VoxEU Column Labour Markets

Employment Protection Legislation and Economic logic

Most of the justifications for politically popular Employment Protection Legislation (EPL) ring hollow. Governments ought to tackle adverse selection, risk aversion, and unequal bargaining power issues with more finely tuned tools.

Someone stuck holding a hammer soon decides that every problem is a nail. In European countries saddled with Employment Protection Legislation (EPL), many labour market analysts have taken the politically popular rigidity in hand and then set about searching for market failures that it might address. However, firing taxes are ill-suited to correct most of the labour market imperfections cited as possible justifications for EPL. Governments ought to directly tackle issues such as adverse selection, risk aversion, and unequal bargaining power in labour markets with finely tuned tools, rather than bluntly hammering them with EPL.

EPL and empowerment

One of the most popular justifications for EPL is the claim that it empowers employees, who suffer relative weakness in bargaining with their employers. If labourers are indeed at such a disadvantage, a firing tax is inappropriate to correct the problem.

First, from a welfare perspective, redistribution should take place between individuals, not between an individual and an institution. A firm is an institution, not a rich person – it may be owned by pensioners who are in fact poorer than the employees.

Second, tax incidence matters. Under free entry, pure profits vanish, so it is not possible to redistribute from firms to workers: policies that reduce profits simply reduce entry and lower employment. If there is no free entry and firms enjoy monopsony power in bargaining with workers, then the appropriate remedy is promoting competition, not regulating the labour market, which is likely to further deter entry.

EPL and relationship-specific investment

EPL proponents have also defended the regulation on the grounds that, by promoting greater employment duration, the tax encourages relationship-specific investment by firms and workers. If people expect to stay longer with the firm, they are more willing to invest in firm-specific human capital. If they lack such expectations, they do not invest, and the firm’s incentive to dismiss the workers is higher – a self-fulfilling prophecy.

However, these effects are entirely internalised by the firm/worker pair, and government policy is not necessary to correct them. Firms can voluntarily provide protection to their employees. If they do not, it means the losses of specific investment do not outweigh the gains from flexibility, and there is no reason this should be different from a social welfare perspective. Moreover, if the specific investment is indeed too low, the appropriate correction is a separation tax, not a firing tax. Workers should be equally taxed for voluntarily leaving the firm.

EPL and adverse selection

A third argument made for EPL is that government should impose a minimal level of employment protection, because the “market” for it is imperfect. Firms and workers do prefer a contract with employment protection, but the first firm to offer such a contract would only attract the most risk-averse, low-productivity workers. Therefore, the market under-provides employment protection. Government can solve this coordination failure by imposing a minimum standard of protection.

This “market for protection” story is inadequate for three reasons. First, fundamentally, we expect workers to value security generally, rather than at a particular job. In a fluid, full-employment labour market, demand for employment protection will be low. Second, adverse selection models in which workers are offered a menu of contracts trading off wages and tenure tell the opposite story – low-productivity, risk-averse workers are able to enjoy job security while high-productivity workers must accept too low protection in order to differentiate themselves in the separating equilibrium. Moreover, it is not obvious that these workers would be aided by a government-imposed pooling equilibrium, under which they would receive lower wages due to the other type’s lower productivity. Third, employment protection creates new adverse selection problems that are probably more severe. In the absence of protection, firms may lower the wages of workers observed to be less productive. If firing is made more difficult, then adverse selection problems at the hiring stage are exacerbated, along with the firm’s incentive to spend resources in screening potential employees.

EPL and inefficient separations

EPL can also be defended on the grounds that the private cost of labour is larger than its social opportunity cost. In a variety of models, wages are too high because workers earn rents and because their private opportunity cost of labour is too large. As a result, separations are inefficiently high: jobs with positive net social continuation value are discontinued. A firing tax corrects this by discouraging separations.

However, that argument ignores the other side of the problem, which is that hirings are also too low. Indeed, the separation tax further reduces hires. The real issue here is that employment is too low because wages are too high. That hirings are too low and separations are too high is just a corollary of this fact. The first-best solution is a wage subsidy to offset the wedge between private and social opportunity cost of labour. This efficient outcome may be replicated by the combination of a firing tax and a hiring subsidy, both equal to the wedge divided by the interest rate. Ironically, most European governments impose large payroll taxes – the very opposite of a wage subsidy.

EPL and risk aversion

A final suggested justification for EPL is the argument that workers are risk-averse and they need to be compensated for a negative income shock. This generates two questions:

  • What kind of shock requires compensation?

  • Why are markets inadequate to provide insurance against such shocks?

In a Walrasian labour market, job loss does not imply welfare loss, as workers immediately find another job paying the same wage. Thus, there is no need to insure that event. While this sounds extreme, to the degree that real world labour markets function well, they alleviate the need for job loss insurance. In the United States, unemployment duration averages two months, so the market provides considerable insurance to workers.

Should we intervene in the labour market in order to provide further insurance? Cohen (1999) has estimated the present discounted value of losing one’s job in France and the US and finds that it is larger in France, despite much more generous unemployment benefits. This suggests that the distortions created by these benefits are so large that they outweigh their direct effects on insurance.

The US experience suggests that the major costs associated with job loss are due to lower wages in future jobs, not long unemployment spells. These wage losses suggest rents exist. If these rents are from firm- or sector-specific human capital, then it makes sense to insure against job loss, which devalues that human capital. However, employment protection is an inefficient instrument, as it insures against the loss by keeping workers in the wrong sectors. More desirable means of insurance provision are feasible.

Short your factory

One approach would move away from EPL to a system of insurance provided by owning a portfolio of financial instruments whose return goes up when the value of one’s occupational, sector-specific human capital falls. If you work in a shoe factory, be short in shoe stocks. Workers would then be insured against a fall in the price of the good they produce and against a fall in the average wage in their occupational category. Since a worker’s compensation would not depend upon his or her individual actions, such a system would not have the adverse effects of employment protection, which maintains people in the wrong jobs and creates incentives to shirk.

The preceding discussion demonstrates the importance of distinguishing between the goals of correcting labour market distortions and of insuring workers. With regard to the former, much of the literature takes employment protection as given and finds second-best justifications for it, in contexts where it has no reason to be the best possible solution. In some circumstances, EPL actually exacerbates the market failure identified as its justification. As for insurance, maintaining people in the wrong sectors is clearly an inefficient means of provision. Well-designed financial instruments are much more promising.