -3282.9    -4415.8     -5481.7        -6049.2
            Of which: Duty free                       -840.8     -1594.4     -2416.2    -3258.3        -3579.4
            Smuggling                               -525.1 -1096.4        -670.5      -648.4      -693.7
            Source: IMF (2008a)
            *Preliminary estimate
            ** Excludes opium exports and, due to limited data availability, flows associated with U.S.
            Army and most ISAF activities.

            By South and Central Asian standards, Afghanistan has a very open economy. The country now
            has 12 tariff rates, ranging from 2.5% - 40% (see Table 5). This is dramatically different from
            the situation in 2002, when Afghanistan had an average tariff rate of 40% and a maximum rate
            of 150%. The country has recently joined the South Asian Free Trade Agreement area, and it is
            pursuing other bilateral agreements which will further increase the amount of trade
            opportunities with its neighbours. The process to gain full accession to WTO is expected to
            start in 2010.

             Table 5. Summary of Import Tariff Schedule (as of January 10, 2008)
                            Reform (Aug-06)                 Revision (Mar-07)                Revision (Jan-08)
             (%)                                    Type of product (Broad categories)
                0                                                     New Machinery, wool, wool dyes
                                                         Industrial raw materials        Raw Materials with no dual
                                                        and intermediate products                 purpose
                        Essential food Products and
               2.5                                                        Essential food products
                              Raw materials
               3.5                                                            Cooking oils *
                5                       Nonessential food products, intermediate goods and chemicals

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                               Fuel products
                        communications equipment,                 Kerosene lubricants, communications equipment,
                           leather products and                       leather products and (bathroom fixtures
                             bathroom fixtures
                10                                              Manufactured products
                12                                                           Petrol diesel, aviation fuel
                16                        Non-priority products, luxury products, arms and ammunition
               20                                                               Mineral water, marble
               25                                                                  Motor vehicles
               40                                                            Non-alcoholic beverages **

             Source: IMF (2008a) using official data from Ministry of Finance
             * 3.5 percent tariff band was created in 2006 on a temporary basis, to be reduced to 2.5 percent in 2008
             ** To be eliminated by March 31, 2009

            Trade statistics were reported by Afghanistan only until 1977 which make any long term
            analysis almost impossible. To present a long view of Afghan external market we used data
            reported by Afghanistan‘s trading partners. In other words, we consider as Afghan exports all
            the imports other countries report as deriving from Afghanistan. Similarly, Afghan imports are
            estimated by summing all exports reported from third countries to Afghanistan. Graph 2
            present four decades evolution on Afghanistan‘s external sector in real 2000 dollars (please see
            Section 5.3.1 for discussion on the limitations when using this data). The huge and
            unsustainable surge of imports is noticeable. One key component of the external market is the
            large dependency on regional countries, today Pakistan (34 percent) and India (20 percent)
            account for over one-half total exports, Russia, the United Arab Emirates and Iran follow as
            the main destinations. This was not the case in the past when the U.S., the U. K. and Germany
            received more than 70 percent of Afghan exports. See section 5.3.1. for further analysis.

            Graph 2. Afghanistan, Export and Imports (Real U.S. 2000 dollars)

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                                                                   Total Export and Imports


          Real US$ Million



















                                                                                           Imports                    Exports
                             Source: COMTRADE, 2008
                             Note: To avoid distorting the imports, special operations (SITC Section 9) were ignored. B y
                             construction, only recorded (legal) trade is considered. Data does not include re-exports.

                             4.4          Sector analysis
                             A sector analysis in the post-Taliban period shows services as the main sector (38 percent of
                             GDP in 2006), followed by agriculture (32 percent) and industry (27 percent) (IMF 2008a).
                             Services have been growing steadily since 2001 and in 2004 surpassed agriculture as the main
                             contributor to GDP (see Graph 3). This is driven by the aid injection in the public sector and
                             key investments, in particular in telecommunications, restaurants and aviation and other sub -
                             sectors that donor activities and the presence of the international community tend to increase
                             demand for. In particular, telecommunications is the star performer of the Afghan economy,
                             accumulating the lion‘s share of private investments (see Table 6).

                             In the agricultural sector performance has not been so successful, but mostly due to exogenous

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            factors like the recent drought that affected cereal production and also the interaction
            between legal agriculture and opium production. Agriculture‘s share of GDP has been
            diminishing since the mid-90s, but even after a period of drought causing negative 20 percent
            growth in the sector in 2006/2007 (IMF, 2008a), agriculture still accounts for over one-third of
            Afghanistan‘s GDP and it accounted for almost one-half of Afghanistan‘s exports in 2006 (IMF,
            2008c). It is also the main occupation and mean of subsistence of the vast majority of the
            population, although livelihoods reliance on agriculture as single source has dropped as
            households in some areas have responded to drought by diversifying their livelihoods
            strategies (FEWSNET, 2007).

                   Table 6. Afghanistan, GDP at constant prices, annual percentage change
                                                           2003/     2004/    2005/    2006/
                                                           2004      2005     2006     2007*
                   Agriculture                             16.9     -17.1    6.7      -21.1
                   Cereals                                 15.9     -20.7    9.5      -15.3
                   Other crops and non-foods               71.6     -5       …        …
                   Livestock                               1.3      1        -7.3     2
                   Industry                                11.9     32.4     23.9     21.3
                   Mining                                  -4.2     26       17.7     10
                   Manufacturing                           7.4      21.9     19.5     16.4
                   Construction                            25.8     60       32.2     30
                   Electricity and power                   132.8    69.9     20       1.3
                   Services                                13.8     34.6     14.6     18.5
                   Trade                                   -5.5     30.1     6.7      14.1
                   Transport and telecommunications        41.1     50       10.5     26.6
                   Public administration                   14.2     24       59.9     21.3
                   Other services                           1.5       15.6      ...       ...
                   Source: IMF 2008c (Fund staff estimates using Afghan authorities data)
                   Note: Afghan national accounts numbers include opium production in the
                   agricultural sector.

            Next to agricultural activities, livestock represent the largest employment activity (30% of
            Afghan households). It constitutes the primary employment activity for the Kuchis population
            (81%). In addition, livestock products are the second largest export group after agricultural
            products (11.3% of all exports in 2006).
            Cultivation of poppies remains the largest and most important industry in Afghanistan
            (UNODC, 2007). Over 4 percent of the agricultural land (193.000 ha) is devoted to opium
            farming and some 3.3 million Afghans are involved in opium cultivation. Opium income to
            farmers was $760 million in 2007, while potential exports of opium were $3.1 billion during the
            same year (Byrd, 2007). However, farmers are receiving a lower percentage of the potential
            value of the crop, from 48% in 2002 to only 24.5% in 2006. This could imply that mor e
            processing is being conducted in Afghanistan, or that lower prices are being paid to farmers. In
            addition, the share of potential exports of opium to total GDP declined from 62.4% in 2002 to
            44.0% in 2006. 8 Opium production has been increasing rapidly in Afghanistan, from 3,400
            metric tons in 2002 to 8,200 metric tons in 2007 (UNODC, 2008). The Afghan opium crop now

                Using opium figures from Byrd (2008) and GDP figures from the IMF

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            constitutes 93% of the word‘s total opium output. Despite this growth in output, the number
            of provinces growing opium has decreased from a high of 34 in 2004 to 21 provinces in 2007,
            most of which are located in the south of the country (see Table 7). Cannabis production is
            increasing, with some 70.000 hectares cultivated in 2007 (UNODC, 2007). Farm-gate value of
            opium accounts for 13 percent of official GDP, whilst the whole opium/heroin sector amounts
            to 35 percent. Once opium is taken out of the official agriculture figures, legal agriculture is
            reduced to 19 percent of official GDP and it becomes the sector with the smallest contribution
            (although we know there is a large informal, worth ~30 percent of official GDP, subsistence
            agriculture and livestock sector, see The Informal Economy below).9
             Table 7. Summary Statistics on Afghanistan’s Opium Economy
                                            199    200     200    200     200    200    200     200    200
                                              5      0      1       2       3      4      5       6      7
             Production (tons)              2,30   3,30           3,40    3,60   4,20   4,10    6,10   8,20
                                              0      0              0       0      0      0       0      0
             World Market share (%)         ~52     70     11      74      76     87     87      92     93
             Number of provinces
                                             8      22     11      24     28      34     26      28     21
             producing opium
             Area under opium poppy
                                             54     82      8      74     80     131     104    165     193
             (thousand ha)
             As % of total agricultural
                                            n/a    n/a     n/a    n/a     1.6     2.9    2.3    3.65   4.27
             Area under poppy/Area
                                             2      3.2    n/a     3.2    2.8     5.9    n/a    n/a     n/a
             under cereals %
             Gross farm income per ha       1,00   1,10    7,40   16,2    12,7   4,60   5,40    4,60   5,20
             (US$)                            0      0       0     00      00      0      0       0      0
             Gross potential value of                             2,50    2,30   2,80   2,70    3,10
                                            n/a    850     n/a                                          n/a
             opiate exports (US$ million)                           0       0      0      0       0
             Gross farm income from
                                                                  1,20    1,00                         1,00
             opium                           50     90     60                    600     560    760
                                                                    0       0                            0
             (US$ million)
             Downstream income in
                                                                  1,30    1,30   2,20   2,14    2,34
             Afghanistan                    n/a    760     n/a                                          n/a
                                                                    0       0      0      0       0
             (US$ million)
            Source: Byrd (2008) based on UNODC and Government of Afghanistan
            Industry, driven by construction, but also manufacturing, has been one of the most dynamic
            sectors during the last five years, accounting for over one-quarter of 2006 GDP. Construction
            is one of the main driving forces behind the growth of the sector, and was the most dynamic
            sub-sector in 2006/2007, with 30 percent growth (IMF, 2008b). This reveals the important
            resources devoted to infrastructure reconstruction, but also how the large amounts of drug
            money are reinvested in real estate in Afghanistan (the rest is invested outside the country).
            Textiles, clothes and leather represented the second largest sub-sector, but it was the largest
            sector in terms of exports (including carpets and skins). There has been growth in
            manufacturing of food and beverages, of which beverages are likely to take the lion‘s share,
            with a large investment by a Dubai-based Afghan businessman in a Coca Cola factory in 2006
            and with the existence of other smaller bottled water factories in Kabul.
            It is worth noting that even in the presence of strong foreign inflows the manufacturing sector
            has been able to keep up with the rest of the economy in terms of growth, supporting the idea
            that exchange rate management has been adequate or in general the benefits associated from
            investment and stability that go directly to reinforce country‘s competitiveness. However,

              The authorities' national accounts numbers include opium production in the agricultural
            sector (IMF, 2008c).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            growth in all these sectors is influenced by the very low levels from which the sectors sta rted,
            following a process of deindustrialisation through conflict that, in the last year for which data
            is available, 1995, had left Afghan industry with only 10 percent of GDP. 10
            The Informal Economy
            Is been reported 80-90 percent of country‘s GDP is informal, including 35 percent of illegal
            opium related activities and an array of other informal activities (World Bank, 2005c). No
            accurate estimations exist, but more than one-half of this 50 percent black economy (non-
            opium) is subsistence level agriculture and livestock (World Bank, 2005c).
            The black economy includes also barter trade, manufacturing and commerce and other non -
            recorded services (e.g., the hawala system or electricity from small-scale generators), but also
            smuggling and re-exports. The smuggled exports include gravels and construction materials,
            gems and semi-precious stones, logging and other non-recorded exports, but this has decrease
            from 5 percent of official GDP in 2002/03 to 2 percent in 2006/07 (see Table 4). Smuggled
            imports were estimated at US$690m in 2006/07, but the contribution from that to the GDP
            should only include the value added generated in the country and consider the lost in tariffs
            revenues. We roughly estimate this to be around 2.5 percent of legal GDP in 2007. 11 A large re-
            export activity exists between Afghanistan and neighbouring countries (mostly Pakistan), but
            the value added from this activity and hence its contribution to the Afghan GDP is also very
            low, probably between 2-3.5 percent of official GDP in 2007. 12 Based on these rough
            approximations, the whole smuggling (imports and exports) and re -export sector would
            represent no more than 8 percent of official GDP in 2007.
            The main components in the informal economy are: non-recorded agriculture and livestock
            (~30 percent of official GDP, World Bank 2005c), opium production and processing activities
            (35 percent of official GDP, World Bank 2005c), illegal trade (6.5-8 percent of official GDP, as
            per our estimations) and other non-recorded activities (including small manufacturing and
            The official figure from the government for unemployment is 40 percent (ANDS). But little
            more is known about the size and composition of the labour market in the country. Using
            NRVA 2005, we estimated the importance of different sectors in the occupational profile of the
            economy (see Annex 2 for the list of occupation and a description of our aggregated
            definitions). Table 8 present the results. At national level, 66 percent of the households depend
            on agriculture (including opium 13) and livestock as the main source of income, this number is
            66 in rural areas and 14 in the cities. Trade and services employs 15 percent of the national
            population but is –as expected—the main source in urban areas (64 percent as first source, 14
            percent as second source). Manufacturing is just a 1 percent at national levels, with 3 percent
            in urban areas, being the third largest occupation.

            Table 8. Afghanistan, occupational profile (% of households reporting the activity as first

              This is at the lowest world levels, comparable to Myanmar, Liberia or Ethiopia in that year.
              Assuming a (high) margin of 30 percent as value added generated by smugglers in
            Afghanistan, and deducting the lost in tariffs collection (assuming 5.3 percent average tariff),
            the sector would amount to 2.3 percent of official GDP in 2007.
               Assuming an average tariff of 5.3 percent is paid and 10 percent margin for re -exporters
            (assuming they are Afghans), the contribution to GDP from this activity was 2.5 percent in
            2007. An upper bound of 20 percent including margin and tariffs contribution could be set at
            3.3-3.5 percent.
              NRVA is based on self-reporting and therefore all figures relative to opium are considered to
            be underestimated.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            or second most important source of income)
                                                                 Total Population
                              Livestock   Agriculture   Opium   Trade       Manf    Remittances   Other    Total
         Livestock                3%          3%         0%       1%          0%        0%          2%       9%
         Agriculture             17%         12%         2%       6%          1%        2%         13%      53%
         Opium                    0%          2%         0%       0%          0%        0%          1%       4%
         Trade and Services       1%          4%         0%       5%          1%        1%          3%      15%
         Manufacturing            0%          0%         0%       0%          0%        0%          0%       1%
         Remittances              0%          2%         0%       1%          0%        0%          1%       4%
         Other                    2%          5%         0%       3%          0%        1%          2%      14%
         Total                   24%         28%         2%      17%          3%       5%         22%      100%

                                                                 Urban Population
                              Livestock   Agriculture   Opium   Trade       Manf    Remittances   Other    Total
         Livestock               0%           0%         0%       0%         0%         0%          0%       1%
         Agriculture             3%           4%         0%       3%         1%         0%          2%      13%
         Opium                   0%           0%         0%       0%         0%         0%          0%       1%
         Trade and Services      1%           3%         0%      40%         6%         2%         12%      64%
         Manufacturing           0%           0%         0%       2%         1%         0%          1%       3%
         Remittances             0%           0%         0%       0%         0%         0%          0%       1%
         Other                   0%           1%         0%       9%         2%         0%          5%      18%
         Total                   4%           9%         1%      54%        10%        2%         20%      100%

                                                                 Rural Population
                              Livestock   Agriculture   Opium   Trade       Manf    Remittances   Other    Total
         Livestock                1%          3%         0%       1%          0%        0%          2%       8%
         Agriculture             19%         13%         2%       6%          1%        2%         14%      58%
         Opium                    0%          3%         0%       0%          0%        0%          1%       4%
         Trade and Services       1%          5%         0%       2%          0%        1%          2%      11%
         Manufacturing            0%          0%         0%       0%          0%        0%          0%       1%
         Remittances              0%          2%         0%       1%          0%        0%          1%       4%
         Other                    2%          5%         0%       3%          0%        1%          1%      13%
         Total                   24%         31%         3%      13%         2%        5%         22%      100%

                                                                 Kuchi Population
                              Livestock   Agriculture   Opium   Trade       Manf    Remittances   Other    Total
         Livestock               33%          7%         1%      2%           1%        1%         13%      58%
         Agriculture              8%          4%         1%      1%           0%        0%          4%      18%
         Opium                    0%          1%         0%      0%           0%        0%          0%       1%
         Trade and Services       3%          0%         0%      1%           0%        0%          1%       5%
         Manufacturing            0%          0%         0%      0%           0%        0%          0%       0%
         Remittances              1%          0%         0%      0%           0%        0%          0%       2%
         Other                   11%          1%         0%      2%           0%        1%          0%      16%
         Total                   56%         14%         2%      7%          1%        2%         19%      100%

            Source: Authors calculation using NRVA 2005
            Note: To read the tables, please consider:
                Sum of the each row gives the % of households engaged in the category as a first source
                Sum of each columns (minus) row/column observation gives the % of households engaged
                 in the category as a second source of income
                The number in the same column/row combination is the % of households engaged in that
                 as a single occupation (no second source of income)
                Total occupation in one category: sum row (plus) sum column (minus) row/column

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            4.5          The Investment Climate
            In 2005, the Investment Climate Assessment (World Bank, 2005d) identified electricity,
            access to land and corruption as the three most important barriers for business i n the country
            (see Table 9). Again, it is clear than many of the issues captured by the assessment in 2005 are
            still unresolved, and although some of them will not be discussed within the GD framework, it
            should not be assumed that we consider these factors to be unimportant.
            Since 2005, some progress has been reported on electricity, and little on land-related issues
            (i.e., other than the industrial parks), but we found no evidence of the situation being worse.
            Because we are trying to explain the reduction on private investment, we cannot use a
            ―constant‖ as explanation, and access to land and electricity problems have been constant ly
            bad or even improving but not worsening since 2001. On the other hand, both the IMF (2008a)
            and the governance indicators from the World Bank (2006) highlight how the cost of doing
            business is now compounded by a ―difficult institutional environment‖ (IMF, 2008a) that has
            weakened in the areas of political stability, government effectiveness, and the rule of law.
             Table 9. Overall and sector-specific ranking on obstacles to business, the World Bank Investment Climat
             Assessment, 2005
                                                                         Other                      Construction
                                    Overall    Manufacture     Traders              Construction                    Transpor
                                                                         Services                   Materials
             Electricity            64    1     79        1    57   3    65    1     45       4      76        1    41     5
             Access to Land         60    2     57        2    70   1    55    3     58       2      67        5    69     2
             Corruption             53    3     45        4    66   2    45    4     62       1      72        4     71    1
             Access to Finance       51   4     52        3    54   4    56    2     53       3      75        2    27     11
                                    34    5     34        6    28   13   30    7     45       5      43        9    41     6
             Customs and Trade
                                    33    6     38        5    49   5    17   10     40       6      54        7    43     4
             Regulatory Policy
                                    32    7     31        8    35   10   32    6     21       9      62        6    40     7
             Tax Rates              32    8     33        7    40   8    33    5     20      11      47        8    37     8
             Tax Administration     27    9     29        10   40   7    16    11    17      13      33        10   44     3
             Telecommunications     26    10    26        11   33   11   18    9     21      10      24        11   29     10
             Transportation (25, 11); Macro Uncertainty (18, 12); Cost of Finance (17, 13); Crime, Theft and Disorder (16, 14)
             Business Licensing and Permits (11, 15); Legal System (10, 16); Skill and Education of Labour Force (9, 17)
             Labour Regulations (5, 18).*
             Source: Investment Climate Assessment (World Bank, 2005d)
            *In the benefit of space, only the most important ten barriers are detailed. For the rest,
            percentage of respondents voting the barrier and its ranking are presented.

      5.     Growth Diagnostic
            The GD differs from the comprehensive policy reform approach of the Washington consensus
            as it seeks to pinpoint the two or three highest level binding constraints faced at a country
            level. Thus, it allows for a more practical and realistic approach to tackling the problem of
            economic growth - or the lack thereof - as opposed to imposing wholesale reforms that are
            essentially driven by ideology rather than informed analytical diagnostic. The model is also
            being used increasingly to guide and advise donors in aid sector allocation and sequencing
            within the framework of the Paris Declaration. As HRV (2005) quote, ―by focusing on the one
            area that represents the biggest hurdle to growth, countries will be more likely to achieve
            success from their reform efforts.‖

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            The primary question that the model attempts to address is why private investment is low, in a
            given context. In the Afghan case, we consider the change in investment trends since 2005 as
            the key period for analysis. The GD attributes this to one of two issues; either the country has
            a high cost of finance or low returns to economic activity. In the next sections we perform the
            GD analysis for Afghanistan, starting with the cost of finance and then moving into returns to
            the economic activity.

            5.1        Finance
            The cost of finance in Afghanistan is not excessively high. The International Financial
            Statistics (IMF) reports 18.8 percent as the relevant lending rate in the third quarter of 2007. 14
            Our informers in Kabul mentioned interest rate fluctuating from 15 -30 percent, depending on
            firm size and covering large to microcredit firms. These numbers are high, but not out of range
            for a post conflict and low income country. This level is in line with El Salvador a few years
            after the peace agreement (16-19 percent, WDI 2008), and better than Sierra Leone in 2005 (24
            percent, WDI 2008), it is also similar to countries with under $300 per capita income, and
            even close to Iran (16 percent, WDI 2008) even though the country has nine -times Afghan per
            capita income.
            In April 2008, the IMF reports the average Afghani lending rate was about 20 percent, while
            that on Afghani deposits was 7 percent (IMF, 2008a). This high interest rate spread is a
            consequence of the general cost of doing business in the country and the small scale of the
            financial sector. It also suggests lack of competition in a dynamic but still anaemic banking
            sector. The financial system is basic, exclusively urban and provides simple instruments. ATM
            machines are scarce and credit cards do not exist. According to the Investment Climate
            Assessment (World Bank, 2005d), only one-third of the surveyed sample had a bank account
            and only three firms (less than one percent) reported having a bank credit in 2005. In 2005, the
            country had four private banks, three state-owned and four foreign-owned subsidiaries
            (compared with six banks in 2001), but by end-2007, state-owned banks remain the same
            whilst the number of private banks had doubled accounting now for 60 percent of total
            financial system assets and 64 percent of total deposits. Total system assets account for 15
            percent of nondrug GDP in September 2007. Banks had 171 branches in 20 provinces, but
            exclusively in urban areas 15 and serving mostly donors and the international community. By
            end-2007 over three-quarters of total deposits and loans were denominated in U.S. dollars. All
            interbank deposits are denominated in foreign currencies (80 percent in U.S. dollars) (see
            Table 10).
            There are other nonbank financial institutions (15 microcredit licensed institutions, one credit
            union and one leasing company). The micro-finance institutions, working through the
            Microfinance Investment Facility (MISFA) do provide loans for asset building to small scale
            enterprises. The MISFA website recorded a total of US$252.8 million in loans disbursed to
            314,208 active borrowers by May 2007. Loans can be delivered to individuals with or without
            collateral, to groups with or without collateral and may or may not contain a savings
            requirement. Interest rates for micro-loans are around 25-30 percent.
            Even if the cost of lending is not out of range, access to formal finance is very limited. This is
            noted in the World Bank Investment Climate Assessment when businessmen report ‗cost of
            finance‘ among the least important barriers, but ‗access to finance‘ as barrier number forth.
            This is due to market imperfections and not due to a lack of savings or credit rationing (see
            Table 3). Banks prefer to keep their business in a more secure realm by providing services
            mostly to the international community and local elites, or lending for trading activities with
            secured cash collaterals. In the GD language, the inadequacy of finance is due to low
            intermediation in the financial sector (see Figure 1).

               IFS reports lending rate in domestic currency to be 17.9%, 17.5%, 16.2% and 18.8% (per
            annum) for Q42006, Q1-Q32007 respectively.
               DAB has 75 branches, 14 of which are in Kabul. The vast majority of DAB‘s branches offe r
            commercial banking services (IMF 2008a).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            In the absence of a risk profile, borrowers are forced to provide collateral, often way above the
            value of the loan. 16 If collateral is available, borrowers are requested to pay an extra 12 percent
            cost on top of the interest rate as processing fees. 17 And even in these cases, banks are often
            reluctant to accept collateral, as the enforcement mechanisms and the judicial and court
            systems are unreliable due to corruption, inefficiency and lack of clear property rights
            especially in land and other immobile assets.

                  Table 10. Afghanistan, the financial sector
                                    2004/05           2005/06                 2006/07             Sep. 2007
                                                                 Total number of:
                  Banks                  11                  12                    15                   16
                  Private                4                   4                      7                   8
                                         3                   3                      3                   3
                  owned                  4                   5                      5                   5
                                  US                  US                    US                   US
                                              Share               Share                 Share                Share
                                  Mill                Mill                 Mill                 Mill
                  Banks           388          100    614          100     1,081         100    1,318         100
                  Private         82          21.1    238         38.8     629          58.2    789          59.9
                                  212         54.6    226         36.8     224          20.7    229          17.4
                  owned           94          24.2    150         24.4     228          21.1    300          22.8
                                  US                  US                    US                   US
                                              Share               Share                 Share                Share
                                  Mill                Mill                 Mill                  Mill
                  Banks           182          100    394          100      812          100    1,022         100
                  Private         51          27.8    194         49.2     520           64     652          63.8
                                  47          25.8    65          16.5      74           9.1     84           8.2
                  owned           84          46.4    135         34.3      218         26.8    286           28
                 Source: IMF (2008a)
            According to the Investment Climate Assessment (World Bank, 2005d), only three out of 338
            firms in five major cities in Afghanistan (Kabul, Kandahar, Mazar-e-Sharif, Jalalabad and
            Herat) reported having bank credit for new investments. Yet, over 70 percent of the surveyed
            firms reported new investments during the previous year. When compared to traditional
            sources of credit, the formal system is simply irrelevant. Seventy -eight of all investments
            reported by the survey where financed with investors‘ own resources, while family and friends
            appear as the second most important source (13 percent from family in Afghanistan, 3 percent

               Some banks will value the collateralized assets with a 50 percent discount, for an implicit
            collateral requirement of 200 percent.
               Expanding the Outreach of Financial Services: The Experience of Financial In stitutions, Case
            study prepared for the Enabling Environment Conference, Aga Khan Development Network,

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            overseas). Overall, private sector investment is almost exclusively self-financed, with only 1
            percent of new investments being financed with bank credit compared to nearly 94 percent of
            new investment funded by either internal funds or money from family and friends. Informal
            credit mechanisms and financial services for larger scale businesses are poorly understood, but
            the World Bank Investment Climate Survey (2005) and anecdotal evidence suggests they may
            be significant. The role of the opium economy in providing finance for business directly and
            indirectly is a murky and difficult subject. Thompson (2006) has shown a link between opium
            and liquidity in the Hawala (money transfer) system, on which many businessmen who se main
            activities are licit, are also dependent for financial services and credit.
               Table 11. Source from which the household got the largest loan last year
                                     Family / friends
                          In                Outside                 Shopkeeper/          Local    land
                          Afghanistan       Afghanistan             traders              owner
               Rural           66%                 2%                      26%               3%
               Urban           82%                 2%                      10%               1%
               Kuchi           81%                 1%                      16%               1%
               Total           68%                 2%                     24%                3%
              Source: Authors calculations from NRVA 2005
            At the household level, only 8 percent (8 percent rural, 11 percent urban) of total surveyed
            population responded that it ―would not be able to access any credit‖ (NRVA 2005). Thirty-
            eight percent (41 percent rural, 10 percent urban) of the households reported having taken
            credit the year before the survey, again using family as the main source (68 percent) (see Table
            11). Shopkeepers and traders were the second most common source of financing with 24
            percent of respondents securing a loan from these sources (26 percent rural, 10 percent urban).
            Quantitative and qualitative studies suggest this small scale informal credit is widely a vailable,
            and widely used as a livelihood strategy (NRVA, AREU). However, most of this credit is used
            not for investment, but for consumption and to respond to household shocks such as illness or
            weddings (see Table 12). Only 3 percent (2 percent rural and 12 percent urban), of respondents
            used their informal loans for business investment.
                           Table 12. Main use given to the loan
                                                  Rural      Urban    Kuchi      Total
                                                  4%         1%       0%         4%
                                                  2%         12 %     5%         3%
                           Land purchase          1%         1%       0%         1%
                           House purchase
                                                  4%         9%       2%         5%
                           Or construction
                                                  4%         9%       2%         4%
                           Food purchases         63 %       45 %     68 %       61 %
                                                  8%         10 %     10 %       8%
                           Bride price
                                                  6%         6%       6%         6%
                           Funeral                3%         3%       2%         3%
                          Other                4%       5%    6%                 4%
                        Source: Authors calculations from NRVA 2005
            Using NRVA 2005, we analysed the urban households reporting ‗business investment‘ as the

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            main use given to the loan. Not surprisingly, most of them have ‗trade and services‘ as the main
            source of income, more importantly for those below the official poverty line. Twenty-two
            percent of the non-poor urban households investing in a business did so into agriculture;
            suggesting access to land and poverty are negative correlated (see Table 13).
                    Table 13. First source of income for urban households investing in
                    a business
                                              Trade and
                           Agriculture                      Manufacture      Other
                    Poor                       Services
                    Yes         2%              89 %            0%            9%
                    No         22 %            71 %           2%              4%
                         Source: Authors calculations from NRVA 2005
            The limited availability and use of formal finance inhibits the develo pment of licit
            opportunities. In particular, it is reported SMEs lacking working capital are forced to buy
            inputs in the local market with deferred payment conditions but highly onerous conditions. If
            the credit constraint were relaxed, they could import inputs and save up to 25 percent on
            costs.18 With the exception of a few deep-pocket investors, many existing firms are unable to
            further expand operations without bank credit. New start-ups are also limited by self-funding.
            This is one of many factors mitigating against an open playing field for entrepreneurs in
            Afghanistan, and supporting the further concentration of wealth and assets in the hands of a
            Both, financial intermediation and credit to the private sector in Afghanistan continue to be
            the lowest in the region. In the current conditions it is unlikely that firms can fully benefit
            from adequate overall liquidity given the underdevelopment of the financial system and the
            high degree of informality existing in the economy. As the country o utgrows the
            reconstruction phase and requires more private initiative to generate economic opportunities
            this might become a more relevant constraint, but we don‘t see the financial sector acting as
            the most binding constraint in Afghanistan these days. The lending interest rate is not out of
            range and in spite of the limited formal financial services, the World Bank survey, data from
            NRVA and anecdotal evidence suggests that there is no serious shortage of capital in the
            Afghan business community. This is especially valid amongst the Afghan Diaspora. In asking
            what is constraining private investment in Afghanistan the most, the GD analysis suggests it is
            not finance. We then discard this and move into the ―low returns to economic activity‖ section
            of the GD.

            5.2          Low Social Returns
            Complementary factors to production (among other factors), like human capital or
            infrastructure will determine the country‘s competitiveness and impact directly on expected
            private returns. Geography, as included in the framework, is also considered both as a source
            of advantages or disadvantages with impacts on both domestic and international trade. This
            section discusses each one of these components for the case of Afghanistan.

            5.2.1      Geography
            Afghanistan is located at the centre of Asia, landlocked and mountainous. It contains most of
            the Hindu Kush mountains, the westernmost extension of the Karakoram and the Himalayas,
            which further isolates regions within the country. Bounded by 5.529 km of borders,
            Afghanistan shares with Pakistan a border of some 2.640 km (the contested Durand line),
            almost 1.000 km with Iran, and the remainder with Turkmenistan, Uzbekistan, Tajikistan and

              Expanding the Outreach of Financial Services: The experience of Financial Institutions. Case
            study prepared for the Enabling Environment Conference, Aga Khan Development Network,

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            The country also has abundant natural resources including natural gas, oil, coal, copper,
            uranium, gold, silver, chromate, talc, barites, sulphur, lead, zinc, iron ore, salt, precious and
            semiprecious stones. 19 Some major mineral reserves are largely untapped, much existing
            mining is illicit, and the recorded activity in the sector contributed less t han one percent of
            GDP in 2006.20 However, there are predictions that the contribution of minerals to GDP and to
            government revenues could increase dramatically (World Bank, 2004b)(see Section 5.3.1).
            Afghanistan has large variety of climate and soils, from plains in the north and southwest to
            large desert areas near the southern border with Pakistan. Around 12 percent of the land is
            arable, but only one-half of this is used due to lack of irrigation (Guimbert, 2004). Still,
            agriculture is the second most important sector in recorded GDP and the main occupation for
            the vast majority of Afghans. Winters are harsh and earthquakes, flooding and droughts are
            frequently recorded. During the last 10 years, the country has registered recurring droughts
            (1998 to 2002, 2004, and 2006) aggravated by deforestation and soil degradation. Each year,
            400,000 people are seriously affected by natural disaster. The NRVA 2005 reports some 6.6
            million Afghans (one-quarter of total population) do not meet minimum food requirements .
            Location was Afghanistan‘s main asset for centuries, during which the country served as a land
            bridge in the region, connecting south and central Asia‘s trade routes. And since the fall of the
            Taliban, much has been made of the hope that the country will regain its position as a land
            bridge and cross-roads of trade by transporting energy and other goods in the region. Thus far,
            Afghanistan has not be able to exploit either its geographical location or legally develop its
            natural resources. Infrastructure capacity, security issues, neighbours relations (mostly with
            Pakistan) and inadequate regulatory frameworks and enforcement mechanisms are the main
            limitations for these to materialize. 21 Even more, large parts of the country are isolated limiting
            markets clearing within the domestic market. Land transportation costs across borders, the
            most expensive in the region, are often cheaper than domestic transport within the country .
            A study of trade in the wider region surrounding Afghanistan has flagged up signi ficant costs
            of being land-locked for this region (Byrd & Raiser et al., 2006). As well as additional costs,
            dependence on the transit infrastructure, cross border political relations, peace and stability
            and administrative practices of neighbouring states may constrain growth (Faye, Mcarthur,
            Sachs & Snow, 2004). Afghanistan has often been perceived as a threat by its neighbours as it
            is seen as an exporter of conflict and political instability as well as opium and other illegal
            goods. As a result, neighbouring countries have periodically closed their borders with
            Afghanistan, limiting commerce. Pakistan frequently closes the border to Afghan exports, for

               Afghanistan Geological Survey, 2007
               By November 2007, Afghanistan announced the State-owned China Metallurgical Group
            won the rights to exploit the Aynak copper field in Logar Province, south of Kabul. This
            investment is likely to generate significant revenues.
               ANDS foresees the promotion of Afghanistan as the centre of a regional transit network by virtue
            of its strategic geographic position to gain access to international markets and to link major trading
            powers by affording them the shortest access to the sea. For this materialize, the country would
            build (1) the North-South Transport Corridor (to connect Tajikistan, Uzbekistan and
            Turkmenistan with the Pakistani ports of Karachi, Qasim and Gawada and further (through
            Wagah) with India and South Asia) and (2) the East-West Transport Corridor (to connect
            Tajikistan, Uzbekistan and Turkmenistan with the Iranian ports of Cha Bahar and Bandar Abbas.)
            The main transit routes and products going through Afghanistan‘s various border points are: from
            Turkmenistan to Pakistan (scrap metal); from Iran to Pakistan (vehicle spare parts); from
            Uzbekistan to Pakistan (cotton, scrap metal and iron rods); from Tajikistan to Pakistan (scrap
            metal); from Pakistan to Iran (wheat); and from Pakistan to Uzbekistan (cement). (‗Transit
            Transport in Afghanistan‘ presented by Mr. Hedayatullah Watanyar, Head of the Directorate of
            Transit and Trade Facilitation, Ministry of Commerce and Industry to the UNCTAD Expert
            Meeting on Regional Cooperation, September 2007).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            various reasons such as trucking strikes by Pakistani truckers, on whom licit Afghan exports
            are reliant. Afghan trucks are not allowed to use Pakistani roads and also access to Iran is
            limited. This forces exporters to transship cargo from Afghan trucks into Pakistani or Iranian
            trucks for transit into those countries, with large cost in money and time, an d huge losses on
            cargo (e.g., up to 60% on fresh fruits, EPAA). As reported by the World Bank Doing Business
            report (2008), to export a container in Afghanistan costs US$2.500 (compared to regional
            average of US$1.180) and it takes 67 days (compared to regional average of 32.5). Based on this
            report, Afghanistan ranks 174 (out of 178) in terms of trading across border. (See also next
            section on ‗infrastructure‘). Yet, we see how official exports are more dynamic than GDP (see
            Table 1) and expected to keep growing.
            The implications for long-term growth, poverty and development, not to mention security,
            deriving from Afghanistan‘s geography and location are massive. The combination of poverty,
            lack of infrastructure and environmental degradation is a serious c hallenge for the –mostly
            poor– 80 percent of Afghans living in rural areas (CSO &UNFPA, 2006) 22 and for the economy
            as a whole. Huge investments in infrastructure, including roads, energy and irrigation are
            required to connect the country within and across the region, and to reduce vulnerability and
            volatility in the agriculture sector. Regional transit agreements need to be enforced and
            enhanced, and international support is to be provided to Afghanistan to solve border issues
            with Pakistan. 23 The Central Asian Republics have also tight border and visa restrictions hence
            increasing exports to Russia may need to include improved relations with the CARs, as much
            as the transit through Pakistan is key for exporting to India. This is one of the reasons why
            Afghanistan joined CAREC (Central Asian Regional Economic Cooperation), regional
            organization geared towards improving cooperation in certain key areas (trade facilitation,
            customs, energy and transport), but again more cooperation and international support woul d
            benefit trade in the area.
            Afghanistan‘s geography, location and topography impose serious short and mid -term
            challenges on the country. But geography is also a potential source for long term benefits, in
            the spheres of trade, agriculture and minerals development. Geography is revelled as a critical
            constraint to long-term growth in Afghanistan, one that requires massive investments in
            infrastructure to be overcome, and for trade, agriculture and mining to actually benefit the

            5.2.2      Infrastructure
            Poor infrastructure figures highly in most lists and ratings of the largest barriers to private
            sector growth in Afghanistan. Lack of access to infrastructure is exacerbated by geography and
            climatic issues. In rural areas and areas rich in natural resources, damaged irrigation systems
            and poor access to roads limit the potential for profitable investments. Access to infrastructure
            such as roads and electricity varies considerably across Afghanistan, with isolated areas
            particularly affected by poor infrastructure. 24 But infrastructure is also one of the most
            benefited sectors from foreign aid, with important investments being done and steadily
            improvement since 2001.

               At the time of writing, it was not clear whether all the population statistics in the cited
            report had been officially approved.
               For example, the Pakistani National Trade Corridor Improvement Program (NTCIP),
            supported by the World Bank, includes key policy reforms along with a comprehensive
            investment program to be implemented in the period 2005-2010. Coordination and
            collaboration between the World Bank projects in Afghanistan and Pakistan could be a good
            starting point for technical considerations and cross-border logistics improvement.
               Uneven economic development in Afghanistan has been identified as a key challenge to
            state-building. Key informant interviews in Kabul, April 2008.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            In spite of large investments in road reconstruction, the country is still ranked at the very
            bottom (150 out of 150) of the World Bank Logistic Performance Index (2007) and is very
            distant from its immediate neighbours and similar countries (see Table 14). Afghanistan has
            one of the worst road systems in the world, and no rail system. Among landlocked developing
            countries, Afghanistan has one of the world longest distances to a seaport, more than 2.000
            km, over harsh terrain (World Bank). It therefore does not come at a surprise that according to
            the World Bank Doing Business report (2008) exports from Afghanistan cost twice the
            regional average and take more than twice the number of days.

                 Table 14. Afghanistan and selected countries, Logistic Performance Index (2007)

                 LPI                                               Internatio   Logistics   Tracki    Domestic
                                           Custo    Infrastructu
                 Ran    Country     LPI                                nal      competen     ng &     logistics
                                            ms      re
                  k                                                shipments       ce       tracing     costs
                 39    India        3.07   2.69         2.9           3.08        3.27        30        3.08
                 68     Pakistan    2.62    2.41        2.37          2.72        2.71       2.57       2.86
                  78    Islamic     2.51    2.5        2.44          2.59         2.69        2         2.93
                 Low income
                                    2.29    2.12        2.06          2.32        2.29       2.25       2.99
                 129                2.16    1.94         2            2.07        2.15       2.08       2.91
                 146                1.93    1.91         2             2           1.9       1.67       2.33
                 149                1.71    1.63        1.67          1.5          1.6       1.67       3.33
                 150                1.21    1.3         1.1           1.22        1.25         1        3.13
            Source: World Bank, Logistic Performance Index 2007
            The scores are from one to five, one being the worst performance for the given dimension.

            Lack of access to electricity affects both rural and urban businesspeople, and figured as the
            single most important obstacle in the 2005 World Bank Investment Climate survey. In 2005,
            seventy-four percent of the sample surveyed was connected to the grid. But public supply is of
            poor quality (low voltage, intermittent dispatch and blackouts) and lasted on average 6.5
            hours a day. In addition, to be connected to the grid, bribes must be paid. Seventy-six percent
            of respondents reported owning at least one generator compared to Syria (73 percent), India (64
            percent), Pakistan (42 percent), China (18 percent), Uzbekistan (2 percent) and Tajikistan (0
            percent). The survey (World Bank, 2005d) reports firms lose on average about 18 percent of their
            merchandise value due to power disruptions, and this number can go as high as 30 percent in
            provinces more dependent on the public system such as Kandahar.25 Only 23 percent of all
            households have access to electricity. From that group, 14.5 percent has access to public
            supply grid, 3.8 percent to a community generator and 2.6 percent to personal generators
            (diesel) (MRRD and CSO, 2007). Still, almost 77 percent of the households have no access to
            any source of electricity, and considering installed capacity and population, Afghanistan is the
            country with the lowest per capita consumption of energy (without accounting for on-site
            The existing power supply infrastructure is fragmented with four isolated grids clustered

              AREU research on barriers to rural SME development found that in Bamyan province in the
            Central Highlands, lack of electricity coupled with mountainous terrain and delayed road
            reconstruction stymied significant potential for processing of agricultural products that could
            otherwise access markets such as Kabul that were not so remote in terms of distance (AREU,

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            around Kabul, Mazar-e-Sharif, Herat and Kandahar. It is estimated that only about 70% of
            transmission and distribution systems survived the decades of war, poor maintenance, and
            total disregard.26 There have been some improvements since 2001. Capacity almost doubled
            between 2002 and 2007 (see Table 15), and several important projects are expected to add
            more power to the main cities during the next years. USAID, India, KfW and ADB are
            supporting a multi-donor effort to import over 100 Mega Watts (MWs) of power to Kabul
            from Uzbekistan by end-2008. In addition, the US-financed Infrastructure Rehabilitation
            Program is currently building a 100 MW diesel power plant in Kabul, expected to b e
            operational by end-2008. The Ministry of Energy and Water expect that by October 2008
            150MW will be imported from Uzbekistan to Kabul (part of the IRP) with another 150MW
            expected by March 2009. But this is still uncertain. Even if most of the infrastruc ture is in
            place, the power purchasing agreements (PPA) with Uzbekistan are still being negotiated.
            Imports from Tajikistan (up to 300MW) are expected by spring 2010 for six summer months
            until 2013 for year round 300MW supply thereafter. Sheberghan Gas Po wer station is
            currently under development and expected to be operational within three years adding
            100MW to the grid. The Aynak Copper project is expected to generate for its consumption and
            to provide the grid. Another line connecting the country with Turkmenistan for imports of up
            to 300MW is being considered. By 2012, the country is expected to add 1.019 MW to the
            existing installed capacity of 769 MW in 2007, but all this remain contingent on issues like
            security, bilateral agreements and the availability of funds for large-scale investments.
                   Table 15. Sources of Electricity Supply (Operating
                   Capacity in MW)
                   Year    Hydro Thermal Imported Hydro Total
                   1978      259       137          0        0      396
                   2002     141       16         87         0       243
                   2007      184      87           96       15      464
                   Source: Power Sector Strategy for the ANDS, Ministry
                   of Energy and Water, DRAFT (April 2007)
            The unreliability and expense of electricity provision is a source of great disadvantage and a
            huge toll for Afghanistan‘s competitiveness, and generators represent an extra overhead that
            may be prohibitive for potential smaller businesses. Power generation is heavily dependent on
            imported fuel and one of the biggest cost items in power production. On top of the internationally
            high prices for fuel, the diesel market in Afghanistan is characterized by abuse of market power and
            barriers to new entrants, corrupt allocation of import licenses, and unregulated imports of poor-
            quality fuel (Paterson 2005). However, with the exception of energy-intensive processes (e.g.,
            marble cutting), generators are a viable option. 27 As AREU research has indicated, ‗generators
            can be factored into a business plan with some degree of predictability‘ (Paterson, 2006). In
            spite of uneven results thus far, several large scale programmes for energy provision and use of
            industrial parks28, should improve the situation for the energy-intense sectors. 29

               Power Sector Strategy for the ANDS, Ministry of Energy and Water, DRAFT (April 2007)
               A marble businessman (EPAA) suggested that it is not only more reliable, but also cheaper to
            use generators (even diesel) compared to grid power once you incorporate the systematic
            bribes that have to be paid to be connected to the grid. Also, they can absorb a predictable
            cost, but not the uncertain level of bribes.
               Industrial parks have had mixed-evaluations in Afghanistan so far. Although several are still
            under construction and more critical mass and time is needed for a full assessment, location
            and accessibility, as well the costs are issues reported against the success for parks thus far.
            Given the limited amount of available land, and the severe limitations electricity imposes on
            energy-intensive sectors, in theory Industrial Parks are a good short-run solution.
                 A representative from the carpet exporters, reported energy as a non-issue for its members

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Value added chain
            For a rural economy, and a country where agribusiness is the most important and dynamic
            export sector, the abysmal absence of value-chain related infrastructure, like cool-storage
            facilities, warehouses, export terminals, etc., in addition to the energy and transportation
            issues, is a constraining issue for exporters. We have no statistics, but informers report Afghan
            exporters using Pakistan infrastructure for storage or the final stages in the value chain. Cool-
            storage facilities across the border are used to store Afghan perishable exports (fruits and
            vegetables) to then re-import again for domestic offseason consumption. Marble is exported as
            rough hewn blocks to Pakistan where it is processed (cutting and polishing) and then re -
            imported back to Afghanistan (or exported from Pakistan) to supply the growing market for
            construction materials. A very large number (above 90 percent, EPAA) of Afghan carpets are
            transferred unfinished to Pakistan where cutting and washing facilities exist in Peshawar.
            Pakistan traders then export them worldwide as a ―made in Pakistan‖ product. 30 Based on export
            rate growth and the evolution of specific sectors (see Section 5.3.1) this seems to be limiting
            country‘s capacity to add value, but not its capacity to increase exports.
            Irrigation infrastructure appears to be a major obstacle for sustainable investment and growth
            in Afghanistan‘s agriculture sector (including formal and informal this is roughly 40 percent of
            countries economy). The sector is key for the economy as a whole, not only because it provides
            for the large rural population (80 percent) and represents the largest job creating sector, it is
            also the most promising one in terms of international competition for exports. Historically low
            yields and low productivity are evidence of low technology and inputs, including irrigation,
            although recent reconstruction efforts are having a positive impact (see Graph 3).

                Graph 3. Af ghanistan, Iran and Pakistan, Yield (wheat, hg/HA)











                                      Afghanistan         Iran      Pakistan

                Source: WDI (2008)

            Afghanistan is a drought-prone country, recording droughts from 1998-2002, in 2004 and
            2006. Crop failure, pests and death of animals due to diseases is one result, but there are other
            negative outcomes out of the lack of irrigation infrastructure. Seasonal flash floods are being
            reported in the same areas affected by drought, indicative of absence in storage and
            distribution capacity (Shobair and Alim). By early 2001, drought placed large parts of the
            country near famine conditions as the traditional sharing and coping mechanisms are

            operating from industrial parks (EPAA).
               Cutting and washing facilities require several acres of land that has water, drainage and
            transportation infrastructure on site. Industrial parks in Afghanistan are more and more being used
            for Afghan producers, but it is still a small fraction.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            insufficient to cover long periods of drought and large parts of the country.
            In 2003, FAO estimated an acute decline in water flow (66 percent), which caused a 58 percent
            reduction of the irrigated land (FAO, 2003). The study considered the severity of the droughts
            and the reduction in main water sources, improper operation and maintenance of canals (siltation
            in the canals, damaged regulating systems, etc.), mismanagement in irrigation water causes low use
            efficiency, lack of sound irrigation structures (90 percent of the systems and water conservation
            technologies are traditional). To show how constraining droughts are in terms of agricultural
            production and export, graph 4 presents an index with aggregate exports in agriculture and
            livestock from Afghanistan since 1998. The negative impact from drought is evident from 1998 -
            2002 and then again in 2004.

            Source: COMTRADE, 2008
            Infrastructure is the most important issue in long term perspective for Afghanistan‗s
            competitiveness and development. As could be expected from a war-torn country,
            Afghanistan‘s infrastructure was much damaged. For a landlocked rural country and economy,
            the implications from this are significant, in particular when interacted with geography,
            including the geo-politics of the region, and considering the urgent necessity for country‘s
            reinsertion into the world economy. But at the same time, infrastructure is one of the sectors
            most benefited by aid and donor-financed reconstruction efforts, with steady improvements
            since 2001. We see how exports and in particular agribusiness-related exports are extremely
            vulnerable to climatic events, suggesting irrigation is the most constraining issue for the
            sector31 (see Graph 4), but at the same time they are the most important exports in terms of
            level and dynamism (see Section 5.3.1 and IMF (2008c) for export performance analysis).
            We see currently exporters and traders are forced to use neighbouring countries‘
            infrastructure preventing them from locally adding value and from increasing their margins,
            but at the same time exports (official) have grown faster than GDP since 2001 (see Table 1).
            This suggests infrastructure is preventing the country from adding value , but not from
            increasing exports. There is no data to confidently conclude, but the differential growth rates
            in industry and in exports (over for example, services, a less infrastructure intensive sector, see
            Table 6) suggest these sectors are not constrained right now. Nevertheless, if governance
            issues are improved, and better efforts to connect the country to world markets are secured,
            investment and growth in the export sector will face in infrastructure the most constraining
            barrier in the medium and longer term.

                 Including also the informal agriculture.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            5.2.3       Human Capital
            By 1960, Afghanistan was ranked 21st amongst the world‘s worst countries in terms of total
            average years of education in the population over 25 (Barro and Lee, 1999) (see Table 16). By
            1990, the last date when data was available, it had become the fifth worst country. 32 By 1990
            the population had fewer years of average years of education than in 1960, and Afghanistan is
            in fact the only country--at such low levels--that reduced its total years of education from 1960
            to 1990. By 2006, adult illiteracy amounted to 72 percent (the fourth worst in the world after
            Mali, Burkina Faso and Chad, WDI 2008), while youth illiteracy stood at 65.7 percent (the
            third worst in the world after Mali and Burkina Faso, WDI 2008). 33 Female literacy stands at
            only 19 percent34, and rural illiteracy is three times higher than amongst the urban
                 Table 16. Average years of education in population over 25, selected
                                  1960   1965       1970        1975      1980     1985     1990
             Niger             0.2    0.17          0.16        0.39      0.39     0.44     0.55     2.75
             Mozambique       0.22    0.26          0.35        0.47      0.53     0.59     0.74     3.36
             Mali             0.25    0.24          0.26         0.3      0.46      0.6     0.76    3.04
             Gambia             …       …             …         0.5       0.64     0.75     0.89    1.78*
             India            1.45    1.49           1.9        2.35      2.72     3.12     3.55    2.44
             Iran             0.61    0.84          0.98        1.43      1.85     2.58     3.29    5.39
             Pakistan         0.63    0.93          1.68        1.62      1.74     1.82     2.29     3.63
             Afghanistan      1.06    0.89          0.76        0.73      0.79     0.88     1.01    0.95
            Source: Barro and Lee (1999)
            *Considering 1990/1975
            Enrolment data presents a similarly depressing story. By 1991, Afghanistan was the second
            worst country in the world in terms of primary education enrolment, and this record had
            further deteriorated by 1998 when data became available again and continued at very low
            levels until 2001. In the post-Taliban period (see Table 16), enrolment rates have grown at
            impressive rates and Afghanistan is now even outperforming countries with similar GDP per -
            capita and performing as well as others with higher income in terms of primary enrolment (i.e.,
            Pakistan with 87.3 percent primary enrolment in 2005)(see Table 17).

                          Table 17. Primary enrolment, selected countries (%)
                                                1991       2001        2003      2004     2005
                          Afghanistan           25.3       19.0        83.4      92.9     86.5
                          Eritrea               …          60.3        66.6      66.5     64.1
                          Niger                 25.8       33.4        40.6      44.7     46.7
                          Mali                  25.9       52.8        61.2      63.8     66.4
                          Ethiopia              30.0       69.1        72.8      77.0     93.4

               After Niger, Mozambique, Mali and Gambia.
               Of note, among adults returnees (and even though based on a small sample) only 47 percent
            are illiterate.
               Seventy-eight percent of women reported never attended schooling. IRC Afghanistan Labor
            Market Survey, 2003.
               ILO, Afghanistan Country report

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                           Uzbekistan              81.4        99.4       99.8          99.8          …
                           Tajikistan              91.0        96.6       99.7          99.9          101.2
                           India                   98.1        98.3       107.4         116.2         …
                           Iran,     Islamic
                                                   109.4       93.2       92.1          103.0         110.7
                           Pakistan                …           72.1       76.4          82.1          87.3
                          Source: WDI, 2008

            With almost one-quarter population in the 6-13 year cohort, Afghanistan has the highest
            proportion of school age population in the world. 36 The combination of one of the world's
            lowest participation rates (especially for girls) and the highest proportion of school age
            population implies that the tasks after 2001 were - and remain - daunting. Tertiary enrolment
            has not shown such improvement as primary and secondary and is still the fourth worst in the
            world (see Table 18). 37
                          Table 18. Secondary and tertiary enrolment, selected
                          countries (%)
                                               1991        2001       2003        2004          2005
                          Afghanistan          14.1        11.1       11.2        15.6          16.2
                          Eritrea              …           27.8       29.3        29.3          31.4
                          Niger                5.8         5.9        6.5         7.9           8.7
                          Mali                 6.7         15.0       20.2        22.3          23.9
                          Ethiopia             12.9        21.0       25.4        27.8          31.3
                          Uzbekistan           99.4        95.5       94.7        94.6
                          Tajikistan           102.1       75.8       80.9        81.8          81.8
                          India                44.2        48.0       52.3        53.5          …
                          Iran, Islamic
                                               57.5        77.4       78.5        81.9          80.6
                          Pakistan             24.7        25.7       24.6        27.2          26.9
                          Afghanistan          …           …          …           1.1           1.1
                          Eritrea              …           1.5        1.4         1.1
                          Niger                0.6                    0.8         0.8           0.9
                          Mali                 0.6         2.4        2.2         2.1           2.6
                          Ethiopia             0.7         1.4        2.2         2.5           2.7
                          Uzbekistan           30.4        15.2       15.3        15.3
                          Tajikistan           22.1        13.3       15.3        16.4          17.3
                          India                6.0         10.5       11.5        11.8          …
                          Iran,     Islamic
                                               10.2        20.2       20.3        22.5          23.9

                 MRRD and CSO (2007)
                 After Maldives, Malawi and Niger. Data for 2004.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                       Pakistan          3.4             2.6     3.2     4.6
                        Source: WDI (2008)
            Skilled labour force
            The Investment Climate Assessment (World Bank, 2005d) reports human capital at the
            bottom of its ranking of barriers to business. This suggests that the output composition and
            technology used by the sectors absorbing the bulk of the labour force (like agriculture,
            construction, and manufacturing) demands mostly low-skilled and unskilled workers. This is
            not only the case in low-skill sectors like construction, even services rank education in the
            labour force 15th out of 17th in the list of barriers to business. This is again clear as only 5
            percent of Afghan firms offer on the job training (ICA, WB 2005). Whilst 98 percent of
            managers in India and 96 percent in Pakistan have secondary or tertiary education, that
            number is 62 percent in Afghanistan (ICA, WB 2005).
            Migration is a way to alleviate labour market shortages, and one heavily used by Afghan
            households to diversify risks and provide incomes 38, and by Afghan companies to import
            specific expertise not found in the country. While Afghans go overseas (e.g., to Dubai, Iran or
            Pakistan), foreigners come to the country, providing a mechanism by which different skills are
            exchanged. Such mechanisms, even if unpopular among the unemployed, should be
            maintained. In that sense, temporary migrant agreements and treaties in ―movement of
            persons‖ in general should also be pursued by the Afghan government with neighbouring
            countries and more broadly in the region, including the Gulf and other low-skill-labour
            absorbing countries.
            Returns to education
            The ICA assessment notes that poor investment in education is a v icious circle, arguing that
            ―Afghanistan is caught in a trap‖, as lack of opportunities discourages workers from investing
            in education, but ―opportunities are lacking‖ in part due to the fact that investors are
            constrained by the lack of skilled labour (WB ICA, 2005). We do not see this cycle operating
            except for the impact it has on the public sector as mentioned below. Firms do import skill
            capacities not provided in the country, thus investors are not constrained by local labour force
            education39. In addition, we do see a large increase in enrolment, suggesting education is
            valued as a good investment. To assess this perception, we estimated a Mincer equation for
            returns to education in Afghanistan. 40
            The results suggest there is a high premium on education. Using NRVA 2005, from a sample of
            urban men, we estimated a premium 17 percent for secondary, 22 percent for high school and
            20 percent for tertiary (see Annex 3). These results are very high, proving a tight market for a
            scarce resource, and a possible binding constraint. But this contradicts the business people‘s
            perception on the scarcity of educated labour. More importantly, it contradicts the World
            Bank ―trap‖ as they prove the dynamic incentives are in place for people to invest in education
            as a profitable decision. 41 In effect, if students perceive further education as a profitable

               As shown before (see Table XX), 9 percent of the population responded having remittances
            as the first (5 percent) or second (4 percent) most important source of income.
               With 40 percent estimated unemployment (ANDS), importing workers (both skilled and
            not) becomes a sensitive political issue. At the same time, Afghan workers migrate to Iran,
            Pakistan, and the Gulf.
               A standard Mincer earnings equation estimates returns to education (in our case completed
            levels) by estimating earnings as a function of years of education and years of potential labor
            market experience (age minus year of schooling minus six).
               Based on our qualitative assessment and interviews in Kabul during April 2008, this
            impression was confirmed. There is still a perception that senior managerial positions are
            occupied by returnees or foreigners, but there is also an increasing supply of skilled local
            human capital. A large mobile phone company in the country recently reported a large

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            decision (this of course conditional on they finding a job once graduated) , they will continue
            to consecutive levels and will postpone the decision to enter the labour force, in expectation of
            a higher wage. The results can also be rationalized within the context of increased demand for
            educated workers by donor agencies and NGOs, distorting the labour market. Bearing in mind
            2004/2005 was a peak period for aid-money entering Afghanistan, and a period of sharp real
            exchange rate appreciation, such explanation can be considered (this will be captured by
            NRVA 2005, especially in urban areas). We also collected via informers‘ anecdotal evidence of
            such distortion in cities where NGOs and donor agencies operate.
            By any measure, human capital in Afghanistan ranks amongst the lowest levels in the world. In
            spite of this, evidence suggest lack of human capital is not the most binding constraint for
            private investment at present time, given the low levels of skills required by the current
            technological structure and the mobility of workers in the area. Yet, it might limit medium and
            long term performance in specific sectors by impeding product and exports diversification an d
            sophistication or further expansion in the services industry, but only if migration and labour
            mobility is limited. In that sense, the improvements in school enrolment after 2001 should be
            maintained and quality should be targeted as the next goal, even if it will take a generation for
            that effort to yield benefits in terms of changing the skill composition in the labour force and a
            longer period to catch up with its neighbours or other societies of similar income. Vocational
            training is a pressing need, as a way to insert the growing labour force into the market.
            More importantly, the lack of human capital is a severe limitation for the government and its
            agencies, and it is through this channel how it affects private sector more brutally. A more
            educated civil service is a precondition to almost any further government reform , but this, as
            framed by the GD does not impact low returns to economic activity, but its appropriability.

            5.3        Low Appropriability of Returns
            5.3.1      Market Failures: Self-Discovery
            As noticed by Hausmann, Rodrik and Velasco (2005), ―the development process in less
            advanced countries is largely about structural change: it can be characterized as one in which
            an economy finds out -self-discovers- what it can be good at, out of the many products and
            processes that already exist.‖ As a small and rural economy with one of the lowest incomes in
            the world, the Afghan domestic market (aid-driven demand included) is unlikely to be
            sufficient to pull the country out of its low productivity and low investment equilibrium. 42
            Therefore, country‘s integration into the world economy will be a key component of its long -
            term growth strategy, most likely the only one. 43
            The ‗self-discovery‘ issues are ―potentially more important and the payoffs of addressing them
            are much larger‖ (HRV 2005) in the tradable sector, as the tradable sector‘s productivity can

            expansion of its legal department from two to a dozen college graduates, including several
            graduates from Kabul University.
               For instance, two-thirds Afghan dried fruit and nuts production is consumed within the
            country, and only one-third is exported. It is clear, if the sector is to grow, it has to be done
            supplying the international market. Accessing the international market will not only allow for
            larger volumes, but also better prices. (The Rapid Expansion of Dried Fruit and Nuts Export,
            Case study prepared for the Enabling Environment Conference, 2007).
               We analyzed together ―self-discovery‖ and ―coordination failures‖. In a country this the level
            of development and after so many years of war and devastation, it is clear the market is not
            well functioning and therefore ill coordinated. Having that in mind, we consider most of the
            coordination failures, for example a private sector joint effort to build storage or other sector
            specific infrastructure is more related to failures in the contract enforcement environment and
            therefore more related to government failures than to market failures. Most of the section is
            devoted to ―self-discovery‖.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            be scaled up to supply world demand. In order to develop new ideas that lead to new sectors,
            countries require specific public goods and private capacities, like infrastructure, changes in
            regulations or financing schemes as well as human capital, which are all underdeveloped or
            lacking in Afghanistan today. Cooperation among countries is required as well, in particular
            for a landlocked country like Afghanistan. For a landlocked country to be competitive,
            infrastructure developments in general (roads, electricity) and export-oriented in particular
            (transit infrastructure, sea ports access, greenhouses and cool-chain infrastructure 44), cross-
            border political relations (e.g., Pakistan malpractices regarding Afghan transit goods exported
            to India), peace and stability and administrative practices of neighbouring states are also of
            utmost importance. Both domestic and international issues are constraining export g rowth,
            transit (using Afghanistan as a land bridge) and diversification in Afghanistan today.
            Afghan exports, composition and destinations
            We used imports from Afghanistan reported by its partners to analyze export composition and
            characteristics from 1962-2006. There are several limitations with the data being used in this
            section. Official statistics do not give the full picture, as misreporting is rampant and
            smuggling very large. 45 Opium exports are not included, re-exports are also ignored. Afghan
            exports to Iran (less than 4 percent of total exports in 2007, Afghan Customs Agency) are not
            reported in COMTRADE, and therefore ignored in this analysis. But e ven if Afghanistan‘s
            external sector suffers from misreporting and a large volume of exports and imp orts are not
            captured by the statistics, the data at least gives a fair representation of country‘s economic
            capacity and ‗revealed comparative advantages‘ as well as its trade partners. Assuming the
            illegal exports represent a fraction on each sector, we have a good assessment of country‘s
            comparative advantages, albeit the share and ranking of these products is not informative at
            high confidence levels. 46 Of note is the fact that smuggling has been decreasing steadily if
            compared to official exports. If official exports represented in 2002/03 57 percent of the
            amount smuggled, that ration reverted in 2004/05 and since then official exports are more than
            twice smugglings.
            As noted, illegal trade and smuggling is not included in the data. Electronics and other re-
            exports, logging, gemstones and opium account for the bulk of illicit trade. 47 There is no way
            to account for local products being exported but not reported, with a high risk of biasing our
            conclusions. Still, we consider the information from the recorded trade to be valuable as it
            shows products and destinations for Afghan exports. For the purpose of this section re -
            exports add no valuable information as they are not produced in Afghanistan. On the other
            hand, re-exports have no impact on local GDP other than the fraction contributed by the
            transport services and the arbitrage premium, assuming these are provided by local
            businesspeople. Some have argued that enforcement actions against re -exports may have a
            rather negative effect on customs revenues. 48 Opium, of course is another story, but it is

               For instance, these two (among others) are subsidized by the Pakistan Horticulture
            Development and Export Board (WTO, 2008).
               In 2006/07, smuggling was estimated at US$150 millions, while official expor ts were
            reported to be US$416 millions. In 2002/03, smuggling (US$177m) was greater than the official
            exports (US$100m). Source, IMF 2008a.
               There are records small quantities of gemstones and semi-precious stones, as well as logs
            officially exported every year. But as these and other items are also smuggled we just don‘t
            know the real volume and the ranking on each item. The main point here is to acknowledge the
            country is producing and exporting a certain basket, but that is done in part legally (offici al
            exports) and illegally (smuggling).
                 A small fraction of logging and gemstones is reporter as official trade also.
               Indeed, re-exports differs from transit trade. As per agreements between Pakistan and
            Afghanistan, trade transit imported through Pakistan is not taxed by that country (with
            higher tariffs than Afghanistan), but they are taxed at the Afghan border. Then they are re-

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            already accepted that the country has comparative advantages on that.
            Using this partial picture of Afghan exports, it becomes clear that the profile of Afghan
            exports has not changed dramatically since data was first available (1962). Just by looking at
            Table 19, we can see exports are more or less the same at disaggregated levels, and always
            within agriculture and agro-based industry sectors. Although there has been some change, for
            example gas exports to the former Soviet Union were significant from the 1960s and garments
            in the 80s, those sectors did not survive in relevant quantities. In real 2000 US dollars, the top
            10 exports in 1962 amounted to over US$300m, compared to total 2006 exports worth
            US$160m. In fact, the most important export item in 1962, fur skins, accounted for more than
            the total (recorded) exports in 2006, whilst today volumes of fur exports are 10 percent of
            their 1962 levels. Similarly, carpet exports in 2006 are less than 10 percent of t heir traded
            volumes in 1962. With the exception of ―natural gums‖, ―spices‖ and ―oil seeds‖, country‘s
            exports are today as they were 45 years ago. And even those new exports are with the
            agriculture and agro-based industry.

                 Table 19. Afghanistan, top ten exports in 1962 and 2006 (2000 US dollars, millions)
                                             1962                                     2006
                                                                      Dried fruit, dehydrated
                 1    Fur skins, undressed                    173     artificially                    24.0
                 2    Carpets, carpeting and rugs, knotted    55.2    Fur skins, undressed            18.3
                                                                      Natural gums, resins,
                 3    Edible nuts, fresh or dried             23.8    balsams                         12.8
                 4    Dried fruit, dehydrated artificially    21.0    Edible nuts, fresh or dried     8.5
                                                                      Raw cotton, other than
                 5    Sawlogs and veneer logs (non conifer)   17.8    linters                         8.0
                 6    Sheep‘s and lamb‘s wool, degreased      4.8     Grapes, fresh                   7.1
                                                                      Spices (ex. pepper &
                 7    Raw cotton, other than linters          3.7     pimento)                        7.1
                 8    Materials of animal origin              3.0     Iron & steel scrap              7.0
                 9    Fresh fruit                             2.3     Other fresh vegetables          6.6
                      Fine animal hair (ex.wool) not carded           Oil seeds, oil nuts & oil
                 10   or combed                               1.3     kernels                         6.2
                      Total                                   314     Total                           161
            Source: COMTRADE, UN, 2008
            One key feature of current external performance is the dependence on neighbouring countries
            as destinations for Afghan exporters. From top market destination in the past, developed
            countries are now the least important, suggesting damage to capability during war and lower
            capacity in the export sector to reach more distant and higher standard markets. In 1962,
            United States, Germany and United Kingdom accounted for 73 percent of exports, and India
            and Pakistan for only 20 percent, by 2006 Pakistan and India where the two main commercial
            partners receiving over one-half of recorded exports (much more if illegal trade is considered).
            In 2006, India was the destination for almost 20 percent of Afghan exports, sec ond to Pakistan
            with 34 percent (COMTRADE, 2008). 49 With a combined population of 1.3 billion, and

            exported back to Pakistan.
              Using Afghan customs data, in 1385 (2006/2007), Afghanistan main partners for exports
            where: Pakistan (44.8), India (27.6), Russia (3.8), Iran (3.1), United Arab Emirates (2.1),
            Uzbekistan (1.8), Kazakhstan (0.4), Turkmenistan (0.3) and others (16).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            income per capita of almost $600 compared to about less than 25 million and $300 in
            Afghanistan, to supply these markets seems to be a reasonable strategy. In particular, India is a
            key market to be exploited. The fourth largest world economy and a star performer in terms of
            growth during the last 25 years is an active trading partner beyond its size and income. Afghan
            agribusiness exports are known and appreciated there (quality issues are less of a concern) in
            addition, the two countries signed a preferential trade agreement that should further be
            exploited by Afghanistan. 50

            Russia, Iran, the United Arab Emirates, Arab as well as Central Asian Republics and other Gulf
            countries should also be considered as promising markets, but again several issues are
            interplaying here. Russia for example, a growing destination for Afghan exports (e.g., the main
            market for Afghan raisins in 2006) can be accessed via Uzbekistan (Hairatan Border). But even
            after the two countries signed a Trade, Transit, and Railroad Memorandum, progress and trade
            flows are reportedly low.
            At current income and competitiveness levels, the best opportunity for the country is to
            capitalize its comparative advantages by increasing its traditional exporting sectors.
            Comparing the exported volume in 1962 and 2006, it‘s easy to see the country has market
            participation to regain and space to do it. It will not be easy or quick, as it demand s better
            quality and certification in order to compete in high end final markets, but it is more
            promising in the short-term than sector diversification. Sector diversification should be
            maintained as a long-term goal, but it will be costly and slow, and conditional to
            improvements in governance and infrastructure. Improvement in infrastructure, in particular
            energy and road and air connections 51 is a necessary condition to increase country‘s
            competitiveness by upgrading the value chain and sophisticating export capac ity (e.g., by
            adding warehouses, cool-chain infrastructure, packing facilities, certificates, export terminals,
            etc.). Pakistan's cooperation in order to allow Afghan products to reach India or its closest sea
            port in Karachi is a precondition for destination and product diversification, but this cannot
            be sorted at the Afghan level alone. The ―low hanging‖ fruit in terms of export strategy in the

                Actually Indian consumers pay a premium for Afghan dried fruits and nuts over local
            produces (The Rapid Expansion of Dried Fruit and Nuts Export, Case study prepared for the
            Enabling Environment Conference, 2007).
               In the absence of infrastructure or good relations with neighbours, and given the level of the
            insecurity in large parts of the country, air transport can provide access to at least the Gulf and
            India for higher value goods.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            short and mid-term would be to increase volume in the current markets (especially in the
            region) and diversify trading partners by exploiting country‘s comparative advantages. Once
            the export sector becomes solid and based on its most competitive products, then product and
            sector diversification can be pursued.
            Markets and competition
            It goes beyond the scope of this study to assess the full competitiveness capacity for
            Afghanistan. We judge an export-led growth strategy initially based on rural and agribusiness
            traditional sectors, and focused on regional and traditional partners to be the right one in
            inserting Afghanistan to the world economy. Currently, the country does not have the
            necessary conditions and factors to diversify into manufacturing, and mining even if promising
            is still conditional to attracting FDI and improving infrastructure.
            Product-destination decomposition suggests that one-half of total value exports in 1977 was
            not being exported in 2006. As can be seen from Graph 6 this is mainly due to the loss of
            market destinations not due to a change in comparative advantage on the specific products.
            This collapse in Afghanistan‘s exports is mostly a collapse in its capacity to export, not in its
            capacity to produce the commodities or a collapse in the international market in that product.
            In addition, the surviving trade collapsed, in some part replaced by other commodities more
            suited to money making in war but also due to destruction of agriculture and livestock
            through war and drought. In 30 years, one-half of export value was lost due to the incapacity
            to reach developed markets. There may be several factors behind this, but the analysis suggests
            that it is not due to a structural transformation in the country. Still, the impact on the sector
            after all these years of war is present and relevant; the surviving varieties in closer markets
            have reduced their value to 28 percent of the 1977 value. Again, there may be several
            explanations, including misreporting and illegal trade, but this suggests that the traditional
            export sector is still alive in the country and is to be considered a way to increase cou ntry‘s
            This in fact is recognized by local authorities. The Export Promotion Agency of Afghanistan
            recognizes ―to build on export potential and capacities in traditional export sectors‖ is the
            most reasonable way to increase Afghan exports. 52 They focus on agribusiness, carpets,
            handicraft and marble, minerals and gemstones, and are particularly interested in neighbouring
            markets. Again, the selection responds to the ‗revealed comparative advantages‘ the country
            can exploit with higher chances of success, given the current capacity and infrastructure
            limits. It also makes sense from an occupational profile point of view, as by linking the rural
            areas to a dynamic sector, more and better opportunities for rural workers will exist. Sector
            diversification should be regarded as a sequential strategy, building upon a more solid
            traditional export sector.


DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Consider, for instance, the dried fruits market. In the U.K., the world largest importer of dried
            fruits, Afghanistan supplied in 2006 0.02 percent of the market (see Graph 7). But in India,
            Afghanistan is the third largest provider after Pakistan and Iran, occupying 7 percent of the
            Indian market in 2006. Or in Russia, where since 2001 Afghanistan has managed to double its
            raisin exports and has now a solid 9 percent of the local market. To prove what potential the
            country has when exporting to the Russian market, graph 9 shows the evolution of dried fruits
            imports in Russia. The market size has doubled since 2002, but the impact from 1998 -2002 and
            2004 droughts on Afghan exports is visible, emphasizing the need for larger investments in
            agriculture, in particular irrigation. Still, the competitive capacity of the Afghan industry is
            evident and a good base to be enhanced. Again transit through Uzbekistan is an additional
            issue to be considered.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            The ―edible nuts market‖ provides another interesting example (graph 10 and 11). Worldwide,
            this is a market bigger than Afghanistan‘s GDP with US$9 billions traded in 2006 (and almost
            doubling in size every ten years since 1980). Europe, the US and India are the main importers,
            and even if Afghanistan has preferential access to all these markets, packing, certification and
            other issues prevent the country from being an active player in the US and European markets. 53
            Out of the US$1 billion imported by the US in 2006, Afghanistan only contributed with
            US$200.000, or 0.02 percent. 54 But in India, Afghan products are considered premium, and the
            country has ample space to grow by competing with other low income countries.55 Again, it
            requires trade facilitation commitments for transit through Pakistan, but the market
            opportunity is there.

                In June 2002, preferential access to the European markets was obtained under the
            Everything But Arms agreement, and in January 2003, the United States granted Afghanistan
            GSP access to its domestic market. In March 2003, Afghanistan and India signed a new
            preferential trade agreement by which India granted 50–100 percent tariff reductions on 38
            export items from Afghanistan, and duty-free access has been given to India for eight tariff
            lines. In January 2008, Afghanistan became a member of the South Asian Free Trade Area
            (India, Pakistan, Nepal, Sri Lanka, Bangladesh, Bhutan and the Maldives), as a member
            Afghanistan will receive the benefit of zero import duty by India on 4,536 tariff lines.
            When asked about access to high end market, representatives from the exports association
            mentioned quality or certification are less of an issue, based on them business v isas to travel to
            these countries is what limits their capacity to de business the most.
               Afghan exports to Germany, the largest world consumer in this market, are only one -half
            million dollars, or 0.04 percent of the German market.
               Afghan products are paid US$ 5.60 to US$ 7 per kg for, in comparison to US$ 3.26 per kg for
            Indian products. (OTF (2006) Strategy and Action Plan for Afghanistan‘s Dried Fruit and Nuts
            Cluster cited by The Rapid Expansion of Dried Fruit and Nuts Export, Case study prepared fo r
            the Enabling Environment Conference, 2007)

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            A mentioned before Afghanistan has large mineral resources and mining has the potential to
            generate large investment inflows, exports, fiscal revenues and GDP growth. But all that is
            conditional to large investments in infrastructure, wise public management and sound
            regulation. Large mining in copper, iron or gold is a capital intensive industry, but the
            employment generation expected from construction and the indirect employment may be
            significantly large. Nowadays, mining accounts for an insignificant 1 percent of licit GDP. 56 But
            unknown amounts of precious and semiprecious stones, as well as marble are smuggled eac h
            year out of Afghanistan. The World Bank (2004b) suggests in the medium term (3 -5 years) and
            by means of formalizing existing mining operations and exports, the Government could collect
            taxes and royalties estimated at up to US$20 million per annum.
            Deposits of precious metals and stones, coal, oil and gas and other resources are known to be
            large and commercially viable. The Hajigak iron ore deposits are ranked ―world class‖ by the
            British Geological Society and reportedly the largest undeveloped iron ore deposit in the
            Middle East.57 The Aynak copper mine, the second largest unexploited deposit in the world, to
            be exploited by the China Metallurgical Group, may propel Afghanistan into the list of the
            world‘s top 15 copper producers. Investment is estimated to be above US$3 billion or almost 30
            percent of nondrug GDP in 2007 (IMF, WEO 2008). Royalties and other contractual details
            are to date confidential, but the project is expected to generate US$390 million per annum to
            fiscal coffers. 58 The potential revenues from gold mines are $150 million per year by 2015
            (World Bank, 2004b).
            Mining is a potential source for long-term exports, GDP growth and fiscal sustainability in
            Afghanistan, but as mentioned projects like these demand large infrastructure investme nts and
            regulatory and enforcement capacity, not to mention security. The risks associated to windfall
            revenues management, including macro volatility, exchange rate appreciation and corruption
            are all exacerbated in Afghanistan.
            Trade Policies
            We noticed on our field work in Kabul government and donors alike having conflicting ideas
            on the desirable and feasible trade policy the country should follow. Import substitution and
            export-led growth, including full-accession to WTO are incompatible. On the one hand, the

               Consisting primarily of coal, quarry materials, marble and dimension stone, industrial
            minerals, some metals, and semi-precious stones.
               Afghanistan Geological Survey, http://www.bgs.ac.uk/AfghanMinerals/
               Huntzinger, 2007.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            pressing need to increase revenues and the current importance of trade taxes (customs duties
            represented 40 percent of total revenues in 2006/2007) may sound attractive to some as an
            easy way to increase fiscal capacity. On the other hand, the low competitiveness of the Afghan
            economy, with high cost of doing business and currency appreciation due to inflows of donor
            funds and drug monies, may suggest to some the only way to compete is to find a niche not
            being occupied (and different from traditional exports, since neighbours are producing the
            same basket) and increase tariffs to protect local industry from imports. Both visions are not
            only wrong, but policies based on it are doomed to fail.
            In both cases, proponents of these strategies neglect how porous borders are (illicit trade is a
            daily testimony of that) and how high evasion is for the currently low tariff schedule. 59 High
            tariffs will only increase the payoff of eluding or bribing the custom agency and the arbitrage
            will induce illicit trade to compete with the local industry it tries to protect. Also, a tariff -
            protected urban-manufacturing sector will generate unsustainable employment in urban areas,
            leaving rural population (80 percent of total) behind.
            We consider the export-led growth strategy initially based on rural and agribusiness
            traditional sectors, and focused on regional and traditional partners (again, this seems to be
            the strategy EPAA has chosen) to be the right one in inserting Afghanistan to the world
            economy. These sectors are mostly based on rural labour, with added value in urban areas and
            demand the skills being provided and abundant in the country. They will be less disruptive in
            terms of migration and employment opportunities and build upon a long tradition with some
            re-emerging but solid trading networks (as proved by the successful boom in exports, growing
            faster than GDP since 2001). In the best case scenario, it is also the strategy that would
            complement strategies toward reducing opium dependency in rural areas. In th e medium and
            long term, diversification of both partners and sectors should be targeted once a more solid
            trading capacity, export-linked value chain and infrastructure has been constructed. An
            export-growth strategy is not spontaneous, as many coordination failures and externalities
            have to be corrected. Some of these will demand a direct intervention from the government in
            building up or subsidizing export terminals, irrigation systems, connecting roads, industrial
            parks, cool-storage facilities, etc.; but in time, these interventions will be in contradiction with
            full WTO accession.
            Afghan accession to WTO, if followed in 2010 as scheduled, should be context -wise in its
            negotiation, and set a comfortable schedule for accession (i.e., more time and better terms).
            WTO accession might be desirable; as it would be a good way to underline the path reforms
            will go in several sectors and will signal investors and entrepreneurs the government
            commitment‘s to private sector development. It might also give the count ry some leverage and
            a forum to force Pakistan (a WTO member) to comply with trade facilitation and other issues.
            But again, as the HRV framework suggests, self-discovery is a process where government
            intervention is needed in presence of market failures and coordination externalities, and some
            of those interventions will conflict with the WTO agreements.60
            Afghan export package continues to be dominated by a few traditional goods and concentrated
            in regional countries. We consider there is space for increasing market share in those countries

               Take for instance the Pakistan-Afghan border with: two ―established‖ border crossings point
            cover the bulk of legal trade (Torkham and Chaman); twenty border-crossing routes are manned by
            customs officials and clan police (Levies); three hundred and fifty illegal and unmanned crossings
            exist, plus hundreds of foot and goat paths used by smugglers, locals, and nomads (Johnson and
            Mason, 2008).
               For example, domestic support (e.g., tractors or seeds support) conflict with WTO if they
            are focused on specific sectors. Or for instance antidumping charges might be raised against
            the country if the Afghan Government supports a sector (e.g., other countries raisin exports
            could challenge Afghanistan measures to boost raisin exports on the basis of dumping).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            (especially India and Russia) and diversify destinations while constructing a solid exports
            base on traditional agribusiness. This short-term strategy is already showing its payoffs in
            terms of export growth, but as mentioned, lack of specific infrastructure will reduce the
            chances for these exports to locally add-value and expand. In particular, drought prone areas
            should invest heavily on irrigation as a way to allow for the scale -up of agro-exports. Trade
            facilitation issues with Pakistan should also be top in the priorities for the Afghan government,
            as it undermines the opportunities for Afghan exports to consolidate and expand its market
            participation in India and beyond. Similarly, relations with the Central Asian Republics need
            to be strengthened (for exports to Russia), customs procedures harmonized and transit
            facilities improved. In the long term, in order to achieve external sustainability given the
            current levels of imports and to substitute opium as the most relevant cash crop and export,
            the tradable sector will need to develop new export opportunities (i.e., diversify into new
            sectors) and expand further into the international market. Mining is potentially an important
            source of exports and fiscal revenues, but as the other sectors country‘s capacity to really
            developed the sector is dependent on governance and infrastructure.

            5.3.2      Government Failures
      Macro Risks
            There is consensus that macroeconomic policy has been an area of success in post-Taliban
            Afghanistan and the currency reform in 2002/03 is frequently heralded as one of the
            achievements of post-Taliban economic policy. Fiscal policy is also solid, but the low revenue
            base makes fiscal sustainability an issue, with donor support a necessity in the medium to
            long-term. The government covers a minimum share of government expenditures and cannot
            meet the requirements of the annual operating expenditures (payroll and maintenance)
            without external aid, let alone infrastructure and development expenditures.
            Foreign aid has helped the economy to be able to stabilize with low to medium levels of
            inflation and low volatility in the exchange rate. These conditions are necessary to promote
            private investment to foster economic development in the medium and long term. After debt
            reduction, the stock of debt is now considerably smaller than compared to the average
            developing country, allowing the country enough room for an effective macroeconomic
            management in the future. There are considerable fiscal and macroeconomic risks in general
            associated with volatility in foreign aid inflows. Even more worrisome as far as long -term
            sustainability is concerned, is the external financing of police salaries (LOTFA) and defence
            expenditures (USA).
            Macroeconomic Assessment
            In general the macroeconomic management is prudent with limited intervention in the
            economy, being able to slowly build institutional confidence needed to achieve economic
            stability. Fiscal accounts are now balanced thanks to foreign aid, but a substantial effort is
            going to be needed in order to move towards domestic sustainability without depending on
            donors‘ money in the long run. This is aggravated by the fact that an important share of fiscal
            revenues come from sources like tariffs and perceived over taxation to the formal sector.
            However we need to recognize the effort made in recent years to reduce the core operating
            budget deficit successfully through a combination of relatively careful public sector spending
            and a significant increase in fiscal revenues. While the importance of the objective is
            acknowledged by the government, the means are less carefully considered.
            Even though the macroeconomic situation is currently highly dependent on foreign aid, since
            foreign aid is expected to continue being high for the foreseeable future we expect
            macroeconomic conditions not to be an effective constraint to growth. Currently, there is no
            evidence in the macro variables of reactions from the private sector on this. As foreign aid
            starts to decrease it is likely that the exchange rate will depreciate stimulating even further the
            export sector and in consequence compensating for the reduction in foreign aid income, but
            again opium related hard currency inflows may counterbalance this.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Imports are considerably high reflecting the reconstruction efforts and low domestic
            production. Machinery and equipments have ranked consistently as the single leading import
            category. Total imports have also been affected by the energy bill associated with recent hig her
            prices of hydrocarbons. However, recently export growth rates have been higher than import
            growth rates, making external sustainability without foreign aid a possibility in the future.
            The role of drug money in the external balance remains a contested issue, but both the World
            Bank and the IMF anticipate a negative macroeconomic impact coming from counter -
            Fiscal Sustainability
            Public debt sustainability is not high on the list of Afghanistan‘s barriers to growth. The
            projected path for public debt in Afghanistan does not seem predominantly worrisome,
            particularly after debt reduction. However, much depends on assumptions that are
            considerably uncertain. Based on IMF projections, growth is expected to remain strong thanks
            to donor-financed reconstruction efforts and an expected improvement on security, a stable
            macroeconomic outlook and reforms favourable to the development of the private sector.
            The sustainability analysis is shown in Table 20. The first table summarizes the assumptions
            on key variables: r (real interest rate), g (real GDP growth), PB/GDP (primary balance as a % of
            GDP, according to IMF medium term estimates) and public debt to GDP ratio. We consider
            three growth and interest rates scenarios (a central scenario and a low and hi gh bound). Given
            the assumptions, the second table (Required Primary Balance) computes how big the primary
            balance would need to be in each case (combination of growth and interest rate scenarios) to
            achieve the public debt to GDP target (30%). The third table (clockwise), computes the
            difference between the medium term IMF estimate fiscal primary balance and the required
            fiscal balance needed to stabilize debt to GDP around 30%. The final table (Steady State Debt
            to GDP ratio) shows the debt to GDP level towards which the economy would move in each
            scenario assuming IMF‘s primary balance estimates.

                                                   Fiscal analysis
            Table 20. Afghanistan, fiscal sustainability sustainability
                                       Assumptions and initial data                                                                  Required Primary Balance
                                                                                                                                                        Economic growth

                                         High           Central            Low                                                               High (7,0%) Central (5,5%)   Low (4,0%)
                                                                                                       Real interest rate

                   g                     7.0%            5.5%             4.0%                                               High (2,0%)       -1.50%        -1.05%        -0.60%

                     r                   2.0%            0.0%             -2.0%                                             Central (0,0%)     -2.10%        -1.65%        -1.20%

             PB/GDP                                   Public Debt/
                                         -2.2%                            30.0%                                              Low (-2,0%)       -2.70%        -2.25%        -1.80%
              target                                   GDP target

                                                                                                    Assuming Public Debt/GDP target= 30%

                                             Steady State Debt-to-GDP ratio                                                                Fiscal Resource Gap
                                                                     Economic growth                                                                    Economic growth

                                                      High (7,0%)     Central (5,5%)   Low (4,0%)                                            High (7,0%) Central (5,5%)   Low (4,0%)
                Real interest rate

                                                                                                       Real interest rate

                                      High (2,0%)        44.0%            62.9%         110.0%                               High (2,0%)       0.70%         1.15%          1.60%

                                     Central (0,0%)      31.4%            40.0%          55.0%                              Central (0,0%)     0.10%         0.55%          1.00%

                                      Low (-2,0%)        24.4%            29.3%          36.7%                               Low (-2,0%)       -0.50%        -0.05%         0.40%

             Assuming PB/GDP= -2.2%                                                                 Assuming Public Debt/GDP target= 30%

            Source: Authors calculation using IMF (2008a, 2008c)

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            According to our sustainability exercise, mostly based on IMF assumptions on the behaviour
            of fiscal revenues, expenditure and growth, the primary deficit stabilizes around 2.2% of GDP.
            This fiscal result is enough to stabilize the public debt to GDP ratio around 40%. It is worth
            noting that the result is driven by the assumption that fiscal revenues stabilize around 12.7% of
            GDP in the long run, which is low even for developing countries. But of note, it is much better
            than the 4.7 percent of GDP in 2003/04, or even the recent 8.2 percent in 2006/07. If the primary
            deficit were to stabilize around 1.7% of GDP then the debt to GDP ratio would stabilize below
            a 30% level, this would be the case if revenues were able to reach 13.2% of GDP, above the IMF
            long term estimates.
            In other words, debt to GDP remains within the range of what is considered reasonable for
            developing economies. Expected low interest rates on concessional loans based on the poverty
            reduction facility agreement, World Bank and ADB contribution in the form of grants, high
            economic growth in the next few years and moderate growth in the long run and a small initial
            stock of debt, all play in favour of fiscal sustainability. Moreover, it is to expect that if the low
            growth scenario realizes, concessional debt will represent a higher share of total debt making
            real interest rates lower; this would act as a buffer mechanism in favour of fiscal sustainability.
            Even though the overall fiscal picture looks relatively favourable from a sustainability point of
            view, it is important to notice that improvements on fiscal revenues are very fragile. Currently
            about half of fiscal revenues (3.7% of GDP) come from taxes on international trade and
            transactions (see Table 22). This is regarded as a distortionary form of taxation and there will
            be pressures to further reduce tariffs in the future, 61 forcing the government to develop other
            sources of revenue. Even more, current levels of imports are not sustainable in the long term as
            they depend mostly on donor funds and to a lesser extent on opium as sources for hard
            currency. If we recognize external support and opium will be reduced in the long-term, then
            we acknowledge no long-term fiscal strategy can rely on tariffs as the main source for revenue
                 Table 21. Afghanistan, revenue portfolio
                                        2003/      2004/     2005/     2006/   to         2006/
                 Type of Revenue        2004       2005      2006      2007    2006       2007
                                                                                  As share of
                                                                                total revenues
                                                  in million dollars                  (%)
                 Total Revenues          203.4       256     413.0     573.2      100      100
                 Domestic Taxes           13.7        46        57     207.8      12.5    36.3
                 Customs Duties           111.5     144.9    221.8     231.6       54     40.4
                 Administrative Fees      39.9       20.0     81.6      76.8      15.4     13.4
                 Sales of Goods and
                 Services                 36.2       38.5     35.2      41.6      14.2      7.3
                 Other non-tax
                 Revenues                 0.58        0.3     13.4       8.3       2.5      1.4
              Social Contributions       1.4    6.3       4     7.1      1.4     1.2
            Source: Ministry of Finance, Treasury (data in Afghani, converted to US$ using Afg50 as the
            exchange rate)
            This baseline scenario assumes the government is successful in developing these alternative
            sources of taxation creating minimum distortions. In this regard, the public sector needs to
            work together with donors and international organisations in order to ensure a stable path of

              More so if the country starts pursuing free trade agreements with its more important
            partners, as these agreements demand bilateral preferences and therefore impact revenues
            collection from custom.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            contributions to avoid transmitting instability into the economy. While the IMF maintains
            that due to the continued availability of external resources, the governme nt does not face short
            to medium term fiscal constraints, the lack of government discretionary finance and increased
            pressure to mobilize national revenues to cover ordinary budget spending undermine this
            In recent months, government efforts to mobilize revenues from the limited formal sector have
            become increasingly aggressive. As a result, some private sector interests related to the large
            tax payer office are currently considering relocating their business entities to reflect what is
            seen as an overly aggressive stance of government vis a vis tax collection. 62 Pressured to
            increase revenues, the government is taking inappropriate steps by overtaxing the formal
            sector (and offering no public goods in exchange), instead of broadening the tax str ucture (see
            next section below). Even if the fiscal vulnerability has thus far not been evidenced by any
            macro variable, as inflation, exchange rate (although credits and deposits are dollarized), etc.,
            it distorts the incentives and the behaviour of the government
            Macroeconomic volatility associated with fiscal or monetary mismanagement is not a
            constraining issue for the Afghan economy at present. It is important for the government to
            continue its prudent fiscal policy and macro-economic reforms, especially in times when the
            imported inflation due to food and fuel would test the real capacity of the authorities in terms
            of policies. Fiscal sustainability is a serious source of instability in the medium to longer term
            as the donor subsidized budget is bound to be reduced, and government capacity to attract
            investment and collect revenues is still uncertain. But we don‘t see currently any sign of
            macroeconomic imbalance or expectations of macrorisks and volatility.

        Micro Risks
            Incentives provided by government policies and behaviour can have a big impact on observed
            levels of investment in any given society, and can determine entrepreneurs‘ decisions to
            participate in the market, and whether to participate as a formal or informal actor. The
            institutional framework affects the capacity of investors to appropriate the returns of their
            productive efforts through the imposition of (or the expectation of) both formal and informal
            taxes on economic activity. If investments cannot be protected and appropriation of private
            returns is not guaranteed, it will obviously affect economic decisions, acting as a direct
            disincentive for investment and entrepreneurship. The informal economy is also affected by
            property rights issues and corruption. Even if not directly affecting the informal sector, the tax
            environment is weakening the incentives for the 10-15 percent (see The Informal Economy in
            Section 4.4) we estimate as informal trade, commerce and services that is potentially able to be
            formalized, with an additional negative impact on economic activity and investment. 63
            Entrepreneurs in the informal sector have the incentives to investment in lower risks projects
            (e.g. in real estate and trade) or transfer profits outside Afghanistan. Informality also prevents
            them from reaching the optimal size (i.e., no economies of scale) and force them to have an
            inefficient portfolio (i.e., diversified to manage risk), which prevents them from adopting more
            modern technologies (World Bank, 2005c).
            Following the GD, and supported mostly by qualitative analysis and perception-based data64,
            we consider the most binding constraint to investment in Afghanistan can be identified as
            micro risks under the umbrella of government failures. The framework suggests three key areas
            where microeconomic risks can affect investment decisions: property rights (rule of law),
            corruption and taxes. All three factors are deemed to increase uncertainty and worsen business
            environment by effectively reducing the appropriation of private returns to investment. The

                 Kabul hotel tax raid sparks cash exodus, The Financial Times, April 7 2008
                 We consider the subsistence agriculture and opium cannot be formalized.
              Analysis based upon select key informant interviews with large business owners in Kabul,
            donors and government officials and surveys and other sources as mentioned in the text.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            expectation of reduced appropriability will act as a deterrent for potential investors, reduce
            the investment of operating firms and may induce firms to leave the market or become
            Property Rights
            In Afghanistan, property rights protection is not only undermined by weak government and a
            corrupt judicial system but also by an increasingly insecure environment, which directly
            affects investment through the imposition of both additional risks and costs associated with
            security efforts. These basic ‗market creating institutions‘ preconditions for any growth
            strategy, are missing in today‘s Afghanistan. We have mentioned under ‗low social returns‘
            three complementary factors (human capital, infrastructure and geography) that are alread y
            lowering capital accumulation in the country by reducing the profitability of investments. If on
            top of expected low returns, businesspeople anticipate they will not be appropriated,
            investment will not be executed.
            Since 2005, there has been a sharp increase in security-related issues, linked both to
            criminality (abductions of middle-class businessmen and their family members, attacks on
            shipments and extortion) and to the insurgency (suicide attacks). ―If at least Kabul were
            secure‖, complained a businessmen. Lack of security and the perception of high insecurity,
            affects property rights and are to be considered as a micro risk investors will factor into their
            business plan as the expectations for returns appropriation is reduced. 65 But only large and
            very profitable firms can afford the security related costs and absorb the increased risk.
            Additionally, there has been an increase in security-related incidents related to the insurgency
            mostly in the East and South of the country; this has a negative impact on reconstruction
            efforts (reducing the crowding in effect of aid) and the development of licit private sector
            activities in the insurgency-affected areas. It also impacts negatively on confidence and
            perceptions of political and security risks in Afghanistan.
            Access to land, titling and property rights has been mentioned as one of the most severe
            impediments to investment in Afghanistan. The World Bank Investment Climate Assessment
            ranked it as the second most important issue limiting investors.66 The Investment Climate
            Survey (World Bank, 2005d) found that ‗nearly 56 percent of the surveyed firms who had tried
            to acquire new land in the past three years were unsuccessful and more than 10 percent
            reported having ongoing land disputes‘. The lack of reliable property rights is an important factor
            increasing the risk of owning land. 67 Widespread corruption and incapacity of the judicial sector
            make things much worse.
            Some companies solve the issue in part by the use of industrial parks. But more t ime is needed
            to thoroughly evaluate the role of industrial parks in overcoming the land and infrastructure
            constraints, as they are in the most parts still under construction, running at low capacity, or
            In Afghanistan, foreign investors cannot own land. They are also limited by 50 years,
            renewable for other 49 years when renting land. Given this short period and the uncertainty
            thereafter, it is not surprising than investors are reluctant to take the risk. The way most large
            companies had solved the issue is by negotiating with the government, who then allocate state

               The Afghanistan Investment Support Agency (AISA) reported a decline in the number of new
            businesses registrations from 2005/2006, and for the first time since 2001, private investment
            effectively diminished as a percentage of GDP (see Table 2).
               AREU qualitative research on markets confirmed the role of land access in inhibiting the
            growth of small manufacturers in Kabul (Paterson, 2006).
               Challenges presented by the Legal and Regularity Framework governing private for-profit
            activity, Background paper prepared for the Enabling Environment Conference, 2007.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            owned land by Presidential Decree. Mid size and ―not connected‖ companies simply cannot
            Policy uncertainty
            Uncertainty is exacerbated by the government and the donor community. There appear to be
            conflicting views within the cabinet and within donors regarding the role of government in the
            economy, exacerbating unpredictability for long-term investments. Vestiges of obsolete and
            non-market systems remain, and the total commitment to its elimination is not perceived as a
            priority by the Government, especially when some officials are advocators of such non-market
            policies. In addition, already poor levels of dialogue and coordination between government
            ministries and other bodies responsible for private sector policies, regulation and support have
            actually deteriorated. ―We are two steps forward, but three steps back‖, summarized it one
            Afghanistan‘s inherent fiscal weakness is not new in the country‘s history, as it has been
            always dependent on foreign aid (Output I, Political Economy Analysis, Understanding
            Afghanistan). Significant steps should be considered to reverse this dependency as a necessary
            condition for growth and development. The country should broaden the tax base by
            formalizing a large share of GDP that is still informal and therefore untaxed, and start
            preparing a strategy to shift from incidence and imports to consumption and income. But this
            is easier said than done, as no incentives to be formal are currently in place (actually, all
            incentives are now aligned to informality), reducing revenue from income but also the
            potential revenue from consumption taxes. At the same time, the potential revenues from
            natural resources exploitation are conditional to attract FDI into those sectors, and unlikely to
            be significant in the short term.
            With a 20 percent flat rate income tax for individuals and corporations, tax rates in
            Afghanistan are not high (compared to 35 percent for the region). 68 We in fact did not receive
            any complaint over the level of the corporate tax rate from our private sector informers. This
            current tax schedule is the result of a policy and administrative reform adopted in 2004 and
            early 2005, by which the corporate tax was reduced from 25 to 20 percent, and wages
            withholding and Business Receipts Tax (BRT) on selective services where introduced. The
            reform also created the Large Taxpayers Office (LTO). In terms of revenue collection, the
            reform did have an impact, moving collection from 4.7 percent of GDP in 2003/04, to 6.4 percent
            of GDP in 2005/06 and to 8.2 percent of GDP in 2006/07.
            The rationale for a LTO is well understood in the public finance literature as it is an
            institutional reform meant to isolate the large taxpayers from corrupted and incapable tax
            administrations. From the government‘s point of view, it is a way to reduce tax leakages and
            corruption and increase tax collection without increasing the tax rate, but instead by increasing
            voluntary compliance and reducing the costs and burden of paying the tax. According to the
            Ministry of Finance ―Taxpayers can expect responsiveness, courtesy, and professionalism in
            their dealings with LTO employees. This type of quality in service will allow the taxpayer
            access to the information he needs for full compliance with the tax laws.‖ As per our
            interviews, this was also the initial expectation from taxpayers, and the office was initially
            welcomed as it was expected to be an elite unit with skilled and transparent staff and with
            direct communication (as opposed to the Mustofiat). 69 To organize tax administration by

              Although this tax should be corrected by the level of public goods provided by the
            government. Firms pay a ―low‖ tax compared to other countries in the region, but a lso receive
            lower (virtually none) public goods, increasing the effective taxation.
               The Mustofiat is a provincial branch of the Ministry of Finance. These provincial finance
            offices are in charge of the provincial revenue collection, with responsibilitie s also in customs
            (clearance, valuation and classification) and perform broad tasks of financial management. A
            common issue in the tense relation between Kabul and the provinces is the transfer of tax

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            differentiating size and other characteristics is also a way to reduce heterogeneity and to tailor
            the management to the specific group. All foreign-owned entities with capital investments in
            Afghanistan exceeding US$10m or operating in oil, gas, banking, finance or international air -
            transport; state-owned enterprises in oil, gas, banking or air transport; and all other entities
            with annual sales over US$1m are categorized as large taxpayers.
            In fact, in line with this strategy, a new office to serve medium sized taxpayers office (MTO)
            was opened in September 2007. The new MTO will serve: businesses with sales over US$1m or
            US$10m ―but whose income is exempt or mostly exempt‖, diplomatic and international
            organisations, medical and educational institutions, foreign owned businesses, legal person
            (i.e. corporations, limited liability companies etc.) regardless of their gross revenue or invested
            capital, individuals and unregistered partnerships with annual gross revenues of 5 to 75 million
            afs (US$ 100.000 to 1.5 m) or invested capital of 15 to 150 million afs (US$ 300.000 to 3 m)
            Even if in principle the reform sounds the right thing to do, the results in practice hav e been
            quite different. Since the reform, the government efforts to mobilize revenues from the limited
            formal sector have become increasingly aggressive. This is being carried out by adding new
            taxes, often based on decades old legislations or retroactive modifications. ―Nuisance taxes‖ in
            forms of several administrative fees and some retroactive taxes are being applied and charged
            by sector ministries and other agencies, many at provincial level. These taxes are not
            necessarily in line with the MoF‘s policies, a reflection of the lack of coordination amongst
            ministries and the lack of capacity in the judicial sector to set clearly the legislation in place by
            reviewing and revoking old laws. A large company in Afghanistan and one of the main
            taxpayers described the volatile and unpredictable tax environment as the most constraining
            issue for business. ―We are running on hope‖, –was the conclusion. Another company had to
            increase seven-fold the number of layers just to deal with the increasingly creative capac ity
            from the LTO to find new taxes or challenge lack of compliance. As a result, private sector
            interests related to the large tax payers‘ office are currently considering relocating their
            business entities to reflect what is seen as an overly aggressive stance of government vis a vis
            tax collection, including current accounts withholding based on estimated back payment.
            Unfortunately, the expected elite and transparent units are also under scrutiny due to
            corruption allegations (although never at the level of the Mustofiats). The two main
            complaints among large taxpayers are the ―selectivity‖ and unpredictability which with the
            LTO enforce the law. Well connected businesspeople and companies are often liberated from
            the law and simply do not pay taxes or pay a fraction that is never challenged by the LTO. The
            reduction since 2006 on FDI, private investment and the number of AISA -registered companies
            (more so for foreign capital companies) registered is, we argue, not a coincidence but a result
            of this strategy. There has been at the same time an increase in perception of insecurity and
            security-related events, but as per our key informers, security costs –even if high and rising—
            could be included into the business plan, but not the uncertainty on tax legisla tion
            enforcement and other forms of corruption.
            Failures in LTOs are due to: (i) discretionary taxation due to political intervention in favour of
            those who have the ‗right connections‘, (ii) unpredictable and volatile regulations, (iii)
            potential gains (and impunity) from involvement in corruption for officials and taxpayers, and
            (iv) poor pay and working conditions. The right model for Afghanistan should limit political
            interference by the Ministry of Finance or other power holders and free the tax administration
            from the constraints of the civil service system (including low wages). A combination of strong
            political support, managerial experience, administrative and financial autonomy, radical
            personnel reform (all staff members should be given the option of voluntary retirement or applying
            for a position in the new agency and taking a rigorous exam) including salaries based on
            comparable employment in the private sector.

            collections to the Central Government. Under the Tax Law 2005, the Mustafiats no longer
            collect taxes from firms listed under the Large Taxpayer Office.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            In 2005, the Investment Climate Assessment (World Bank, 2005d) placed corruption just
            behind access to land and electricity as the third most important barrier for business, with 58
            percent of respondents citing it as a severe barrier. Firms reported spending 8 percent of sales
            or the value of public contracts on bribes (compared to less than 2 percent in Pakistan).
            Foreign investors and Afghan Diaspora returning from overseas were identified as particularly
            threatened by corruption. By then (2005), Afghanistan scored 2.5 in Transparency
            International‘s corruption perception index, ranked 117 out of 158 but ahead of neighbouring
            Turkmenistan (1.8), Tajikistan (2.1), Pakistan (2.1), and Uzbekistan (2.2). But in 2007, it was
            ranked 172 out of 179 and the second worse in the region after Uzbekistan. According to the
            2007 Integrity Watch Afghanistan survey, the ―opium economy and other illicit activities‖
            coupled by large inflows of international aid 70, ―pressures to spend it quickly‖ and with little to
            no accountability (i.e., outside the national budget) provide for new opportunities for
            Afghanistan‘s constitution as presidential democracy places extended powers in the executive,
            but in practice, Afghanistan is politically fragmented and dominated by non -state powers with
            a tradition of strong decentralized local administration in the form of tribalism or warlordism.
            This leaves the state vulnerable to capture and patronage and thus to policy decisions tilted in
            favour of influential private interests, limiting the ability of non-traditional sectors to compete.
            Government credibility and commitment to the rules of the game, and certainty of enforcement
            can thus be questioned by investors and perceived as an added risk factor. This is suggested by
            investors‘ and other informers‘ perceptions of undue influence and favouritism in government
            decisions, suggesting the political strategy is skewed toward targeting benefits to a few as a
            mean to secure power (e.g., preferential tax conditions or even exceptions, licensees, market
            power allocation, etc.). (See the non-technical paper for a more elaborate debate on the issue).
            In addition to their impact on entrepreneurship and investment, perceived discretionar y rule
            enforcement has particularly important implications for political stability. The Integrity
            Watch Afghanistan survey (2007) reports: ―Some 60 percent of respondents perceived
            President Hamid Karzai‘s administration to be more corrupt than that of the Taliban,
            Mujahiddin or the Communist periods. While under previous government‘s ethnic ties or
            political leanings enabled corruption, today money has become all-important, and those with
            access to ready cash can buy government appointments, bypass justice or evade police.‖ 72
            According to the survey, the Ministry of Justice (courts) (52.6 percent), the Ministry of
            Interior (42.5 percent) and municipalities (37.9 percent) are perceived as the most corrupt
            sectors in public administration. Agencies delivering public services such as education, health,
            transport, water and electricity, and also customs are considered very corrupt ins titutions.
            Corruption has developed beyond the traditional gift (Baksheesh) considered to a be co-
            payment for the public service to compensate the extreme low wages of civil servants or to ―get
            things done‖, or the pre-war ―traditional‖ corruption based on patronage networks and tribal
            linkages and new forms of corruption are increasingly being ‗institutionalized‘ as a bribe
            extracting mafia connected by corruption networks (Band-Bazi). This takes the form of
            bureaucratic burden with increasing requirements of licenses and permits from state and non-
            state agents. 73 Power holders of various kinds create these opportunities and rents by limiting

               The country now receives aid inflows close to US$4 billion per year, compared to the US$
            300 million in the 90s.
               The link between drugs and corruption is direct and mutually reinforcing; corruption has
            been a major factor explaining the limited success of counter-narcotics efforts thus far and the
            booming opium industry has contributed to extend corruption and increased crime (Shaw, 2006).
               Yama Torabi, Lorenzo Delesgues, Afghan Perceptions of Corruption A Survey Across
            Thirteen Provinces Integrity Watch Afghanistan, January 2007
               The World Bank Doing Business report captured this issue by downgrading ―Dealing with
            licenses‖ for Afghanistan from 139 (2007) to 141 (2008) (out of 178 economies).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            access to resources and limiting business opportunities, which is noticeable as several
            ministries and sub national representations now issue sector-specific licenses and local
            permits as a means to increase revenues. Corruption is endogenous to this multilayer of
            authorities and agencies (e.g., it takes 20 days of paperwork to export to Pakistan, and it
            requires almost 50 signatures), and finally rooted in an inherent weakness in the
            administrative system and law enforcement that create an environment of impunity. Knowing
            the country will depend for many years on aid, and may become a large exporter of natural
            resources only make this worse, as the expectation of large rents coming from future aid
            inflows and country‘s mineral resources makes rent seeking and cronyism an even better
            The worsening in the perception of corruption happened in spite (or as we wi ll argue next:
            because) of several attempts to streamline official procedures for business registrations, the
            creation of anti-corruption agencies and the intention to reduce tax leakage by improving the
            tax collection process in a centralized way. In a patrimonial society, based on rent distribution
            and patronage, every attempt to reduce corruption or state capture will be challenged by new
            sources of informal finance meant to cover the lost revenue. In addition, efforts aimed at
            reducing corruption have a localized impact as in an inherently weak institutional framework
            characterized by an ill defined set of agencies and responsibilities, corruption always find its
            way out again. We argue this process is happening in Afghanistan currently. In fact, this ca n
            rationalize why the country is now perceived as more corrupt than ever, even when anti-
            corruption ―agencies‖ are in place. 74 Or how, at the same time the World Bank Governance
            Indicators improves Afghanistan index on anti-corruption, but the perception of corruption
            worsens (Transparency International (2008), Integrity Watch Afghanistan (2008)). We argue
            this is not only ―more‖ corruption, but also ―new‖ corruption being used as informal taxation
            (please refer to Output I for a theoretical discussion on patrimonial societies, and to Output III
            non-technical paper for more elaboration).
            The impact on investment from these practices can be easily anticipated. Entrepreneurs and
            investors will limit their exposure. Different strategies can be followed to cope w ith this issue,
            but none favourable to the Afghan economy. Either investment will be reconsidered, reduced
            to minimize being targeted or pursued informally. In any case the result will be the same, and
            the one already mentioned, less formal investment. Even more, this has ramifications that are
            critical for any market economy to perform. It has been mentioned that the high costs of
            securing credit by means of collateralisation and the way banks are reluctant to consider them
            based on Courts‘ inefficiency and corruption. Informers in our field-trip to Kabul referred to
            corruption in the electricity company as one factor leading them to use on -site diesel
            generation, since even if more expensive they are at least certain. The links between land
            titling and corruption had been also mentioned.
            Pressured to increase revenues, the government is taking the wrong steps by overtaxing the
            formal sector (while offering no public goods in exchange) and increasing tax uncertainty

               In 2004, the General Independent Administration of Anti-Corruption (GIAAC) was
            established by law as the leading anti-corruption agency with a broad mandate to promote
            awareness, preventive policies and investigate corruption. But GIAAC lacks high -level
            political support, and is not an autonomous agency (no separate budget), in addition the
            Parliament voted for it dismissal. In 2006, the Attorney General (AG) was given a mandate
            from President Karzai to investigate corruption in the public sector (is was i nitially intended
            to receive the corruption cases from GIAAC and to enquire and deciding sending them to court
            or not). In 2006, President Karzai appointed members to a High Level Anti -Corruption
            Commission (high level leaders from Ministries, Parliament and the judicial sector) with a
            mandate to address public sector corruption and to propose an action plan to be implemented
            as part of the ANDS. In some cases, especially when cases involve political figures, the Supreme
            Court has a role on oversight corruption cases.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            especially to large taxpayers, instead of broadening the tax structure by means of increasing
            investment and formalizing part of the legal but informal economy. This perverse equilibrium
            of over-taxation, corruption and selective law enforcement, coupled with under -provision of
            all public goods (including law and order and property rights) will inhibit further investment
            and alienate potential formal investors, thereby setting in motion a vicious cycle leading to
            greater informality and the need for more tax aggressiveness. Revenue generat ion capacity, by
            means of increasing the tax base should be secured, but the temptation to overtax the formal
            sector should be curbed. It is by bringing the legal-informal economy into a taxable scheme
            and attracting new formal business under a predictable tax structure and stable economic and
            business environment that a long term stable revenue base will be created.
            Based on the available evidence used to distinguish between the most immediate constraints
            and those that are less important in the short term, we consider that micro risks, under the
            ‗government failures‘ branch of the GD framework is at present time the most binding
            constraint to growth in Afghanistan. In particular, we identified the tax reform in 2004/2005
            as one possible shock explaining the reduction and stagnation in private investment (including
            FDI) since then. As the government becomes more centralized and tax collection more efficient
            (even if ―selective‖), illegal sources of income are being eliminated, but other forms appear as
            ―nuisance‖ taxes, administrative fees and increased corruption. It is this combination of tax
            voracity and uncertainty on formal and large tax payers and the perception of rampant
            corruption what limits private investment in Afghanistan today. And this, we argu e is the most
            binding constraint to growth.

      6     Conclusions
            We applied the GD methodology developed by Hausmann, Rodrik and Velasco (2005), to
            identify Afghanistan‘s most constraining issues to private investment. The HRV framework
            seeks to prioritize the economic distortions a country should eliminate or reduce in an effort to
            enhance and sustain economic growth. Paucity of data and the post-conflict situation makes it
            difficult to conduct a thorough study of growth barriers in Afghanistan and t he dominance of
            the illegal (opium) economy inhibits an accurate holistic understanding of the Afghan
            economy and its dynamics. Absence of primary data impeded efforts to conduct more robust
            empirical tests that would more accurately support our findings, and therefo re their
            interpretation and the rankings proposed should be taken with caution. Nonetheless, we have
            used all available information to rank the most pressing urgencies in terms of private
            investment, or in the GD jargon, the most binding constraint. This paper presents the technical
            part of the HRV, to be complemented with further analysis in the non-technical report.
            Economic activity in Afghanistan returned to double-digit growth rates in 2007 as a
            consequence of both a recovering in the agriculture sector after the 2006 drought and large
            demand impulse based on donor financed spending and drug money. Growth in the medium
            and long term is expected to decline, and will depend on improvement on governance and a
            successful integration into the world economy. Still, private sector investment has not
            recuperated from peak levels in 2005.
            The findings of the GD analysis remind us how fragile the Afghan economy is and that much
            more should still be done in order to put the country on a path for sustainable development.
            We find severe deficiencies in every aspect covered by the GD framework, maybe with the
            exception of macro volatility. We have shown access to formal finance is extremely limited and
            most investments are self-funded. When analyzing the financial market, we conclude this is
            due to poor intermediation and not lack of domestic savings. Banks are not providing credit to
            the private sector and limit their activities to basic services focused on urban areas, foreign and
            donor community. This is in part due to severe property rights and rule enforcement issues,
            also discussed.
            We then moved to analyze returns to economic activity. The country has very low quantity
            and poor quality of human capital, but there is no evidence this is a constraining factor for the
            private sector at present time. We have mentioned the lack of human capital is more of a
            reason for government failure and low efficiency (i.e., it goes in a different branch in the GD).

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Geographical challenges, including being landlocked and the challenges that impose for
            Afghan exports as they should transit into neighbouring countries, add to the list of
            constraining issues. We argue, geography combined with the lack of adequate infrastructure in
            particular related to the agribusiness (transportation, irrigation, value-chain links) and other
            potentially exporting sectors, appear to be the second most binding constraint to economic
            activity, and it is the most important challenge for the country in long-term perspective. We
            discussed under ―self-discovery‖ and competitiveness issues how big of an impact
            infrastructure and transit to reach ports and market is for the exporting sectors. Even if the
            country lacks on export sophistication, we found ‗revealed comparative advantages‘ are solid
            on traditional exports and regional markets seem to be the most promising in the near future
            in terms of expanding trade volume. We concluded ‗self-discovery‘ is not the cause for low
            levels of private investment in Afghanistan today. Exports, and therefore investments i n the
            tradable sector, are more constrained by lack of infrastructure in the value chain, and by
            geopolitical issues dealing with transit through Pakistan and the Central Asian Republics than
            for absence of competitive exports. But even these sectors do respond to opportunities and
            show impressive growth and market share recovery, suggesting they are still able to overcome
            the infrastructure limitation even if at high cost in terms of reduced margins and added -value.
            The evidence presented in this study, and supported by the HRV framework suggests that at
            present, the most binding constraint to growth in Afghanistan lies in poor appropriability of
            returns on private investment, due to microeconomic risks in the form of weak rule of law,
            taxation on the formal sector and increased corruption. We argue the decline in private
            investment recorded in 2005/2006 may be due to the change in tax policies into an
            unpredictable and unfair environment, especially for the large taxpayers and foreign
            companies. Because of the projected reduction of donor‘s funding and the international
            pressure on the Afghan Government to increase revenues, the government response by
            reforming the tax system and over-taxing (and under-providing) the large and international
            companies, reduced the incentives for investments in the formal sector. AISA recorded sharp
            reductions on foreign investors and IMF reported stagnation on private and FDI investment as
            percent of GDP may be a direct effect of this. Tax reforms carried out in 2004, and anti-
            corruption efforts sustained since then, created a more centralize d collection system that
            limited ―taxation‖ capacity in informal and sub-national agents. Those agents, we claim,
            embedded into a patrimonial system (Output I, Political Economy Analysis, Understanding
            Afghanistan) reacted by creating new sources of informal funds, in the form of ―nuisance
            taxes‖, fees, licenses and open corruption. If the Medium Size Taxpayers office would behave
            as its predecessor, it is not good news for mid-size companies and for the economy as a whole
            given that they would also reduce investment and go informal when possible. The result would
            inevitably be lower levels of private formal investment, and a vicious dynamic circle by which
            less formal investment will be completed and therefore more voracity should the government
            use to increase revenues.
            We argue the business environment now in place is not conducive to development of the
            private sector, and therefore is not conducive to growth or a long-term sustainable fiscal
            structure. As discussed, tax collection in Afghanistan is characterized by numerous ―nuisance
            taxes‖ that are both low-yielding and highly distortive, illicit revenue collection by local and
            sectoral authorities, double-taxation, selective application of tax legislation, and a widely
            perceived corrupt judicial system. The impact on economic activity from these cannot be
            estimated, but is likely to be very high. Even worse, the dynamic being created here can only
            complicate sustainability in the long-term as it harms directly the large taxpayers, and create
            all the incentives for the rest to remain informal. In the short-term, and a precondition to any
            ―growth-strategy‖ these issues should be solved.
            Conditional on improvement in this urgent issue (micro risks), we have identified
            infrastructure as a medium term constraint. A ―growth strategy‖ we argue that is sustainable
            for Afghanistan is one based on its comparative advantages (export-lead) in the traditional
            sector in the short term, to move into a higher value and diversified export portfolio (including
            mining) in mid to longer term. Proximity to large and dynamic markets (India, China, Russia,

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            the Gulf,) will play in favour of this, but that strategy is conditional on successfully linking the
            country within and to key international markets. If these two constraints are relaxed, then the
            country is likely to hit credit as the most limiting factor for entrepreneurship and investment
            to take off. As of now, we don‘t see human capital (conditional on region al labour mobility
            being allowed), macro volatility (conditional on fiscal and monetary prudency) or market
            failures (conditional on improved business environment) as potential constraints for private
            investment in the medium term.

      7     Policy Recommendations
            The policy recommendations are complex and would not be easy to implement. They demand
            strong political will and leadership, as they relate in a broader sense to the centre -periphery
            relationship and the central government capacity to force sub-national actors to comply. They
            are discussed in greater detail in the non-technical paper, but we provide an overview here.
            In line with our conclusions, we consider the most urgent reforms in Afghanistan should focus
            the energy and political and human capital, and donors support in both technical capacity and
            funding on (i) further reforming the tax collection process and agents, and reducing the
            ―nuisance‖ taxes and multiplicity of fees; (ii) addressing the widespread corruption, (iii)
            focusing investments in infrastructure projects directly related to increase competitiveness in
            the tradable sector, and (iv) securing trade facilitation and transit rights in the region to
            effectively exploit the regional and world market. In particular:
                The tax reform should be reviewed. Afghanistan needs a semi-autonomous revenue
                 authority (like DAB), isolated from political pressures (out of the MoF) and stuffed by
                 elite professionals (out of the civil service). The Large Taxpayers Office is the first stop to
                 address this is issue, as it affects the most important players in the economy with an
                 important multiplier effect.
                Sustained efforts to reduce corruption at all levels should be enhanced. The Roadmap for
                 Strategy and Action to fight corruption proposed by the join donor co mmunity (including
                 DFID) should be considered, revised and implemented.
                The endogenous corruption should be limited by cutting red tape, concentrating all
                 licenses and permits into a one-shop model and empowering one single autonomous
                 revenue authority. A broad action is required to do this, including the strengthening of
                 sub-national governing units and greater compliance to Central Government, as well as
                 greater clarity on the role of provincial administration in planning and executing the
                An autonomous anti-corruption agency with clear mandate to detect and prosecute,
                 independent budget and high-level political support should be created. It should be
                 directly dependent on the President as a way to isolate it from political pressures.
                Prevention and awareness are two other key components of any anti-corruption
                 campaign, and the donors play an important role on it as one of the causes for corruption
                 is aid.
                Investments in infrastructure beyond roads and energy are required. Irrigation, value -
                 added related infrastructure, productivity enhancing projects, connecting roads and other
                 specific investments in the agribusiness industry should be considered a priority. Specific
                 public goods should be provided into the sector, like certification, training, e tc., aimed to
                 enhance the tradability of the Afghan exports and its capacity to reach high end markets.
                 Access to land should be improved by distributing state-owned land or by providing more
                 and more accessible industrial parks.
                A coordinated effort between Afghanistan, the region and the international community is
                 required to secure trade facilitation, trade transit and customs collaboration. Pakistan
                 and the Central Asian Republics trade facilitation is key for any long term export -lead
                 growth strategy.

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            ADB, DFID, UNDP, UNODC and WB (2007), ‗Fighting Corruption in Afghanistan, a Roadmap
            for Strategy and Action‘
            ADB (2007), ‗Islamic Republic of Afghanistan: Establishing a Gas Regulatory Framework‘
            ADB (2008), ‗Asia Economic Outlook: Islamic Republic of Afghanistan‘
            Afghanistan Geological Survey (2007), ‗Preliminary Assessment of Non-Fuel Mineral
            Resources of Afghanistan‘
            ANDS (2007a), Poverty Profile (Draft)
            ANDS (2007b), ‗Power sector strategy for the Afghanistan National Development Strate gy
            ANDS (2008) ‗Private Sector Development and Trade Sector Strategy‘
            Bhattacharya, R., Clements, B., Gupta, S., Mattina, T., Segura-Ubiergo, A., and Tareq, S. (2005),
            ‗Rebuilding Fiscal Institutions in Postconflict Countries‘, IMF Occasional Pa per
            Buddenberg, D. and Byrd, W. editors (2007), ‗Afghanistan‘s drug industry, structure,
            funcionating, dynamics, and implications for counter-narcotics policy‘, UNODC and World
            Byrd, W., Mansfield, D., Oldham, D., and Ward, C. (2008) ‗Afghanistan, Eco nomic Incentives
            and Development, Initiatives to reduce opium production‘, World Bank
            Byrd, W. (2008), ‗Responding to Afghanistan‘s Opium Economy Challenge: Lessons and Policy
            Implications from a Development Perspective‘, World Bank
            Byrd, W. (2007), ‗Responding to Afghanistan‘s Development Challenge, an Assessment of
            experience during 2002-2007 and issues and priorities for the future‘, World Bank
            Byrd, W., and Ward, C. (2004), ‗Drugs and Development in Afghanistan‘ Social Development
            Papers, No. 18 , World Bank
            Byrd, W., and Ward, C. (2004), ‗Afghanistan‘s Opium Economy‘, World Bank
            Byrd, W, Raiser, M, Dobronogov, A and Kitain, A. (2005), Prospects for Regional Development and
            Economic Cooperation in the Wider Central Asian Region. Paper prepared for the Kabul Conference on
            Regional Economic Cooperation
            Barnett, R. (2000), ‗The Political Economy of War and Peace in Afghanistan‘ World
            Development 28(10): 1789-1803.
            Bhatia, M., Lanigan, K., Wilkinson, P. (2004), ‗Minimal Investments, Minimal Results, The
            Failure of Security Policy in Afghanistan‘, AREU Briefing Paper
            Collier, P and Hoeffler (2002) ‗Aid, Policy and Growth in Post-Conflict Societies,‘ World Bank
            Central Statistics Office (CSO) & United Nations Population Fund (UNFPA) (2006),
            Afghanistan: A Socio-Economic and Demographic Profile, Household Listing, 2003-2005, CSO & UNFPA
            Da Afghanistan Bank (2007), ‗Da Afghanistan Annual Bulletin (2006-2007)‘
            Demekas, D., McHugh, J., and Kosma, T. (2002), ‗The Economics of Post Conflict Aid‘, IMF
            Deng, Lual A. and Wema J. Kategile (2004), ‗Institutional Reform and Capacity Development
            in Post-Conflict Countries: Can Countries Emerging Out of Civil Wars Get Institutions
            Right?‘ Paper prepared for the World Bank Institute (WBI).
            DFID (2006), ‗Asia Fact sheet, Afghanistan‘

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            FAO (2003), Afghanistan: Survey of the Horticulture Sector
            FAO (2003a), Afghanistan: Countrywide assessment and inventory survey of traditional irrigation
            Faye, M L, Mcarthur, J W, Sachs, J D & Snow, T (2004), ‗The Challenges Fac ing Land-locked
            Developing Countries‘, Journal of Human Development, Vol. 5, No. 1, March 2004: 31-68
            FEWSNET (2005) Afghanistan Rural Livelihoods Profiles, USAID/FEWSNET
            Goodhand, J, (2003), ‗Frontiers and Wars: a study of the opium economy in         Afghanistan‘,
            DRAFT, University of London
            Grace, J. and Pain, A. (2004), ‗Rethinking Rural Livelihoods in Afghanistan‖, AREU
            Guimbert, S. (2004), ‗Structure and Performance of the Afghan Economy‖, World Bank
            Hausmann, R., Rodrik, D. and Velasco, A. (2005),                        ‗Growth      Diagnostic‘,
            Hausmann, R. and Wagner, R. (2008), ‗"Doing Growth Diagnostics in Practice", Harvard
            Mimeo (unpublished)
            Huntzinger, E. (2007), ‗Aynak Copper Mine, Opportunities and Threats for Development from
            a Sustainable Business Perspective, A Report by Integrity Watch Afghanistan.
            IMF (2008a), Islamic Republic of Afghanistan: 2007 Article IV Consultation
            IMF (2008b), Islamic Republic of Afghanistan: Selected Issues
            IMF (2008c), Islamic Republic of Afghanistan: Statistical Appendix
            IMF (2008d), Poverty Reduction Strategy Paper, Progress Report
            Islamic Republic of Afghanistan, Central Statistical Office (2007), ‗Estim ated population of
            Afghanistan 2007-2008‘
            Johnson, T., and Mason, C. (2008), No sign until the burst of dire: understanding the Pakistan -
            Afghanistan frontier.
            Lister, S., Pain, A. (2004), ‗Trading in Power: The Politics of ―Free‖ Markets in Afghanistan,‘
            Lücke, M. and Rothert, J. (2006), ‗Central Asia‘s Comparative Advantage in International
            Trade‘, Kiel Institute for the World Economy
            Maddison, A. (2001), ‗The World Economy: a Millennial Perspective‘, OECD.
            Maletta, H. (2007), ‗Arable Land Tenure in Afghanistan in the early post-Taliban Era‘, African
            and Asian Studies 6 (2007) 13-52
            Manabu, F. (2004), ‗ADBI RESEARCH POLICY BRIEF NO.12 UPDATE,‘ ADB Institute
            Mansfield, D. and Pain, A., (2007) ‗Evidence from the Field: Understanding Changing Levels of
            Opium Poppy Cultivation in Afghanistan‘, AREU
            McKechnie, A. (2007), ‗Rebuilding a robust Afghan economy‘, In Building a New Afghanistan,
            Robert I. Rotberg, eds., Brookings Institution Press and World Peace Foundation
            Mills, R. and Qimiao, F., (2006), ‗The Investment Climate in Post-conflict situations‘, The
            World Bank Institute
            MRRD and CSO (2007), ‗The National Risk and Vulnerability Assessment 2005‘
            Nixon, H, (2007), ‗Aiding the State? International Assistance and the Statebuilding Paradox in
            Afghanistan,‘ AREU Briefing Paper
            Opel, Aftab (2005), ‗Bound for the City: A study of rural to urban Labour migration in

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            Afghanistan‘. AREU, Working Paper
            Paterson, Anna (2005), ‗Understanding Markets in Afghanistan: A Study of the Market for
            Petroleum Fuels. Kabul: Afghanistan Research and Evaluation Unit.
            UNDP (2007), ‗Afghanistan Human Development Report, 2007‘
            UNODC (2007), ‗Afghanistan Opium Survey‘
            Rodrik, Dani (2004), ‗Getting Institutions Right.‘ A Key note address delivered at the PREM
            Week,‘ The World Bank: Washington, DC.
            Rubin, B. (2006), ‗Peace Building and State-Building in Afghanistan: constructing sovereignty
            for whose security?‘, Third World Quarterly, 27-1, 175-185
            Rubin, B. (2006a), ‗Afghanistan‘s Uncertain Transition from Turmoil to Normalcy‘, Counci l on
            Foreign Relations
            Rubin, B. (2007), ‗Saving Afghanistan‘, Foreign Affairs‘
            Rubin, B. and Armstrong, A. (2003), ‗Regional Issues in the Reconstruction of Afghanistan‘,
            World Policy Journal, Spring
            Shah, B. (2003), ‗Afghanistan‘s reconstruction: issues and challenges, Institute for Strategic
            Studies, Islamabad
            Shaw, A. (2006), ‗Drug Trafficking and the development of organized crime in Post-Taliban
            Shobair, S., and Alim, A., The effects of calamities on water resources and consumption in
            Afghanistan. FAO, WHO
            Starr, F. (2007), ‗Regional Development in Greater Asia: The Afghan Pivot‘, In Building a New
            Afghanistan, Robert I. Rotberg, eds., Brookings Institution Press and World Peace Foundation
            Suhrke, A. (2006), ‗The limits of State building: the role of International Assistance in
            Afghanistan‘, The Chr. Michelsen Institute
            Suhrke, A. (2007), ‗Reconstruction as modernisation: the ‗post-conflict‘ project in
            Afghanistan‘, Third World Quarterly, 28:7, 1291-1308
            World Bank (2003), ‗Trade and Regional Cooperation between Afghanistan and its
            World Bank (2004a), ‗Reforming Fiscal and Economic Management in Afghanistan‘
            World Bank (2004b), ‗Transitional Islamic State of Afghanistan, Mining as a Source of
            World Bank (2005a), ‗Kabul Urban Policy Notes Series‘ n.2
            World Bank (2005b), ‗Managing Public Finance for Development‘
            World Bank (2005c), ‗Afghanistan State Building, Sustaining Growth, and Reducing Poverty A
            Country Economic Report‘
            World Bank (2005d), ‗Afghanistan Investment Climate assessment, Exploiting Opportunities
            in an Uncertain Environment‘
            World Bank (2006), ‗Economic Cooperation in the Wider Central Asia Region‘
            World Bank (2007), ‗Islamic republic of Afghanistan‘, Interim Strategy for the period FY07 -
            World Bank (2008a), ‗World Development Indicators‘
            World Bank (2008b), ‗Doing Business 2008 Afghanistan‘

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

            World Trade Organisation (2008), ‗Trade Policy Review: Pakistan‘

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Annex1            Table of Key Indicators (Time Event since Peace Agreement)
            Variable            t=0      0       1        2       3       4       5         6       7

            GDP growth
              Afghanistan      2001      n/a     28.6     15.1    9.4     16.4    6.1       13.5    9.0
              East Timor                 1999    n/a      15.4    16.5    (6.7)   (6.2)     0.3     2.3
              El Salvador      1992      7.5     7.4      6.1     6.4     1.7     4.2       3.7     3.4
              Ethiopia                   1991    (8.9)    13.3    3.5     6.1     13.2      3.5     (4.0)
              Nicaragua                  1990    (0.1)    (0.2)   0.4     (0.4)   5.0       5.9     6.3
              4.0     Sierra Leone       2002             27.4    9.5     7.4     7.2       6.8     6.5
            GDP per capita
              Afghanistan      2001      122     182      199     252     300     335       383     433
              East Timor                 1999    357      403     468     414     368       355     345

               El Salvador     1992     1.755    1.844    1.914   1.993   1.985   2.026     2.059   2.088
               Ethiopia                 1991     108      93      108     108     111       121     122
               129     Nicaragua                 1990     712     694     679     659       665     688
               717     730     Sierra Leone      2002     195     204     210     218
            Aid per GNI
               Afghanistan     2001     16.4     32.3     34.7    36.7    37.8
               East Timor               1999     n/a      71.2    51.1    60.3    41.8      31.8    33.5
               El Salvador     1992     6.9      5.9      3.8     3.1     2.9     2.5       1.5     1.5
               Ethiopia                 1991     8.2      8.3     12.5    15.6    11.6      9.6     6.9
               8.6     Nicaragua                 1990     33.4    72.1    50.4    23.8      23.7    23.0
               31.1    13.2
            Gross Domestic Investment
               Afghanistan     2001     n/a      n/a      27.6    31.7    35      37.1      31.4    29.6
               East Timor               1999     n/a      34      32      28      23        19      23
               El Salvador     1992     n/a      n/a      4       5.3     3.9     3.6       4.1     2.9
               Ethiopia                 1991     n/a      n/a     n/a     n/a     n/a       n/a     8.3
               7.4     Nicaragua                 1990     n/a     n/a     n/a     n/a       n/a     n/a
               n/a     13.7    Sierra Leone      2002     4.4     4.6     4.6     5.8
               Afghanistan     2001     n/a      n/a      1.2     1.3     9.4     9.6       8.9     9.0
               East Timor               1999     n/a      7       9       7       9         9       9
               El Salvador     1992     n/a      n/a      15.8    15      12.3    12.6      13.2    13.5
               Ethiopia                 1991     n/a      n/a     n/a     n/a     n/a       n/a     8.7
               9.8     Nicaragua                 1990     n/a     n/a     n/a     n/a       n/a     n/a
               n/a     16.8    Sierra Leone      2002     5.7     9       6.1     11.5
            Gross National Savings
               Afghanistan     2001     n/a      n/a      (5.3)   (5.6)   (5.4)   (3.6)     (4.2)   (4.1)
               East Timor               1999     n/a      5       6       6       10        42      53
               El Salvador     1992     n/a      n/a      2       3.2     1.8     1.6       1.2     0.2
               Ethiopia                 1991     n/a      n/a     n/a     n/a     n/a       n/a     6.8
               3.3     Nicaragua                 1990     n/a     n/a     n/a     n/a       n/a     n/a
               n/a     4       Sierra Leone      2002     (4.3)   (3.2)   (1.2)   (1.6)
               Afghanistan     2001     n/a      n/a      30.4    28.3    46.6    45.4      43.9    43.0
               East Timor               1999     n/a      (50)    (51)    (54)    (44)      (36)    (26)

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

                El Salvador      1992    n/a     n/a     14.1    12.3    12.2     13.5      13.9   14.3
                Ethiopia                 1991    n/a     n/a     n/a     n/a      n/a       n/a    1.2
                4.4      Nicaragua               1990    n/a     n/a     n/a      n/a       n/a    n/a
                n/a      (0.7)
                Sierra Leone     2002    (5.1)   (4.2)   (0.5)   1

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Annex 2            Classification of Occupation by Respondents

             1.   Crop production for home consumption                                      agriculture
             2.   Livestock product production for home consumption                         livestock
             3.   Prod & sales of field crops                                               agriculture
             4.   Prod & sales of cash crops (Non Opium)                                    agriculture
             5.   Prod & sales of Opium                                                     opium
             6.   Prod & sales of orchard products                                          agriculture
             7.   Prod & sales of livestock & products                                      livestock
             8.   Sales of prepared foods                                                   services
             9.   Agricultural wage labour (Non Opium)                                      agriculture
             10. Opium wage labour                                                          opium
             11. Shepherding                                                                livestock
             12. Mills                                                                      others
             13. Other wage labour                                                          others
             14. Skilled labour                                                             services
             15. Salary/Government job                                                      services
             16. Small business                                                             services
             17. Petty trade                                                                services
             18. Cross border trade                                                         services
             19. Firewood /charcoal sales                                                   agriculture
             20. Handicrafts                                                                manufacture
             21. Carpet weaving                                                             manufacture
             22. Mining                                                                     others
             23. Military service                                                           others
             24. Taxi/transport                                                             services
             25. Remittances for seasonal migrants                                          remittances
             26. Remittances from family members permanently away.                          remittances
             27. Pension                                                                    others
             28. Other Govt. benefits                                                       others
             29. Rental income                                                              others
             30. Sale of food aid                                                           others
             31. Begging/borrowing                                                          others
             32. Other                                                                      others
            Source: Authors calculation using NRVA 2005

DFID Understanding Afghanistan - Growth Diagnostic - 3.4.1 Draft Final Report - Technical

      Annex 3     Methodological Note for Returns to Education (Mincer
            We use NRVA 2005 to estimate returns to education in Afghanistan. In order to reduce
            measurement errors, we limited our sample to urban male population, and we considered only
            head of the household and income reported for the first source of income.
            NRVA does not report per capita wages, but household income from the source and number of
            members engaged in the occupation. We approximated wages by considering household
            monthly wage as reported from the first source of income and divided it by the number of
            people involved in the respective activity. As we restricted the sample to only male and urban
            population, very few cases where this happened, and it never included more than 3 household
            male members engaged in the activity. We took age and highest level of education from NRVA.
            Experience is calculated as the age minus years of education-6.

                   Mincer returns to education in Afghanistan (primary omitted)
                   Experience          0.00901
                   Experience^2        -6.3E-05
                   Secondary           0.169***
                   High School         0.218***
                   Tertiary            0.201***
                   Constant            8.328***
                   Observations        2.757
                   R-squared           0.016
                   Robust standard errors in parentheses
                   *** p<0.01, ** p<0.05, * p<0.1
            Source: Authors calculation using NRVA 2005