In contrast to the most advanced economies, almost one in two Spanish workers is employed in a micro-enterprise of less than ten people. In Germany, for example, this proportion is reduced to one in five workers. There is no doubt that this predominance of microenterprises in Spain is an obstacle to the productivity and internationalization of Spanish companies: investing in international trade or automation is practically impossible for a firm of such small size.

Why is microenterprise so predominant in our country? One of the main hypotheses is that the regulations that try to favor small businesses are, at the same time, an obstacle to stop being so. When a company grows above 50 employees, obligations such as having a company committee, employing an account auditor (if it meets income criteria), disbursing VAT monthly and increasing the installment payment in companies. Other rules affect companies in a similar way when they pass other thresholds, such as six million euros or 10 employees.

The risk that such controls mean is that if growth means that the company faces higher regulatory costs, the employer prefers to stay below the threshold at which they will be promoted, be it 10 or 50 employees. This type of restriction could help explain the enormous predominance of microenterprises in Spain and thus help explain the low productivity and low internationalization of our economy.

In a recent study in the American Economic Review, we studied in France the economic cost of regulations very similar to these Spanish regulations. In this country, as in ours, companies face a veritable tsunami of regulations when they reach the threshold of 50 employees. Above this limit, but not below, companies must form a works council, offer union representation, have a profit-sharing scheme, pay more in a training budget, and negotiate with the Ministry of Labor if they wish to reduce its size. Other size thresholds exist, but 50 workers is by far the most important in France.

Our research begins by documenting that there is indeed a sharp drop in the number of companies at 50 employees, exactly where these regulations take effect. There are far more French companies just below the regulatory threshold than above 50. In addition, we note in particular that companies that "cram down" to just below 50 employees are relatively productive (which would justify their scaling up).

It would be hard to argue from this evidence that regulation has no effect, but what is its cost? To quantify expenses, we use basic economic theory, which relates the size of the company to the productivity of its technology and its managers. The most efficient or innovative companies have higher productivity and tend to be largefor example, they may have advanced technologies like Facebook or Google or an efficient manufacturing process like Zara or Toyota. They also tend to have better managers.

It is true that truly productive companies will choose to continue to grow and simply absorb the cost of regulation. Bill Gates would not keep Microsoft a micro company just because of a labor law. But even the largest companies tend to be quite a bit smaller than they would have been without the added labor costs. On the other hand, there may be many companies that could have employed five or ten more workers, but do not do so in order to avoid these costs.

Our calculations, based on changes in company size distribution, indicate that the additional regulatory costs above 50 workers are equivalent to an increase of approximately 2.3% in labor costs. Part of the incidence of labor regulation is in the form of lower wages throughout the economy and higher unemployment: the lower demand for labor puts downward pressure on wages and employment. In addition, many workers, given the low wages, will choose self-employment or start a small business instead of working in another company, generating entrepreneurs "by force" who are neither qualified nor really willing to be. The result of all this is that the total production of the economy is reduced by just over 3%.

French companies try to mitigate the costs of regulation in many ways. Our research shows that, to avoid crossing the 50-employee threshold, companies just below that number increase average hours worked, the fraction of top-skilled workers (substituting quality for quantity), investment spending, and some cases are subdivided into smaller entities. These distortions mitigate the cost of regulation, but our calculations suggest that they do not cause regulatory expenditures to fall materially.

Our results do not suggest the need to eliminate the regulations we studied. But they do suggest a concerted effort to eliminate the threshold effects that make every step forward more difficult for companies trying to internationalize.

Finally, there is also causality in the reverse direction: threshold effects explain part of the problem of the low productivity of the economy, but it is also true that companies are small in part because their workers and managers are not very productive, as shown by a recent work by Enrique Moral Benito (Growing by learning). And, as he indicates, one of the main candidates to explain this differential is the low level of training of employers and workers. Spain is in the last positions in Europe in the quality of business management (according to several studies by John van Reenen with several co-authors). As for workers, OECD studies on the educational level of adults show that one in three Spanish adults did not reach the minimum level of skills, either written or numerical, one of the worst results in the entire OECD.

In short, we cannot expect Spain to have the salary levels of our environment without its workers having productivity levels similar to those of these countries. To achieve this, not a single change is worth it: it is necessary to rethink the labor relations framework, the regulation by company size, and to invest heavily in improving the qualifications of Spanish workers and entrepreneurs. It is to be wished and hoped that there will be improvements in all these aspects in the coming years.

Luis Garicano is Professor of Economics and Strategy at the London School of Economics, Claire LeLarge is an economist at the Center de Recherche en Economie et Statistique in France, and John Van Reenen is Professor of Economics at MIT.

• What is the optimal company size?
• Why is optimal company size important?
• Why are microenterprises an impediment to improving the productivity of economies?
• What prevents France and Spain from increasing the size of the company?
• According to the theory, what would be the optimal production and prices under perfect competition?
• What are the different market structures that affect their behavior?