EC3311. Seminar 2. ² Explain how employment rates have changed over time for married/cohabiting mothers and for lone mothers respectively.

Transcription

1 EC3311 Seminar 2 Part A: Review questions 1. What do we mean when we say that both consumption and leisure are normal goods. 2. Explain why the slope of the individual s budget constraint is equal to w. 3. Draw a diagram to highlight the income and substitution e ect of a wage increase. 4. An individual s reservation wage is given by the slope of her indi erence curve through her endowment point. Explain! 5. How can a labour supply curve be backward bending? 6. What is the typical empirical nding about men s labour supply elasticity (with respect to wage)? Part B: Questions relating to Gregg, P. and S. Harkness (2003) Welfare Reform and Lone Parents Employment in the UK, Mimeo, CMPO, University of Bristol. Web-address:?? (FIX) Some Background to the Paper Main welfare system support those who are out of work is called Income Support (IS). In 2002 there were around 1.7 million lone parents in the UK, of who approximately half (850,000) were on Income Support (IS). Alongside IS has been a system of bene ts that can be claimed by those working; this system used to be called Family Credit, but was changed in 1999 and renamed Working Families Tax Credit (WFTC). A speci c problem has been the low employment rates among lone mothers. The employment rate of lone mothers in the UK has been uctuating between percent, which is low in international comparison. These very low employment rates have contributed towards the UK having the highest proportion of children living in workless households in OECD countries in 1996, and one of the highest incidences of child poverty. The incoming Labour government in 1997 initiated a series of policy reforms aimed at reducing child poverty. A key element of this was the move to increase employment rates among families with children, especially among lone parents. The WFTC was a key part of this strategy. The current paper considers whether the welfare reforms (including the move to WFTC) has been successful. ² Explain how employment rates have changed over time for married/cohabiting mothers and for lone mothers respectively. 1

2 A: In the late 1970s employment rates of lone parents and married mothers were broadly similar. However over the last thirty years married mothers employment has increased rapidly, while over the same time those for lone parents have fallen. Figure 3 in the paper uses data from the General Household Survey (GHS) and Household Labour Force Survey (HLFS) to show how employment rates have changed for single and married mothers, and single childless women, between the late 1970s and Married mothers employment rates were only marginally higher that those of lone parents in the late 1970s. However, while employment rates of lone parents fell and then stagnated in the 1980s, for married mothers employment rates grew steadily from around 1984 so that by the mid-1990s employment rates of single childless women and married mothers were broadly similar. This divergence in the experiences of single and married mothers meant that by the early 1990s a gulf in the employment rates of single and married mothers had emerged. ² How does the employment rate among lone mothers in the UK compare with other countries? A: Over the last thirty years the employment of married/co-habiting mothers in the UK rose dramatically, especially for those with young children. et the employment rate of lone mothers was lower in the early 1990s than it was in the late 1970s, at just under 40 percent, and 25 percentage point lower than the employment rates of married mothers. Figure 1 in the paper reports lone parents employment rates among OECD countries. This clearly shows that only Ireland, Poland and Australia have lower rates of employment of single mothers than the UK. While the employment rate was around 40 percent in the UK in 1999, in countries including the US, Canada, Italy, Sweden, Finland and Portugal employment rates were at above 65 percent. Figure 2 in the paper contrasts the employment rates of single mothers with both those of married mothers and single childless women within each country. In both cases the gap in employment for lone parents and other women is greater than in any other OECD country. ² Brie y describe the welfare reforms (in the UK) that the authors focus on (i.e. the changes that occured during ). How was the aim of these reforms di erent from the reforms that we happening in the US? A: Two major policy reforms. 1. The rst was to raise the nancial gains to working for families by means of the Working Family Tax Credit (WFTC). The move (from Family Credit) towftc in October 1999resulted in three major changes: rst the credits became more generous, second variations in credits by age of children were progressively eliminated (they used to be very strongly di erentiated by age), and third the level of earnings at which credits could be retained 2

3 increased substantially. These changes are illustrated in Figure 5 in the paper. (There were also changes tothe treatment of childcare costs. Under WFTC a fraction of childcare costs up to 70% of childcare costs of up to 120 a week were added to the basic credits, thus making them available for those low income mothers working part time). 2. New Deal for Lone Parents (NDLP) introduced active case management into the welfare system, in order to encourage and support single parents to move back to work. The policy attendance to discuss work options with an advisor compulsory for lone mothers on welfare ( Income Support ). The reforms in the US were more directly targeted at reducing the number of individuals on welfare; e.g. time-limits (i.e. restrictions on the amount of time that bene ts could be claimed) were introduced on the main welfare programs, mandatory job search was enforced more tightly, bene t generosity was reduced. The reform in the UK were more directed at helping lone parents both in and out of work (in an e ort to reduce child-poverty in particular). Thus, unlike in the US where in-work bene ts were introduced with the primary objective of welfare caseload reduction, in Britain the dominant policy aim has been to raise incomes for lone parents both in and out of work, with an increased earnings contribution being an important component of the intended income gains. ² Describe the empirical methodology used by the authors (Nb. ou do not need to go into details about the matching process, but you should understand the general idea behind the di erence-in-di erences estimator). A: The methodology used in the paper is a di erence- in-di erence approach to assessing the impact of policy. This methodology cannot precisely disentangle the e ects of individual policies e ects (WFTC vs. NDLP); however, it provides powerful and straightforward evidence on behavioural impacts of the overall impact of the total policy package. As the objective is to estimate the impact of policy reform on lone parents employment, if we denote employment rate by then the impact of policy change on employment is the di erence between the post policy outcome, w ith policy (e.g. the participation rate), and the outcome that would have occurred in the absence of policy changes, w ithout policy. We seek to determine the di erence w ith policy w ithout policy. However, we face the problem that w ithout policy is not observed since all lone parents where exposed to the policy. Hence we need to seek a suitable control (or counterfactual ) group. The idealcounterfactual group should not have experienced any policy shocks e ecting their employment, but should have the same set of observed and unobserved employment attributes, have experienced the same local labour market shocks, and reacted in an identicalway to them. The benchmark group therefore should share as many common characteristics as possible with the focus group (i.e. the lone parents); the only di erence between them should be their experience of policy. As singles without children are relatively una ected by policy 3

4 change (since they are not eligible for WFTC and not included in NDLP), the authors use them as our main control group. In order to make the estimates more precise, the authors use statistical techniques to pick out individuals from the latter group that have characteristics that are similar to the group of lone parents. (Note e.g. that singles without children are in general likely to have di erent age and levels of education than the set of lone parents.) So the key idea is to try to nd, among the singles without children, a set of individuals who have characteristics very similar to the group of single parents. The idea is then that we can compare what happened to participation rate of the group of single parents to what happened to the participation rate of the matched singles without children. Since the former group was exposed to the policy while the latter was not, we can take the di erence in their di erences as our measure of the impact of the policy. To be a bit more formal, let 1 LP be the participation rate of lone parents after the policy was introduced and 0 LP be the participation rate of lone parents before the policy was introduced. Similarly, let 1 S and 0 S be the participation rates of (matched) singles after and before the policy was introduced respectively. 1 S may di er from 0 S e.g. due to a business cycle e ect. On the other hand 1 LP and 0 LP will di er due to a business cycle e ect AND due to the policy reform. Hence if we assume that both groups have been a ected in the same way by the business cycle we have that S S = business cycle e ect LP LP = business cycle e ect + e ect of policy Hence, when we take the di erence-in-di erences we obtain: LP LP S S = e ect of policy (As a side-note, the authors also note that there were some trend of convergence or divergence already before the reforms took place. They account for that be computing these trends on the years prior to the reform and adjust the estimates accordingly.) The authors report results using as pre-period (i.e. the before policy case). ² What impact of the policy reforms on employment rates of lone parents do the authors nd? (See in particular the section on years ). A: According to the Household Labour Force Survey, between 1992 and 2002 employment rates of lone parents rose from 42 to 53 percent. This rise began before the new policy regime came into e ect in 1998, but sped up thereafter. The question is then: how did the development of the employment rate for the lone parents compare with that of the singles? Second there was a decade long sustained rise in employment among the population as a whole from

5 onwards. However, while there was such a rise in employment rates throughout the period for the population as a whole, there was a marked slow down after The di erence- in-di erence method of estimation predicts that since 1998 lone parents employment has risen by 6-percentage points more than would have been expected from a population of singles without children with the same characteristics! If we believe that the relative employment gains made by lone parents were entirely due to policy reforms, then this simple di erence model provides an estimate of the impact of post-98 policy reform. Example of calculation: ² The employment rate of lone parents in 1998 was percent; the corresponding employment rate in 2002 was 53.19, a di erence of 53:19 46:59 = 6:6 The employment rate of (the matched) singles in 1998 was percent; the corresponding employment rate in 2002 was 66.77, a di erence of 66:77 66:13 = 0:64 Hence the employment rate by 6:6 0:64 = 5:96 percentage points more for lone parents, which we attribute as a policy e ect. (The authors modify this estimate downwards somewhat by noticing that there was a pre-existing di erences in trends for the two groups; nevertheless, they argue that the e ect of the policy was roughly a 5 percentage point increase in the participation rate for lone parents. The authors go on to investigate if there were any noticable di erence among groups, in terms of age of children and education, but we can ignore this.) ² What do the authors nd on the e ect of the policy reform on hours worked? What does theory predict? A: A key feature is that a lone parent must work a minimum of 16 hours in order to claim the tax credit, and there is a supplementary credit if they work in excess of 30 hours. Theory therefore predicts that those who had been working fewer than 16 hours a week to increase their hours of work, while those working more than 16 hours may decide to reduce their hours as a result of the windfall e ect of increased welfare payments. The authors estimates suggest those who have entered work as a result of the reforms work relatively fewer hours. They predict that those already in work have reduced their average hours of work marginally, by around half an hour a week, in response to the windfall e ect of increased welfare income a small income e ect. 5