Wednesday, January 2, 2013

Is 49 a Magic Number?

Copyright, The New York Times Company

The aggregate amount of regulation is difficult to quantify, but we learn something about it from the number of businesses that choose to have exactly 49 employees.

I noted last week that government regulations are not as easily quantified as taxes and spending, because regulation has no budget and no obvious accounting method. Some laws are not enforced, while others have little impact because people would follow them even without the force of a law. The most useful regulatory budget would put a lot of weight on the laws that actually matter.

The economists Luis Garicano, Claire Lelarge and John Van Reenen are developing a method to quantify the aggregate importance of employer regulations. They note that small employers are naturally more common than medium-size employers, which are themselves more common than large employers.

Moreover, the frequency of employers of various sizes appears in many situations to follow a “power law” of statistics. If you tell these economists how many employers in a given region have, say, 22 employees, the authors can, with the power law, accurately predict the number of employers with 23 employees.

The chart below, reproduced from their paper using data on employers in France, shows the number of employers of various sizes. For example, about 10,000 employers in their sample have four employees. Their paper confirms the statistical power law with one glaring deviation: the exceptional number of employers with exactly 49 employees and far fewer with 50 employees (another lesser deviation occurs on the margin between 9 and 10 employees).

Luis Garicano, Claire Lelarge and John Van Reenen, “Firm Size Distortions and the Productivity Distribution: Evidence From France.”


The chart shows a couple of odd patterns at the 50-employee mark. First, there are sharply fewer employees (by more than a factor of two) with exactly 50 employees than with exactly 49 employees. Second, although the number of companies usually falls with the number of employees, there are actually more employers with 49 employees than with 45 employees.

The authors show how this pattern reflects deliberate efforts by employers to stay below the 50-employee threshold where several employment and accounting regulations take effect. For example, they note that French companies employing 50 or more workers are, among other things, obligated “to establish a committee on health, safety and working conditions and train its members,” whereas companies with 49 employees are not. France also has regulations kicking in at employment levels of 10, 11, 20 and 25.

Presumably, companies would not adjust their size to avoid regulations that are not enforced or regulations that require an employer to take actions that he already takes. Moreover, mandates that create costs for employers but create nearly equal benefits for employees should not induce companies to change their size, because (barring minimum-wage regulations) employers can pass on the costs of those regulations to their employees in the form of lower cash wages.

The primary reason for size adjustments is regulations that impose costs on an employer significantly higher than the benefits they create for the employees. The larger the net costs, the more company size should deviate from the statistical power law, which is why Professors Garicano, Lelarge and Van Reenen can use their power-law analysis to quantify the private costs of the regulations that kick in at the 50-employee threshold.

This is not to say that the regulations imposed on 50-employee companies are necessarily excessive, because they can create public benefits that more than justify their net costs for an employer and his employees, just as taxes and government spending can. For example, an air-pollution regulation might kick in at 50 employees that creates a significant cost for the employer and little aggregate benefit for his employees but creates a significant benefit for the people of France.

The authors show how, so far, employer sizes in the United States deviate less from the statistical power law, which implies that French regulations kicking in at 50 are more costly (from the point of view of an employer and his employees) than the United States regulations kicking in at that threshold.

But the United States has added some major regulations with its Affordable Care Act and its Dodd-Frank regulations. Beginning this time next year, for example, the Affordable Care Act will put new requirements on businesses with 50 or more full-time employees, whereas businesses with 49 or fewer employees will be exempt. Businesses with fewer than 26 employees may already be eligible for Affordable Care Act tax credits for providing health insurance, whereas larger businesses are not.

Perhaps we should thank the French for their heavy regulation, as it is already helping us account for the impact of regulation in the United States.

No comments: