Stream 3 The Supply of and Demand for Income Security

Document Sample
Stream 3 The Supply of and Demand for Income Security Powered By Docstoc
					 Stream 3: The Supply of and Demand for Income Security: Behavioural
                           Policy Modelling

30. The effect of childcare costs on labour supply

The effectiveness of incentives towards participation in the labour market will depend to a
large extent on the costs and the availability of childcare outside the home. The aim of this
project is to incorporate detailed measures of the costs of childcare into the labour supply
models used by the Melbourne Institute Tax and Transfer Simulator (MITTS). This will
greatly improve our understanding of the barriers to working outside the home for both sole
parents and married mothers. The study will provide the first comprehensive research
linking childcare policies, other tax and transfer programs and labour supply in Australia.

Studies of the labour supply effects of changes in taxes and transfers suggest that a major
component of the response in employment takes the form of adjustments in participation
rather than changes in the hours of work for the employed. Understanding the decision to
enter or leave the labour market is crucial to our ability to predict effects of policy changes.
In households with children, an important component of this decision is the cost of
childcare outside the home. In the current labour supply model underlying MITTS, the
measure used in the fixed cost parameter of participating in the labour market is whether the
youngest child is of pre-school age (less than four years old) or of school-going age (5 to 15
years old). This provides a rough approximation only of the variation in childcare costs to
the working mother.
Detailed information on formal and informal childcare expenditures will allow us to impute
more realistic cost functions into the labour supply model. These costs will depend on the
number and age of the children, the presence of a spouse and other adult family members,
the state of residence and the urban/ rural distinction, and eligibility for government
childcare subsidies. The resulting labour supply model will better reflect the participation
decision facing parents.
More detailed childcare cost functions can also be used to model and understand preferences
between part-time and full-time work. Hourly childcare costs are not constant over all levels
of hours and increased hourly costs at higher levels of hours will induce parents to prefer
lower hours of work. It is possible that a substantial component of what is currently
interpreted as a preference for low hours of work by mothers, is in fact a response to
increases in hourly childcare costs for long hours of care.
Recent changes in government policies on family payments for childcare reflect the view that
childcare programs are important and must be integrated with other publicly provided
benefits. By incorporating childcare costs in the MITTS labour supply model, we are
explicitly modelling the interaction of childcare expenditures and policies with other tax and
transfer programs. The resulting labour supply estimates constitute an important step in
understanding the labour supply behaviour of Australian households and will be necessary to
our understanding of the effects of the large recent changes in childcare policies.

The aim of including a more accurate measure of the costs of working when children are
present can be achieved by using a specialised survey conducted by the ABS. (Information
must be obtained from other sources since the Survey of Income and Housing Costs, used
for the labour supply estimation and in MITTS, does not contain information on childcare
expenditure or usage.) The ABS Child Care Survey 1996 contains information on the cost of
childcare for families living in different parts of Australia and with different characteristics. It
contains data on a large representative sample of 11,419 children under 12 in 6,421 income
units in private dwellings. Information is provided on childcare arrangements including
weekly hours, cost and number of weekdays used by type of care, main reasons for using
formal and informal care, and arrangements other than care; details of requirements for
(additional) formal care including number of days required, main type of care, main reason
required, whether required/available and main reason not used. Furthermore several
household characteristics are included, such as geographical location (state and capital
city/rest of state), parents' age, income, weekly hours worked, labour force status,
occupation and source of income; and family type. In addition it contains detailed
information on the children such as the number of dependent children aged 12 and over;
children's age, and whether attending school. Finally, information is available on receipt and
eligibility for childcare subsidies, that is whether the Childcare Cash Rebate was claimed for
formal or informal care and if it is not claimed the main reasons for not claiming are given;
and whether a family received Childcare Assistance.
The Child Care Survey provides a rich source of information on the levels and the variation
of childcare costs across families. It also includes data, which allow us to match this
information with the households in the Survey of Income and Housing Costs. Finally, the
1996 Child Care Survey gives us information dated at the mid-point of the data underlying
MITTS. Hence we do not have to worry about changes due to time factors affecting the
comparison of childcare costs in the two surveys.
The first step in the study will be to estimate childcare cost functions using information from
the Child Care Survey. Expenditures on childcare will depend on household characteristics,
geographical location and labour supply behaviour. We will separate the variation in total
expenditures originating from usage and from hourly costs. The second step is to impute the
estimated childcare costs for different levels of labour supply for the households in the
Survey of Income and Housing Costs using common household characteristics. The third
step is to re-estimate the sole parents’ and couples’ labour supply models with the extended
childcare cost component. In addition, the MITTS model will be updated to include the
calculation of childcare subsidies for which families are eligible. Finally, after estimation of
the updated labour supply model, simulation of a childcare related policy change will be
performed and reported.
The following recent papers from the UK provide examples of the use of childcare costs
data in studies of labour supply. Blundell et al. (2000a, 2000b) and Duncan and McCrae
(1999) estimate a labour supply model accounting for childcare costs by looking at the
distribution of hourly childcare costs and estimating the number of hours of childcare
needed for different family types at different hours of work. Their results suggest that the
inclusion of childcare costs is important in the measurement of fixed costs of participation
and affects the resulting labour supply elasticity estimates.

Blundell, R., A. Duncan, J. McCrae and C. Meghir (2000a), “The Labour Market Impact of
     the Working Families’ Tax Credit”, Fiscal Studies, 21(1), 75-104.
Blundell, R., A. Duncan, J. McCrae and C. Meghir (2000b), “Evaluating In-Work Benefit
     Reform: the Working Families Tax Credit in the UK”, mimeo, Institute for Fiscal
Duncan, A. and J. McCrae (1999), “Household Labour Supply, Childcare Costs and In-Work
     Benefits: modelling the impact of the Working Families Tax Credit in the UK”,
     mimeo, Institute for Fiscal Studies.
The Melbourne Institute contact for this project will be Dr Guyonne Kalb.
Project is at the draft report stage

Shared By:
Description: Stream 3 The Supply of and Demand for Income Security