Don Paskini claims that higher welfare benefits help people find work, pointing out that between 1996 and 2009, the numbers of lone parents in work increased, at the same time as benefits did.
His commenters are right to say the evidence doesn’t prove his claim, because perhaps even more people would have found jobs if benefits had been lower. But the evidence does prove something - that higher benefits can be consistent with lower benefit claims.
The problem here is that the anti-welfare lobby misrepresents its case. For example, when Neil O’Brien claims that higher benefits encourage people to stay on benefits rather than find work, what he should say is: “other things being equal, higher benefits will, at the margin, discourage people from seeking work.”
Put like this, four questions arise.
1. Is even this true? Yes, higher benefits encourage people to substitute work for leisure. But as Don says, there might also be an offsetting income effect; higher benefits allow people to afford bus fares to Job Centres, or a local newspaper, or child care whilst they are at interviews, or a pleasant-smelling clunge that’ll impress an interviewer. They might, therefore price people into work.
2. How extensive is the margin? One of my big gripes against the right is their tendency to assume that margins are more extensive than they are. And this might be the case here. Some cross-country evidence has found that the link between job search and the ratio of unemployment benefits to wages is statistically insignificant.
3. Isn’t there an aggregation problem here? Let’s concede that lower unemployment benefits do encourage people to look for work. For any individual unemployed person, this increases their chances of getting a job. However, unless aggregate employment increases, their increased chances come at the expense of lower chances of finding work for someone else. The question, then, is: does increased job search lead to increased aggregate employment?
In theory, the answer is yes: an increased supply of labour means lower wages, which encourages employers to hire more - though this raises issues about elasticities. However, the presence of the minimum wage blocks this channel. Unless you call for the abolition of the minimum wage, the expectation that benefit cuts will increase employment is simply the fallacy of composition*.
4. Are other things equal? Don is surely right to say that unemployment depends more upon the state of the economy than the level of benefits.
And this brings me to the key issue. There is always a trade-off between risk and incentives; the low unemployment benefits that give people a high incentive to find work also mean that there is a large risk of losing lots of income if you lose your job.
The question is: what should be our preferences on this risk-incentives spectrum?
In good macroeconomic times, when work is plentiful (not that it ever is, but let that pass) it might be reasonable to prefer a welfare state that provides sharper incentives and greater risk. But this is not the world we live in today, and it might not be for many years.
My fear is that a nasty combination of wishful thinking about the labour market, a desire to stigmatize the unemployed, and unwarranted optimism about the supply-side effects of restricting unemployment benefits is leading the right to overweight the desirability of sharp incentives, and underweight the merits of better risk-pooling.
* I suppose there are other mechanisms through which benefit cuts might increase aggregate employment. But all require considerable levels of elasticity optimism.
His commenters are right to say the evidence doesn’t prove his claim, because perhaps even more people would have found jobs if benefits had been lower. But the evidence does prove something - that higher benefits can be consistent with lower benefit claims.
The problem here is that the anti-welfare lobby misrepresents its case. For example, when Neil O’Brien claims that higher benefits encourage people to stay on benefits rather than find work, what he should say is: “other things being equal, higher benefits will, at the margin, discourage people from seeking work.”
Put like this, four questions arise.
1. Is even this true? Yes, higher benefits encourage people to substitute work for leisure. But as Don says, there might also be an offsetting income effect; higher benefits allow people to afford bus fares to Job Centres, or a local newspaper, or child care whilst they are at interviews, or a pleasant-smelling clunge that’ll impress an interviewer. They might, therefore price people into work.
2. How extensive is the margin? One of my big gripes against the right is their tendency to assume that margins are more extensive than they are. And this might be the case here. Some cross-country evidence has found that the link between job search and the ratio of unemployment benefits to wages is statistically insignificant.
3. Isn’t there an aggregation problem here? Let’s concede that lower unemployment benefits do encourage people to look for work. For any individual unemployed person, this increases their chances of getting a job. However, unless aggregate employment increases, their increased chances come at the expense of lower chances of finding work for someone else. The question, then, is: does increased job search lead to increased aggregate employment?
In theory, the answer is yes: an increased supply of labour means lower wages, which encourages employers to hire more - though this raises issues about elasticities. However, the presence of the minimum wage blocks this channel. Unless you call for the abolition of the minimum wage, the expectation that benefit cuts will increase employment is simply the fallacy of composition*.
4. Are other things equal? Don is surely right to say that unemployment depends more upon the state of the economy than the level of benefits.
And this brings me to the key issue. There is always a trade-off between risk and incentives; the low unemployment benefits that give people a high incentive to find work also mean that there is a large risk of losing lots of income if you lose your job.
The question is: what should be our preferences on this risk-incentives spectrum?
In good macroeconomic times, when work is plentiful (not that it ever is, but let that pass) it might be reasonable to prefer a welfare state that provides sharper incentives and greater risk. But this is not the world we live in today, and it might not be for many years.
My fear is that a nasty combination of wishful thinking about the labour market, a desire to stigmatize the unemployed, and unwarranted optimism about the supply-side effects of restricting unemployment benefits is leading the right to overweight the desirability of sharp incentives, and underweight the merits of better risk-pooling.
* I suppose there are other mechanisms through which benefit cuts might increase aggregate employment. But all require considerable levels of elasticity optimism.
Where do withdrawal rates fit? I would suppose that the difference in income between employment and unemployment is more important than absolute levels. If you are in receipt of various benefits as a result of being unemployed with dependents, you can suffer 100% marginal withdrawal rates in taking low paid employment.
I am not anti welfare, but I am anti aggressive benefit withdrawal rates.
Posted by: Mark | August 29, 2010 at 11:39 AM
3. is a v good, much overlooked point. Although there might be some more things going on at the macro level. I'm not sure about the dynamics of getting to it, but a situation with higher employment might mean higher aggregate demand, lower taxes. So there might be more than "lower wages" to increase the supply jobs.
Posted by: Luis Enrique | August 29, 2010 at 11:39 AM
oh, I see your footnote covers that. I should add I think you're spot on about the right focusing on one margin, and granting it far too much explanatory power.
Posted by: Luis Enrique | August 29, 2010 at 12:10 PM
I don’t think point No. 3 is 100% valid (as your footnote seems to imply). It is true that those on benefits tend to go for jobs at the lower end of the jobs market, but these are nowhere near exclusively minimum wage jobs or even “double the minimum wage jobs”. To illustrate, I am sure a significant proportion of those able to earn double the minimum wage find benefits plus leisure a more pleasant option, at least for a few months. Indeed, I personally know some of these people. Cutting these peoples’ benefits would not induce them to go for minimum wage jobs, thus the min. wage is not an obstacle to their finding work.
Posted by: Ralph Musgrave | August 30, 2010 at 08:15 AM
thank you for an interesting post which challenged my thinking on this subject - it is nice to in a convincing way be told that i am probably wrong about something.
however - contrary to the first sentence of the last paragraph - i don't think that any malicious intent to stigmatize the unemployed is a necessary or even common ingredient in the process of arriving at the possibly false conclusion. in my humble experience, it is rare that my political opponents have arrived at their positions with the intention to be evil; usually we disagree about facts, causality and which potential outcome is more desireable, but share an honest intent to better society.
Posted by: hans | August 31, 2010 at 01:33 AM
here's a relevant recent MR post
http://www.marginalrevolution.com/marginalrevolution/2010/08/barro-v-barro.html
Posted by: Luis Enrique | August 31, 2010 at 09:34 AM
rent £150
jsa £60
cash in hand work £100
----------------------
£310 per week
job @ £20,000
£315 per week
-----------------------
Posted by: james c | August 31, 2010 at 01:49 PM