In many countries, calls for raising the minimum wage are growing louder. The “Fight for $15” movement has picked up momentum since 2012, when some New York fast-food workers began holding demonstrations calling for a higher minimum wage. The drive gained advocates around the world. The city of Seattle, Washington became famous for its decision to gradually impose a $15-per-hour minimum wage. Germany implemented a national hourly minimum wage of 8.50 euros in 2015.
Such measures are seen as a way to combat poverty and inequality. These two social ills can have serious political and even geopolitical ramifications: too much poverty or too much inequality can lead to social unrest, instability and regime change. Seen in this light, such policies appear wise. And while traditional economic theory predicts that raising the minimum wage will have a negative impact on employment, empirical studies have come to mixed conclusions. Some of the results have given activists and politicians who support the movement the justification they were looking for. Where does the debate stand now?
In Germany, the new national minimum wage introduced in 2015 was 8.50 euros, now 8.84 euros after a first hike. The move is said to have had favorable effects overall, with no negative impact on unemployment. Employees got paid more and productivity increased in some cases, not only because of a bigger workload, but also due to a reorganization of labor and a better atmosphere. Yet the measure was passed with transitional periods and exemptions for small businesses, and in a context of very low unemployment and inflation, an aging population and a favorable growth cycle. Noncompliance and black-market labor could also partly explain the lack of effect on unemployment.
On the other hand, a recent study by the National Bureau of Economic Research (NBER) based on the most detailed data ever gathered on the Seattle experiment, seems to throw cold water on the optimism: if the first minimum wage increase from $9.47 to $11 in 2015 only had a slightly negative impact on employment, the second hike, in 2016, from $11 to $13 was associated with a 9 percent decrease in low-wage jobs, and a 3 percent decrease in payroll for such jobs – implying lower earnings of $125 per month in 2016. It seems that past a certain level, a minimum wage does indeed have a negative impact.
As the Seattle experiment shows, while higher minimum-wage policies have good intentions, there are unintended consequences. “Forcing” a change in economic reality by trying to hasten economic progress can end up having the opposite effect: a minimum wage that is too high can hurt workers in several ways.
First, even with no impact on employment, if wages are forced upwards, prices could rise as businesses pass on their higher costs to consumers. Price increases can offset the rise in wages, leaving workers with their real incomes basically unchanged.
Second, and probably more important, is the issue of the gap between workers’ productivity and their wages: to be “valuable” – that is, “profitable” (as unpleasant as it might sound) – a worker must generate more value than his or her cost to the entrepreneur. Otherwise, hiring the employee would result in a destruction of value. In such cases, the entrepreneur simply stops employing the worker. This idea can be understood in two ways.
One is the pure “skill” aspect: people with lower skills are typically less productive. These include young people, those with low levels of education and poor, unintegrated migrants. Paying them far more than the value their skills provide is an impossible strategy. Instead, businesses will react by hiring cheaper, illegal labor or replacing workers with more cost-effective technologies, such as the automated kiosks in fast-food restaurants. If workers are not easily and cheaply replaced by machines or illegal labor, entrepreneurs can either curtail their activity (fewer working hours) or shut down.
Another way to look at the issue is through the idea of “network size.” Different areas or “economic territories” have different levels of economic development, due to historical factors, the division of labor and their development of specializations. As 18th-century economist Adam Smith told us, these are a function of the economic network size: a larger market (with more people and wealth) enables more specialization. And as another economist, Allyn Young, later added, increasing specialization (because of the larger incomes this process generates) allows for larger markets. This simple view can have profound implications for minimum-wage policies.
Think of two workers with the same skills doing the same job in two different settings: one works in a more advanced economic network – typically a “larger” market – and the other in a less advanced environment. The first worker can generate a lot more value than the second. Forcing the latter’s wage too high in the less advanced market means an entrepreneur would not pay the worker, regardless of his or her skills, because the local market size could not support such a high wage. For example, imagine the effects of aligning Bulgaria’s minimum wage with France’s.
However, thinking of countries as homogeneous entities can be misleading, because variation in levels of development even across a single country can be staggering. Countries and even cities are made up of various “economic territories.” A minimum wage that seems benign for the national economy can hurt smaller and less advanced local networks (for example, poor suburbs or migrant communities). If the national minimum wage is too high for their level of development, it can result in long-lasting economic exclusion, with all the associated political and social risks.
Of course, the complexity of economic networks makes it very difficult to put them in “economic boxes.” Where does a network start? Where does it end? How does one decide what is part of this or that “economic territory”? Can a certain place belong to several networks? Given such difficulties, the only measurable and relevant “network” is, in fact, the single contract between a worker and a company, which suggests that there cannot really be a state-mandated minimum wage. That is why in countries like Sweden or Denmark, minimum wages are negotiated between unions and employers’ associations, usually on a sector-by-sector basis. This supposes a “horizontal” social model with a very high rate of union membership.
But what if policymakers and populations insist on establishing a mandated minimum wage? What could be the optimal levels of implementation and remuneration? First, it is wise to choose a decentralized level of implementation, for reasons related to the “economic territories” explanation, despite the difficulties with precisely defining them.
Local authorities, in cooperation with firms and trade associations, have a better knowledge of the local labor market. Rather than a single national minimum wage blindly forced onto diverse labor markets, it is better to apply the “subsidiarity principle” and thus circumvent the issue and debates about the negative effects of a national minimum wage. The Swiss canton of Jura comes to mind. In a local experiment, policy outcomes are more directly visible and corrections can be made more easily.
Institutional competition between jurisdictions would tend to slowly adjust local minimum wages to their “reasonable level,” – not too low to attract workers, not so high as to hamper employment under local economic conditions.
It would also be wise to implement a minimum wage that is not too close to the jurisdiction’s median wage. Minimum wages that stray too close to the median (say, more than 50 percent of the median wage) can have a detrimental effect on employment and compress the wage scale. In Germany, for instance, the new national minimum wage was set at 48 percent of the median wage – versus 61 percent in France.
Rolling back a minimum wage is very difficult politically. In countries with a minimum wage that is high relative to the median wage, governments (like in France) can try to mitigate the negative impact by reducing social contributions on lower wages, to reduce companies’ labor costs while protecting workers’ net earnings. However, this forces other workers to pay disproportionately for welfare benefits. Also, there tends to be a “lock-in” effect on wages around these low levels, which could potentially cause long-term instability.
To avoid the many problems of centralized minimum wage systems and leave enough room for policymakers to maneuver, one option could be to set a low national minimum wage, with higher local minimum wages – if deemed necessary and appropriate given local conditions. This is already the case in the United States, where not all states have a minimum wage. Berlin, like Seattle, has decided to set its minimum wage (9 euros) higher than the federally mandated one, for instance.
An alternative path could be to change the social model toward a decentralized regime as in the Nordic countries, with sectors and trades in charge of the minimum wage. Trends will obviously depend on a country’s history and social model, as well as the amount of political pressure for a minimum wage. In the context of a potential increase in immigrants with low skill levels, serious consideration should be given to which option will better absorb these newcomers and mitigate the risks of non-integration.
Emmanuel Martin is the general manager of the French MOOC project “École de la liberté.” He holds a PhD in economics from the University of Aix-en-Provence.