GPP 501 Microeconomic Analysis for Public Policy Fall 2017

Transcription

1 GPP 501 Microeconomic Analysis for Public Policy Fall 2017 Given by Kevin Milligan Vancouver School of Economics University of British Columbia Lecture November 15 th Asymmetric Information: Social Insurance GPP501: Lecture Nov. 15th 1 of 21

2 Agenda 1. Why social insurance? 2. Unemployment insurance GPP501: Lecture Nov. 15th 2 of 21

3 Social Insurance: Arrow s motivation Here is a quote from Kenneth Arrow: (AER 1963) I propose here the view that, when the market fails to achieve an optimal state, society will, to some extent at least, recognize the gap, and nonmarket social institutions will arise attempting to bridge it. Q: Why might markets fail to achieve optimal state? Q: What assumption about markets might be failing? GPP501: Lecture Nov. 15th 3 of 21

4 Social Insurance: When does government insurance help? Why would we ever want social insurance instead of private insurance? We have private insurance for cars, life, houses. Why not for other risks as well? What is different? Three arguments in favour of social insurance: Information problems. Moral hazard and adverse selection. MH is hidden action, AS is hidden type. But these problems affect all insurance markets why is social insurance different? We ll see. Risk pooling. Social insurance allows us to pool risk across regions / generations. But couldn t a firm operate across regions? Can t families provide intergenerational insurance? Economies of scale in insurance / administration. But wouldn t private firms reach their efficient scale for operation? GPP501: Lecture Nov. 15th 4 of 21

5 Social Insurance: What kind of policies? What risks? Here are several types of insurance that are sometimes provided by government. a) Health insurance b) Unemployment insurance c) Public pensions (longevity insurance) d) Welfare (low income insurance) e) Disability Insurance f) Workers Compensation g) Long-term care insurance h) Drug insurance i) Auto insurance j) Mortgage insurance Q: What risk is each of these covering? Q: Are there severe adverse selection issues? Q: Are there severe moral hazard issues? Q: Could private insurers handle this? Why or why not? GPP501: Lecture Nov. 15th 5 of 21

6 Social Insurance: Does public insurance crowd out private? Crowd out definition: As public insurance rises, does private insurance coverage fall? Why is this important?: If public insurance simply crowds out private insurance, then we are just making a social choice about whether production should be in the private or public sector. However, if there is not complete crowd-out, social insurance insures privately uninsurable risks and fills missing markets; social insurance improves efficiency. GPP501: Lecture Nov. 15th 6 of 21

7 Social Insurance: Is there crowd out? Hans Werner Sinn Social Insurance, Incentives, and risk-taking International Tax and Public Finance, Vol. 3, Iss. 3, (1996). Hans-Werner Sinn estimates that about 85% of spending on social insurance covers risks that would not be covered by private insurance. In other words, public insurance largely does not crowd out private insurance. GPP501: Lecture Nov. 15th 7 of 21

8 Social Insurance: Is there crowd out? As GDP goes up by 1%, private insurance premiums goes down by 0.15%. GPP501: Lecture Nov. 15th 8 of 21

9 Agenda 1. Why social insurance? 2. Unemployment insurance GPP501: Lecture Nov. 15th 9 of 21

10 Why is unemployment insurance often public? Why is unemployment insurance mostly public? Can asymmetric information explain why the private UI market doesn t appear to exist? Adverse selection: Those at risk of unemployment would be more likely to purchase, biasing the risk pool. Solution? Mandatory participation. Moral hazard: The insured change their behaviour search less intensely or take employment that is more risky. (Q: Is this good or bad?) Note: Larger firms have severance payments and agreements. Q: Is this like UI in some way? Q: How do they get around adverse selection / moral hazard? GPP501: Lecture Nov. 15th 10 of 21

11 UI and Moral Hazard I: Does insurance affect work? Jane Friesen, The effect of unemployment insurance on weekly hours of work in Canada, Canadian Journal of Economics, Vol. 35, No. 2 (2002). Before 1997, eligibility depended on weeks of work over the period before unemployment. You got credit for a week if you worked 15 or more hours. New policy in 1997 changed this to hours. Does this make sense? Can you predict any behavioural responses? GPP501: Lecture Nov. 15th 11 of 21

12 UI and Moral Hazard: Does insurance affect work? Got a full week at 15 hours, so now not so much benefit being at hours. Increase in value of working many hours in a week. Q: why is this evidence of moral hazard? GPP501: Lecture Nov. 15th 12 of 21

13 UI and Moral Hazard II: Does insurance affect job search? Black, Dan, A., Jeffrey A. Smith, Mark C. Berger, and Brett J. Noel "Is the Threat of Reemployment Services More Effective Than the Services Themselves? Evidence from Random Assignment in the UI System." American Economic Review, 93(4): Evaluated the Worker Profiling and Reemployment Services in Kentucky in the 1990s. Offered job search assistance services to recipients of UI in Kentucky. You got a letter around week 2 of your benefit spell, telling you to come to training sessions days later. Randomized people into this program. Research Q s Does long-run employment increase? Earnings? GPP501: Lecture Nov. 15th 13 of 21

14 The Kentucky letter: GPP501: Lecture Nov. 15th

15 The timing. GPP501: Lecture Nov. 15th 15 of 21

16 What happens to employment? Short-run vs Long-run GPP501: Lecture Nov. 15th 16 of 21

17 Main Findings: 1. The program reduces mean weeks of UI benefit receipt by about 2.2 weeks, reduces mean UI benefits received by about $143, and increases subsequent earnings by about $1,000. Given its very low cost, the program easily passes standard cost-benefit tests. 2. All of the above effects concentrated in the first quarter when the letter received; no long-run difference in employment or earnings. GPP501: Lecture Nov. 15th 17 of 21

18 UI and Moral Hazard III: Subsidizing Risk-Taking Acemoglu, Daron and Robert Shimer, Efficient Unemployment Insurance The Journal of Political Economy, Vol. 107, No. 5 (Oct., 1999), pp They place unemployment insurance in a dynamic search model. UI acts as a subsidy to search. Q: Is that a bad thing necessarily? Why or why not? GPP501: Lecture Nov. 15th 18 of 21

19 UI in a Search Model A standard search model with UI: Workers know the distribution of wage offers. They choose a reservation wage. They wait until they get an offer that exceeds their reservation wage. UI causes reservation wage to increase, thus increasing the expected time looking for work. Key assumption of standard model is that the distribution of wage offers is fixed; doesn t change in the presence of UI. GPP501: Lecture Nov. 15th 19 of 21

20 UI in a Search Model So, in this model, what is different? With UI, workers are willing to take on more risk because they have a silent partner (the gov t). This leads them to search for higher paying jobs. **A&S innovation** This makes firms invest more and supply higher paying jobs. Output is higher. This is different from the standard story because the distribution of wage offers is different in the UI economy. GPP501: Lecture Nov. 15th 20 of 21

21 For next time: We will cover health insurance and pensions. GPP501: Lecture Nov. 15th 21 of 21