Working out the job market
All of us are born unemployed and single, and if we want that to change, we will have to start looking for a suitable match
Supply and demand is a fundamental economic concept, and unemployment is a totemic economic problem. But apply the concept to the problem, and you will not get very far. The logic of supply and demand says that if wages are high, lots of people will want to work, but few people will want to employ them; if wages are low, employers will be hungry to go hiring, but few people will want to work. At equilibrium, the number of hours available equals the number of hours people are willing to work. Unemployment is impossible, unless there is a minimum wage – this suggests, for instance, that unemployment was unknown in the UK before April 1 1999, which is not my recollection. The supply-and-demand approach offers little insight into job-market recessions, or why different countries have such different experiences of employment.
In this year’s Royal Economic Society public lecture, Christopher Pissarides, winner of the Nobel memorial prize in economics in 2010, set out to resolve the mystery. Pissarides, along with Peter Diamond and Dale Mortensen, has developed a model of job-matching that has become the standard way macroeconomists think about labour markets.
The basic insight is nothing staggering. There are job-seekers in the world, and there are job vacancies in the world, and the aim is to match seekers to vacancies to create actual “jobs”, which are matched pairs of former vacancies and former job-seekers. Searching for suitable vacancies, or suitable employees, is costly, and neither jobseeker nor employer knows whether any match will work out.
In such “search models”, unemployment isn’t a puzzle; it’s the natural state of economic existence, just as being single is the natural state of romantic existence. All of us are born unemployed and single, and if we want that to change, sooner or later we will have to start looking for a suitable match.
Once Pissarides, Diamond and Mortensen began to write models that encapsulated some of these commonsense observations, they discovered a natural explanation for the “Beveridge curve”.
The Beveridge curve is a simple downward-sloping relationship between the vacancy rate in an economy, and the unemployment rate. In good times, vacancies are plentiful and unemployment is low; in a recession, the economy slides down the Beveridge curve to a place where vacancies are scarce and unemployment is high. More interesting is the fact that the curve itself sits in different positions for different economies, and it can shift. The Beveridge curve in much of the EU is higher than that in the US, for instance – for any given level of vacancies, there will be less unemployment in the US.
This fits a search-and-matching explanation. If the curve shifts outwards, with both vacancies and unemployment rising simultaneously, that is a sign of some kind of structural failure to match: there are potential jobs but for some reason, the match between vacancy and jobseeker is not occurring as quickly as usual. The US is showing signs of this structural stress: vacancies are on the rise but unemployment is falling more slowly than we would expect based on past experience.
Meanwhile, Germany, whose labour market has been defying the financial crisis, has enjoyed structural gains: unemployment has been falling even when vacancies have not been buoyant. Pissarides credits the delayed effects of Gerhard Schroeder’s labour market reforms, with more flexibility and plenty of incentives to match young people with jobs.
The question, of course, is what feature of Germany’s labour market has proved decisive in this – and what we can transplant into other countries. Even a Nobel laureate was not able to give a convincing answer to that question.
Also published at ft.com.
8 Comments
Roger Hyam says:
My understanding is the the German labour market is highly regulated compared to the US. Employees have more rights and it is harder to hire and fire people.
Does this suggest that lack of regulation is good in a boom but bad in a bust. Surely not that simple…
Germany is also a far more equal society!
15th of December, 2012Jon says:
A quibble. Surely the minimum wage example in the first para is flawed in that the UK’s benefits system prior to the implementation of minimum wage legislation would have had an effect on employment levels so full employment wouldn’t be expected in the absence of minimum wage legislation
Perhaps one could say that a benefits system is a form of minimum wage mechanism (so could vary across countries) but to suggest a point based solely on implementation date of the legislation strikes me as ‘iffy’.
15th of December, 2012Nick says:
Surely unemployment with benefits is just another sort of job. The pay is poor but the hours are good. It is not enough to just consider the weekly wage – the number of hours of work to get that wage is a factor in any job.
16th of December, 2012Jon says:
I reckon German employment has benefited from downward pressure on the euro boosting exports as well cheap capital due to the flight to safety experienced across the Eurozone – just look at the real rates on German debt/yield spreads! If Germany didn’t use the euro surely exports would drop as the currency appreciated?
16th of December, 2012Hamish Atkinson says:
If benefits were unaffected by employment, the labour market would more much more fluid. People claiming benefits often aren’t “unable to get a job” , they have done a marginal analysis on the pros and cons of getting a job: Extra hours worked – obvious con. Loss of benefits – except for the first £5, £1 lost for every £1 earned (income support), then 80p in the £1 after that (housing and council tax benefit). Cost of getting to work.
16th of December, 2012Viewed marginally, the minimum wage isn’t £6.19, it’s approx. 90p-£1.23. Would you work for 90p/hour? No? Well, then that explains unemployment.
Chris Hodder says:
Hamish Atkinson, you are a genius (although you did forget the effect of tax). Many have suggested universal benefits so that all earned income is a marginal gain, but politically this is dead in the water and the economics of hoping that everyone would continue “striving” when all basic needs are met hasn’t been evaluated.
17th of December, 2012James Bartholomew says:
I agree with Jon’s first remark but I am afraid I would struggle to put it so gently. The whole article begins with the astonishing idea that there was no minimum wage before 1999. Of course that is technically true. But it completely ignores the effects of welfare benefits received by millions. For many people, especially in the early 1980s, welfare benefits were worth very close to net earnings or actually more. Even more recently, welfare benefits have offered terms not much worse than legal work. Nobody should even begin to write about unemployment without considering the effects of welfare benefits which can be as powerful or more powerful than minimum wages. You should also remember that the level of minimum wages is crucial. A low minimum wage makes little difference. Do you really think a minimum wages of, say, £40, would not cause unemployment on a major scale?
20th of December, 2012Matt Perks says:
Maybe the reference to the introduction of the minimum wage was a bit of a throwaway comment; that first paragraph is really just dismissing a supply/demand model for labour. However, I think it would be fair to state that benefits for the unemployed were pretty limited in the 1930s and almost non-existent prior to 1911. And there are plenty of examples in the developing world today. Anyone wanting to make a case for a supply/demand model of employment will need to show, for example, that there was no unemployment in the 1930s, which is not my recollection.
21st of December, 2012