Docstoc

rp

Document Sample
rp Powered By Docstoc
					        Corruption Perception Index and Cell Phone Subscription Rates

                                 Andy Schaalman1
                                 SIS, American University
                                 4400 Massachusetts Avenue NW, Washington DC 20016
                                 as0988a@student.american.edu

    Abstract

Since the end of the Cold War – and the subsequent disintegration of the proxy mentality it produced – international
attention has increasingly been paid to corruption within states, both in terms of individual actions by leaders and
more widespread institutionalized patterns of behavior. Consequently, a number of organizations have introduced
new metrics and indexes in attempts to quantify corruption’s presence and provide policymakers with the necessary
insights to combat the problem in the hopes of creating more transparency. The most prominent of these
organizations has been has been Transparency International (TI); its most important and utilized metric measure has
been the Corruption Perceptions Index (CPI).

In today’s twenty-first century world, mobile communication technology serves a number of purposes. Besides
spanning great distances to connect people and businesses on a “micro” level, it also serves as a “macro” catalyst of
sorts, insofar as it provides a degree of social empowerment to people through its utility as an instrument of social
cohesion.

The focus of this paper – and the accompanying quantitative analysis – will be on the relationship between countries’
CPI and their rate of cell phone subscribers. The great issue which lies at the heart of the comparison is the impact the
explosion of social media and the spread of information technology is having on the behavior of governments and
leaders vis-à-vis the people they serve. Many recent events on the world stage – especially during this year’s Arab
Spring – have hearkened to changing dynamics in the balance of power between people and their leaders thanks to
the penetration of social media through information technology, albeit in the face of a host of evidence that
government behavior has not undergone a transformation. Despite –or perhaps because of this – US and international
policymakers have increasingly devoted attention and resources to implement increased “connectivity” and “access”
to information technology in the countries of the developing world in the hopes of altering such behavior. My goal
over the course of this paper will be to assess the feasibility of the notion of producing more transparent and
accountable leadership through the penetration of one courier of information technology – the cell phone. If and
where patterns emerge, I will examine the possible reasons why; I will do the same in the absence of such patterns as
well.

The data for CPI – the dependent variable – is taken directly from Transparency International; the data for cell phone
subscribers per 100 people – the independent variable – is taken from the World Bank’s World Development
Indicators. As control variables, I will use a variety of metrics which will account for the different forces that may
influence – or be intrinsically connected to – the relationship between the variables. Three of these metrics – Internet
subscribers per 100 people, literacy rate and GDP per capita – are drawn from the WDI; the final metric, level of
political freedom, is drawn from Freedom House.




    1
        I am grateful to Zhouzhe Chen for discussions and to Angel and Prof. Assenov for assistance in getting the data for this research



                                                                                                                                            1
1. Introduction



In this section:


    a. State the problem or objective of your research paper.
                    a. What exactly do you want to study?
       I want to study the relationship between the omnipotence of cell phones as a courier
       for social media and the perception of corruption / transparency in governments
       around the world.
                    b.Why is it worth studying?
    b. In the past several years, policymakers have sought to bring greater social
       empowerment to the “corrupt” countries of the world by proposing and funding
       projects to bring the infrastructure needed for mobile broadband and Internet access.
       The backbone idea behind these directives is simple: that increased accessibility of
       information and social connectivity will bring about greater transparency and
       cooperation between these countries’ public leaders and institutions and the people
       they serve.
                    a. Does the proposed study have practical significance?
       Although new social media outlets such as Facebook and Twitter have rightfully
       received the brunt of the credit for sparking the extraordinary run of organized protests
       seen around the developing world in recent months and years, these outlets would not
       have been available to spur those demonstrations without the technology to carry those
       outlets. Hence, the cell phone has played a role in changing the power dynamic
       between people and government.

        Of course, this does not necessarily mean that government has changed or societies
        have changed – just that peoples’ opportunities to engage their governments have
        changed. Nonetheless, policymakers have been intent on pursuing programs to
        develop the infrastructure and markets for information technology in developing
        countries in the hope that the governments in those countries will demonstrate more
        responsible leadership.

        In this study I will examine how the growing role of one device – the cell phone – is
        associated with levels of CPI, an interval-ratio measure of state corruption /
        transparency. The answer and its nuances will shed a great deal of light on the
        feasibility of the policy community’s current designs on combating corrupt
        governments through information technology and social media investment.

    c. Formulate a research question that:

                                                                                              2
                a. Is a clearly stated expression of interest and intent, and that can be answered
        Given the evidence provided by the application of the control variables
    (Internet subscribers per capita, GDP per capita, political freedom) for the
    primary data set and the pattern of the relationship between the primary
    variables over time, can we ascertain that as rates of cell phone subscription rise,
    the perception of corruption will decrease?
                b.Implies a relationship between variables. Think about the quantitative
                    variables whose behavior you are trying to explain, i.e. dependent variable,
                    the variable that is measured to see whether the treatment or manipulation of
                    the independent variable had an effect. Find a variable/s that is/are
                    manipulated by the policy makers to examine its impact on a dependent
                    variable
d. Briefly discuss the structure of your paper, i.e. what follows in section two, section three, etc.


        In my literature review, I will first discuss the dependent variable – CPI – followed

    by corruption more generally, insofar as to give a sense of context to what the gives

    the scores for the variable meaning. In this section, I cite the work of Michael W.

    Collier. Next, I delve into a discussion of the significance of the relationship between

    social media and the political arena, as described at length by NYU Prof. Clay Shirky

    in Foreign Affairs Magazine. Finally, I will cite and analyze the findings of a study

    conducted by Jacob Groshek on the relationship between democracy and the Internet

    to provide some additional insight into the relationship between governance and new

    media.


        In the sections following my literature review, I will present the tables and graphs

    of data outputs which I used to measure the relationship between CPI and cell phone

    usage. Contained in these sections are: histograms of each of the variables, relevant

    descriptive statistics, single linear regression of the primary variables, bivariate, pair-

    wise, and multiple regression analyses, scatterplot for the primary variables, and log

    residuals plot for the control variables.


                                                                                                   3
     2. Literature review
In this section do critical review of the quant articles/papers that are related to your research
topic/question/.

          Over the last two decades, corruption – defined by the World Bank as the “abuse of public

power for private gain” – has been a primary focus of policymakers on the international arena. In

turn, many NGOs, international lobbyists and research organizations have sought to produce

quantifiable metrics to measure the presence and pervasiveness of irresponsible leadership in the

effort to stem the tide in favor of greater transparency and accountability. One of the

organizations at the forefront of the drive to produce such metrics has been Transparency

International (TI).

          Since its inception in the mid-1990s, TI has been producing a statistic called the

Corruption Perception Index (CPI). The CPI metric works as follows: TI ranks countries on a

scale of 1 to 10, with widespread corruption at 1 and corruption-free at 10. As is to be expected

from data collected in such a difficult context and on such a tricky subject, TI is mum on the

specifics they use to construct their ratings, saying only that CPI “draws from a wide array of

surveys and assessments”, including the likes of Gallup International, PriceWaterhouseCoopers,

and Freedom House. While it has generally been praised as a strong metric measure, CPI – by

virtue of its basis in perception and its reliance on multiple sources – is by no means perfect. For

instance, all the countries in a given year’s index are not rated by the same sources, and some

organizations’ figures are given higher weights in the rankings.2 Expert assessments carry the

most weight, often at the expense of households and firms.3 Still, global policymakers hold

sufficient faith in CPI to use it as a policy tool on a consistent basis, so any additional insight into



     2
      Theresa Thompson and Anwar Shah, Transparency International’s Corruption Perceptions Index: Whose Perceptions are they Anyway?
Worldbank.org http://siteresources.worldbank.org/PSGLP/Resources/ShahThompsonTransparencyinternationalCPI.pdf
    3
      Thompson and Shah, 12
                                                                                                                                       4
the ratings and their potential significance for global trends and their policy implications has great

potential significance.

            In order to appreciate what TI has attempted to quantify, we must first appreciate what

corruption is – and how it germinates – by looking beyond its textbook definition. In a paper

written to the annual convention of the International Studies Association in 1999, Florida

International professor Michael W. Collier posits that corruption emanates, in the most basic

sense, from a combination of: (A) social and cultural norms of individual choice, and (B) the

opportunities and ability ruling elites possess to maximize or exploit their share of “social

surplus” according to these social and cultural customs. He explains:

Social rules link people (agents) and society. Rules tell people what they should do, what they must do, and what they
have a right to do. When agents fail to follow rules, other supporting rules bring consequences. In light of their
material circumstances, agents follow or disregard rules in order to achieve their goals. Institutions are simply
patterns of stable rules. Structure is a stable pattern of rules, institutions, and their unintended consequences. 4

            The norms Collier speaks of, and the structure produced from them, create what Johnston

calls the “system of public order” – defined as a state’s “relationships between power and private

interests” and their resulting implications for state resource allocation, power elite accountability,

and political participation of the masses.5 According to Johnston, the interweaving of these forces

produces in each state a particular level of social empowerment in the general populace. Although

social empowerment cannot be created by means of mass communication and social media alone,

they can – and by virtue of their growing presence and importance, do – confer upon capable

people the opportunity to keep corruption in check by maximizing their share of “social surplus”,

argues Clay Shirky of Foreign Affairs Magazine:

            As the communications landscape gets denser, more complex, and more participatory, the networked
            population (gains) greater access to information, more opportunities to engage in public speech, and an
            enhanced ability to undertake collective action. In the political arena…these increased freedoms can help
            loosely coordinated publics demand change…



    4
        Michael W. Collier, Examining Political Corruption. Columbia International Affairs Online http://www.ciaonet.org/isa/com01/
    5
        Michael W. Collier, http://www.ciaonet.org/isa/com01/
                                                                                                                                      5
           The use of social media tools…does not have a single preordained outcome… (because) political freedom
           has to be accompanied by a civil society literate enough and densely connected enough to discuss the issues
           presented to the public…Access to information is far less important, politically, than access to conversation. 6

           Although the Internet remains the backbone of the information and communications

“landscape” that Shirky describes and the personal computer remains the most important courier,

the cellular phone holds an increasingly significant place in the realms of social engagement and

public connectivity because of its portability, its range of capabilities, and its relatively low cost.

Over the last handful of years, the cell phone has exploded in numbers across the developing

world, to the point that in 2008, 50 percent of the world had at least 75 cell phone subscriptions

per 100 people.7

           One country that recently stepped over that threshold of cellular connectivity,

coincidentally, is Egypt. Although studies conducted by the Egyptian government itself and

private research analysts differ in their numbers, at least 80% of the Egyptian population (of about

80.5 million) now subscribes to a cellular service provider, a figure which eclipses the number of

Internet subscribers in the country by about four to one.

           By now, the story seems cliché, but for this context it bears repeating: the protests in

Egypt against the Mubarak government – and the buildup of critical mass which subsequently led

to his stepping down – were organized and sustained primarily through the use of social media;

applications (voice, e-mail, text, camera, video, and social networking) of these media were

utilized by scores of protesters wielding cell phones. Although the Egyptian government – which

has majority control over Telecom Egypt, the national fixed line provider – enacted a shutdown of

all the country’s Internet and cellular services in the wake of January’s protests, the move proved



    6
      Clay Shirky, The Political Power of Social Media: Technology, the Public Sphere, and Political Change. Foreign Affairs Magazine, Jan/Feb
2011 http://www.foreignaffairs.com/articles/67038/clay-shirky/the-political-power-of-social-media
    7
      World Development Indicators, 2008, World Bank Group



                                                                                                                                                 6
feeble in its attempt to defuse the protesters. The reason for the move’s lack of success seems

self-evident: the critical mass of popular sentiment and common cause engendered by the calls to

action prior to the shutdown was sufficient to sustain an emotional fervor and a spirit of

cooperation that could endure through the shutdown itself. Clearly, this would not have been

possible without a sufficiently literate, engaged, and empowered population (the “access to

conversation” angle) – and, it stands to reason, the penetration of mass communication to foster

those dynamics.

            Probing deeper into the issue, Clay Shirky says the following:

           This condition of shared awareness – which is increasingly evident in all modern states – creates what is
           commonly called…”the conservative dilemma”, so named because it applies not only to autocrats but also to
           democratic governments…with the spread of (new) media…a state accustomed to having a monopoly on
           public speech finds itself called to account for anomalies between its view of events and the public’s. The
           two responses to the conservative dilemma are censorship and propaganda. But neither of these is as
           effective a source of control as the enforced silence of the citizens…

           But (when) a government… (shuts) down information access (in this manner), it (risks) radicalizing
           otherwise pro-regime citizens or harming the economy…tools in broad use (are) much harder to censor
           without risking politicizing the larger group of otherwise apolitical actors.8

           To be sure – as the events of last winter attest – Egypt’s growing number of cellular

subscribers had no great bearing on the inclination of the country’s leaders to serve their own

purposes and exploit their “social surplus” potential; otherwise the government would not have

felt compelled to shut down phone and Internet access in the wake of the protests. In addition,

CPI is an aggregate measure of corrupt acts and norms of behavior, so any reductions in a

country’s CPI score cannot be attributed to any benevolent “counterweight” actions undertaken

by already-corrupt public officials.

           That said, it stands to reason that the enormous growth of information infrastructure and

social media worldwide – and its accompanying consequences for public social empowerment –



   8
       Shirky, Foreign Affairs



                                                                                                                     7
has the potential, if properly utilized, to alter the behavior of public leaders and institutions

towards the citizens they serve, even if on a very microcosmic level. As a rapidly-growing

instrument of the larger information infrastructure and a primary courier of social media, a

measure change in the corruption paradigm – however small – could thus be attributed in some

form to the cell phone’s singular presence. An example demonstrating the potential utility of cell

phones as a means of reducing corruption is described by Dan Rice and Guy Filippelli in Small

Wars Journal:

When Americans first entered Afghanistan in 2001 there was little infrastructure and no banking system in an entirely
cash economy. Nine years later it is still a cash economy and 97% of the country remains “unbanked”, but
Afghanistan’s thriving telecom industry offers a way to minimize graft. From a standing start, Afghanistan now
boasts a cellular network of 12 million cell phones in country of 28 million. Mobile technology is the largest legal,
taxpaying industry in Afghanistan and the single greatest economic success story in the country since the fall of the
Taliban… In 2009, the Afghan National Police began a test to pay salaries through mobile telephones, rather than in
cash. It immediately found that at least 10% of its payments had been going to ghost policemen who didn’t exist;
middlemen in the police hierarchy were pocketing the difference…Mobile currency in Afghanistan has already
demonstrated the capacity to support salary payroll, limited merchant payments, peer-to-peer transfers, loan
disbursements and payments. 9

     However enticing this may seem on the surface, policymakers –Shirky cautions – must be

careful not to be lured into interpreting this and other examples of transformed behavior through

new technology as a reason to operate from a top-down approach on the issue of promoting

“connectivity” and “access” around the world. In a great number of incidents, especially

powerful authoritarian governments have been able to leverage their total control over a society’s

communication infrastructure to effectively defuse protest activity through social media and

information technology shortly after it commenced. The central underpinning of this reality is

that the widespread penetration of new technology and its use by the people is insufficient to

produce change: a literate, functioning, and reasonably open civil society must be in place for any




      9
        Dan Rice and Guy Filippelli, One Cell Phone at a Time: Countering Corruption in Afghanistan. Small Wars Journal, 2010
http://smallwarsjournal.com/blog/journal/docs-temp/527-rice.pdf
                                                                                                                                8
form of new media to have significant consequences for a transformation in the behavior of

leaders.

      In his piece entitled Multinational Forecasts of Democratic Forecasts and Internet Diffusion,

Jacob Groshek echoes many of Clay Shirky’s sentiments, cautioning that while technology

provides opportunities, history has shown that it is “only as deterministic as the people who create

and subsequently use (it)” and that whatever change it does bring about is invariably slow. In a

study which has relevant implications for our examination of cell phones and corruption, Groshek

conducted a time-series, cross-national study of the effects of Internet penetration on democracy

for the years 1994 to 2003.10 Despite popular expectations to the contrary, the results of his study

showed no significant relationship between the penetration of the Internet and the advancements

of democratic ideals, and in the four countries where an association seemed to be present there

were additional variables that skewed the data. The principal implication of his study for my

analysis is the following: even if information technology and so-called “new media” contribute to

the growth of transparent governance, there is little or no causation involved and there are a

number of more important variables involved which contribute to any level of association.

     To be sure, regardless of whatever quantitative associations may be produced from my data,

proving associations and patterns between rate of cell phone use and perceived government

behavior will be difficult. Still, it stands to reason that the prevalence of mobile communications

technology has had – and is having, as evidenced by the Arab Spring – an effect on the

relationship between people and governments around the world, even if in a very temporary and

microcosmic way. If the information I present gives a sense of scale to the relationship between




      10
         Jacob Groshek, A Time-Series, Multinational Analysis of Democratic Forecasts and Internet Diffusion.
http://ijoc.org/ojs/index.php/ijoc/article/view/495/392


                                                                                                                9
the two variables and the consequences of that relationship for policymakers going forward, then I

will have produced a successful analysis.


    3. The Model
        Describe the model that you are going to use and the hypotheses that you are formally testing.


        Example of non-generic model


        CPIl= 0 + 1MSper100 +2ISper100+ 3 PF +4 LitRate + 6 GDPpc + ut


where, the dependent variable is CPI, and the independent variables are MSper100, mobile subscriptions

per 100 people, ISper100, Internet subscriptions per 100 people, PF, political freedom, LitRate , adult

literacy rate, and GDPpc, GDP per capita.


First hypothesis: H1: Corruption Perception Index ranking and mobile subscriptions per 100 people are

positively correlated, i.e. corruption level and rate of cell phone subscriptions are negatively correlated.

We expect that 1 is statistically significant and positive.


Second hypothesis: H2: Corruption Perception Index ranking and Internet subscribers per 100 people, are
positively correlated, i.e. corruption level and rate of Internet subscribers are negatively correlated. We
expect that 2 is statistically significant and positive.
Third hypothesis: H3: Corruption Perception Index ranking and GDP per are positively correlated, i.e.
corruption level and GDP per capita are negatively correlated. We expect that 3 is statistically significant
and negative.

Overall hypothesis: Overall statement about the research hypotheses consistent with the above model:
Controlling for Internet subscribers per 100 people and GDP per capita, we expect that CPI is positively
correlated with the mobile subscriptions per 100 of a population; thus, corruption and mobile subscriptions
are negatively correlated.

    4. Data

In this section describe:
What is the source of your data?           World Development Indicators (World Bank), Transparency
International, Freedom House


                                                                                                           10
What is the level of measurement of your dependent variable and your independent variables? Interval-
Ratio
What is the data availability? In the case of cross country study - Do you have the data for all of the
countries? In the case of survey research data - Do you have nationally representative data?

        The cross country study contains observations from between 150 and 175 countries for each
variable with the exception of literacy rate; because Pearson’s r projections for eliminating error in
predicting CPI are statistically insignificant, I did not use this variable in my multiple regression models.

Overall statement: Is the data reliable so that you can make credible inferences from you statistical
findings? Yes




    5. Estimates and Empirical Findings

For Interval-Ratio (I-R) dependent in this section describe:
Do the descriptive statistical analysis for your dependent and independent variables
                    a. Run descriptive statistics for each of the dependent and independent variables
                                i. What is the central tendency of dependent and independent variables?
                               ii. Plot the histogram for each of the Interval-Ratio (I-R) variables and
                                    overlay the normal curve
                    b. Run bivariate correlation tables and plots for all of the variables
                             i. Run matrix scatterplots - scatterplots for all of the combinations of your
                                  dependent variable and X I-R variables and examine the linearity of the
                                  relationship between the variables. Look for linear and non-linear
                                  patterns/relationships
                            ii. Run bivariate correlation tables
                    c. Run multiple regressions, check the regression statistics
                    a. Create the tables with the relevant statistics in your research paper
                               1. Descriptive statistics
                               2. Correlation statistics table
                               3. Table with the estimates from the regression models

Make statements about relationship and regression analysis


        The basic linear regression of the primary variables, relevant plotted charts, and the coefficients

provided by two of three multiple regression models show a moderate level of association between a

country’s Corruption Perception Index score and its number of cell phones per 100 people, and

controlling for the number of Internet subscribers per 100 people, the level of association jumps to highly

significant. Thus, we can conclude that there is some form of a relationship between the two, although –
                                                                                                                11
accounting for the writings of Shirky and Groshek, some data points on the linear regression scattergram

and the inconsistency in the coefficients produced by the multiple regression models – the origins and

the nature of that relationship must be called into question.

    When compared to the basic linear regression for the primary control variable (Internet subscribers

per 100 people), we see that the coefficients and adjusted correlation data for the dependent variable

are far stronger with respect to Internet access than to cell phone usage – the implication being that cell

phone penetration is to a large degree predicated upon the connectivity of countries to the worldwide

web and its adjoining information infrastructure. (It should also be noted that the level of per capita

subscription to cellular providers is presently about 3.5 times that of the equivalent statistical measure

for Internet providers, as shown in the histograms. This is significant because – other things being equal –

low penetration of the Internet to ordinary citizens enables governments to exercise more control over

the larger communication infrastructure and its messages, even where cell phone penetration is high.)

    The relevance of cell phone penetration is further complicated by its diminishing coefficients in

Models 3 and 5; when combined with political freedom and GDP per capita in the latter, its coefficient is

statistically insignificant at the level of measurement (alpha=0.05). Further complicating matters, the

coefficients for Internet subscribers decrease significantly when paired with the other crucial control

variables (political freedom in Model 4, political freedom and GDP per capita in Model 6), indicating the

Internet’s relationship to CPI which is largely predicated upon development in other areas. Most

conspicuously, however, the multiple correlations show a remarkably consistent relationship between

CPI and political freedom at the coefficient level – which should not be considered a surprise given what

has been cautioned about the tenuous relationship between technology and governance. These

cumulative trends were concerning enough that I ran a log analysis on Model 6, but the plotted points

produced by the log were inconclusive to show substantial errors in the data.

    These trends suggest – as Shirky and Groshek argue – that the diffusion of technology alone cannot

bring more transparent government; the proper elements of civil society must be in place first.
                                                                                                             12
    6. Conclusion and policy options



Summarize your findings:

        Based on the Pearson’s r and regression analysis I accepted the research hypothesis that

Corruption Perception Index and cell phones per 100 people are positively correlated, i.e. that

penetration of the technology and transparency are moderately associated. However, there are nuances

to the data and caveats to this conclusion which must be taken into account by policymakers.



                     i. Draw the policy recommendations from you findings:


    The connection between rate of cell phone use and perceptions of government behavior –

individually and as one component of the information technology dynamic – is apparent. Where

information mediums and technology are in abundance and civil society is ingrained, transparency is

high; where neither condition is present, transparency is low. However, as the scatterplot of the basic

linear regression and the multiple regression models show, the level of association between the variables

dips considerably after the points below the variables’ means on the scatterplot, indicating a spread of

the cell phone which has far exceeded its ability to contribute to institutional and behavioral change

where civil society is not as developed. This means that the spread of the device has been so swift –

especially relative to population Internet access – that it has created a large bloc of countries where a

discernible gap exists between the “mobile connectivity” of people and the balance of “social surplus”

between the people and the government.

    The recent protests in Egypt, Iran, China, Myanmar, and elsewhere speak to this gap: the

governments there were not less corrupt than before – the people just had the opportunity and the


                                                                                                            13
means to coordinate together to air grievances in large numbers (and to varying degrees of success). In

these countries and others like them, the development of civil society sufficient to keep corruption (and

the perception of such behavior) in check has been outpaced by the march of technology.

    With this in mind, US policymakers should heed the advice of Shirky and Gorshek and focus on

fostering growth of civil society from the ground up in the developing world, instead of spending large

sums for “access and connectivity” in efforts to spring civil society and responsible governance up

overnight. The United States and its international partners should remain committed to promoting

information freedom and the spread of new technology in all its forms, but must adjust to each individual

case and focus on specific commitments – such as resolving or mitigating one element of information

repression in Iran and slowly expanding and testing the mobile banking system underway in Afghanistan

– rather than making holistic commitments and investments to countries that are “undeveloped” or

“censored”. Only if this is done will the spread of information technology and social media throughout

the world hold true meaning for the future of governance over the long term.




                                                                                                          14
          Appendix. Histograms
                                                                                                                                                      Mobile Phone Subscriptions per 100 People, 2008
                                                     Corruption Perceptions Index, 2008




                                                                                                                                .01
                                 .3




                                                                                                                             .008
                                                                                                                             .006
                                 .2




                                                                                                              Density
             Density




                                                                                                                             .004
                                 .1




                                                                                                                             .002

                                                                                                                                              0
                                  0




                                           0           2                 4                 6         8          10                                      0                  50              100                    150        200
                                                                                 cpi2008                                                                                           Mobile_100people_08
                             Source: 2008 CPI Data, Transparency International                                                       Source: 2008 WDI Data, World Bank




                                                                                                                                                                               Level of Political Freedom, 2008
                                                   Internet Subscribers per 100 People, 2008




                                                                                                                                                      .04
                           .05
                           .04




                                                                                                                                                      .03
                           .03




                                                                                                                                    Density
          Density




                                                                                                                                                      .02
                           .02




                                                                                                                                                      .01
                           .01




                                                                                                                                                                0
                                  0




                                               0            20               40           60             80                    100                                    20                   40               60          80         100
                                                                          Internet_per_100_08                                                                                                       FHstand2008
                            Source: 2008 WDI Data, World Bank                                                                                          Source: 2008 WDI Data, World Bank




                                                                                                                                                                                 GDP per Capita, 2008
                                                    Literacy Rate, Age 15+, 2004
              .05




                                                                                                                                              .3
             .04




                                                                                                                                              .2
             .03




                                                                                                                   Density
Density


             .02




                                                                                                                                              .1
             .01




                                                                                                                                                  0
                       0




                                      20                   40                60                 80                100                                       0              2                    4             6          8          10
                                                                        Read2004                                                                                                                    cpi2008
                    Source: 2008, Freedom House                                                                                               Source: 2008 WDI Data, World Bank




                                                                                                                                                                                                                                         15
Tables

Table 1. Descriptive Statistics

          Variable             Mean          Standard Deviation       Coefficient of       Observations
                                                                        Variation
    CPI                            3.98                2.12                 0.533           170


    Cell Phones per 100            74.58              45.24                 0.607           169


    Internet per 100               26.89              26.29                 0.978           169


    Political Freedom              65.25              27.26                 0.418           168


    Literacy Rate                  80.05              20.15                 0.252           118


    GDP per capita                $13,273            $20,026                1.509           168


Data sources: Transparency International (2008), World Bank’s WDI (2008), and Freedom House (2004)


Table 2. Bivariate Analysis


 Independent Variable       Observations        Pearson’s r;          Interpretation         Direction
                                             adjusted r-squared
    Cell Phones per 100            169q          0.6169; 0.377                              Positive
                                                                        Reject Null;
                                                                        Moderate
                                                                       Association
    Internet per 100               26.89          0.8514; 0.723         Reject Null;        Positive
                                                                          High
                                                                       Association
    Political Freedom              65.25          0.6567; 0.431         Reject Null;        Positive
                                                                        Moderate
                                                                       Association
    Literacy Rate                  80.05           0.3216; .103            Fail to          Positive
                                                                       Reject Null;
                                                                          Weak
                                                                       Association
    GDP per capita                $13,273         0.7885; 0.622         Reject Null;        Positive
                                                                          High
                                                                       Association
Data sources: Transparency International (2008), World Bank’s WDI (2008), and Freedom House (2004)
                                                                                                          16
        Table 3. Pair-Wise Correlation Table


   Variable             CPI         Cell Phones       Internet          Political     Literacy Rate   GDP per
                                      per 100        Subscribers        Freedom                        Capita
                                                       per 100
        CPI                1


                          170


        Cell          0.6169              1
                      (0.000)
 Phones per             169              169

  100 People


     Internet         0.8514            0.747
                      (0.000)          (0.045)             1
 Subscribers            169              168

   per 100                                               169


     Political        0.6567            0.460                               1
                      (0.000)          (0.000)
   Freedom              168              167
                                                        0.600             168
                                                       (0.000)
                                                         167
     Literacy         0.3216           0.5940                            0.2084              1
                     (0.0004)          (0.000)         0.5682           (0.0242)
     Rate               118              167           (0.000)             117             118
                                                         117

     GDP per          0.7885           0.5863                           0.4138            0.3574           1
                      (0.000)          (0.000)                          (0.000)          (0.0001)
    Capita              168              168           0.7766             166               117        168
                                                       (0.000)
                                                         167
Data sources: Transparency International (2008), World Bank’s WDI (2008), and Freedom House (2004)




                                                                                                      17
                                    Table 4. Multiple Regression Models


   Independent
                         Model 1           Model 2            Model 3           Model 4    Model 5   Model 6
     Variables
                           0.029                               0.0181                      0.0460
Mobile Phones per
                          (10.13)                              (6.54)                      (1.92)
  100 People
                           0.001                                0.001
    Internet                                0.0684                               0.0571              0.0334
 Subscribers per                            (20.98)                              (14.79)             (5.52)
   100 People                                0.001                                0.001               0.001
                                                               0.0372            0.0172    0.0286    0.0188
 Level of Political
                                                               (8.20)             (4.61)   (7.83)    (5.02)
    Freedom
                                                                0.001             0.001     0.001     0.001
                                                                                           0. 603    0. 381
 GDP per Capita                                                                            (6.58)    (4.59)
   ($10,000)                                                                                0.001     0.001

Adjusted r-squared         0.377              0.723             0.547              0.751    0.748     0.799


 Observations (N)            169               169                167               167     165       165


              Note: All Statistically Significant Figures Significant at alpha=0.0001




                                                                                                      18
                        Transparency (CPI) and Mobile Phones per 100 People

            10
             8
             6
             4
             2
             0




                       0                          50                 100                        150                    200
                                                              mobile_100people_08

                                                                    Fitted values          cpi2008




                            Scatterplot of CPI vs. Mobile Phones per 100 People
                                cpi2008

                      200

                      100                   mobile_100people_08

                        0
                      100

                      50                                      internet_per_100_08

                        0
                      100

                      50                                                              FHstand2008

                        0
                      100

                      50                                                                              Read2004

                       0
                 100000

                 50000                                                                                                 gdp2008

                       0
                            0      5        100        100   2000       50      100
                                                                                  0       50      0
                                                                                                100       50     100
             Source: Transparency International, The World Bank Group, Freedom House




                                       Predicted Values of cpi2008 vs. Residuals Plot
                   3
                   2
                   1
Residuals




                   0
                 -1
                 -2




                                        2                               4                             6                          8
                                                                            Fitted values




                                                                                                                                     19
References
  •   Theresa Thompson and Anwar Shah, Transparency International’s Corruption Perceptions
      Index: Whose Perceptions are they Anyway? Worldbank.org
      http://siteresources.worldbank.org/PSGLP/Resources/ShahThompsonTransparencyinterna
      tionalCPI.pdf
  •   Michael W. Collier, Examining Political Corruption. Columbia International Affairs
      Online http://www.ciaonet.org/isa/com01/
  •   Clay Shirky, The Political Power of Social Media: Technology, the Public Sphere, and
      Political Change. Foreign Affairs Magazine, Jan/Feb 2011
      http://www.foreignaffairs.com/articles/67038/clay-shirky/the-political-power-of-social-
      media
  •   Dan Rice and Guy Filippelli, “One Cell Phone at a Time: Countering Corruption in
      Afghanistan”. Small Wars Journal, 2010 http://smallwarsjournal.com/blog/journal/docs-
      temp/527-rice.pdf
  •   Jacob Groshek, A Time-Series, Multinational Analysis of Democratic Forecasts and
      Internet Diffusion. http://ijoc.org/ojs/index.php/ijoc/article/view/495/392
  •   Mark Landler and Brian Knowlton, “U.S. Policy to Address Internet Freedom”. New York
      Times, 2/14/11 http://www.nytimes.com/2011/02/15/world/15clinton.html
  •   “Use Mobile Phone Banking to Share Africa Resource Wealth: World Bank”. China Post,
      6/10/11 http://www.chinapost.com.tw/business/africa/2011/06/10/305611/Use-mobile.htm
  •   Transparency International online
  •   World Bank, World Development Indicators online
  •   Freedom House online




                                                                                          20

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:15
posted:8/20/2011
language:English
pages:20