How Do Firms Respond to Size Regulation?

This new NBER Working Paper  documents the growth of the number of French firms with 49 employees.

“In France, firms with 50 employees or more face substantially more regulation than firms with less than 50. As a result, the size distribution of firms is visibly distorted: there are many firms with exactly 49 employees.”

You might think that 49 person firms would be equally likely as 50 person firms but French regulation creates a  discontinuity due to the conscious choice by firms to avoid growing “too big”.

Will U.S firms engage in similar behavior to reduce their exposure to new health care regulation?  If many firms engage in such behavior, does this have aggregate consequences or can thousands of 49 person firms achieve similar economic production as fewer larger firms?

Author: Matthew E. Kahn

Professor of Economics at UCLA.

32 thoughts on “How Do Firms Respond to Size Regulation?”

  1. The federal employment discrimination laws have long applied only to companies above a certain size, which I am too lazy to look up now and have long forgotten since my clients are all well above the threshhold. Maybe there’s a discontinuity there. Certainly it’s a place to look/

    1. Ya – this sort of hard edge has been around for a long time in the US too.
      Personally, I would expect some distortions in distribution around the 40-60 range, but not something giant.

      If your company is doing well enough to hire the 50th worker, it will probably take some extra time to get the paperwork all lined up.

  2. I think this is Obama’s clever plan to rebuild the middle-entrepreneur class!! Lots of people working at 3/4ths speed on their 45-person firms, and volunteering in their churches and running Scout Troops. The Fifties will come back!! and all because of Obamacare and the 50-person limit!!

  3. From the WSJ: “But with fewer than 50 full-time equivalent employees in each half of this business, he hopes to avoid paying the penalties that otherwise could amount to at least $40,000 a year.”

    Obviously that’s a noticeable cost to hiring worker No. 50. At the same time, it doesn’t seem prohibitive if he has serious growth prospects.

    On the other hand, that plant ain’t making its employees rich. $1.6 million in sales with 40 employees? Taking out his rent and input costs and cetera, they would indeed as the Journal says be “low paid.”

    1. By ‘penalties’, does the article mean premiums for employee benefits? It’s not as if the company is to be punished, it’s just that employees are being given benefits. Why is this element of employee benefits considered a penalty when other elements, or basic pay, is not? It all comes out of gross income and is money that the employer can’t take home.

      Is there no benefit to being a better place to work, in attracting employees?

  4. I wonder what the natural limit of firm size, without regulation, is for firms which can get down to 49 employees. It seems to me that once you get to 75 employees, say, it becomes unrealistic to think you can cut back to 49. Actually, I suspect the upper limit is less than 75, maybe 60 or so.

    It would be helpful to look at the histogram showing the numbers of French firm sizes in the 45-65 employee range.

    1. Indeed; you’d likely expect a depression versus the expected distribution curve in the 50-70 range, complementing the spike at 45-49 and together cancelling out.

  5. I would expect any ACA edge effect to be much less sharp. There is still a consensus in the modern American workforce: a “real job” offers health insurance. Employers whose hiring isn’t mostly at the bottom of the labor market know that health insurance is part of the expected package. So to them, health insurance is a good, albeit a good with a cost. In this, it is unlike most other mandates, which confer no particular benefit to the obligor on the mandate.

    Of course, there are some industries–such as retail–who hire at the bottom of the market. They might have to reconfigure.

    1. Hmmm … how many Walmarts would they have to create if they follow this strategy to limit them to 49 employees?

        1. Even universities are doing this. The poorly paid “adjunct professors” who teach so many of the course are now being limited to 29 hours of work per week, which makes their pay even worse. Problem is, even if you decide that these (and so many retail jobs) aren’t “real jobs” they are part of what keeps many people able to eat. When your hours are cut back because otherwise your employer considers you too expensive to hire, you’re worse off than if somebody hadn’t tried to help you by coercing your employer.

  6. This phenomenon is not new and is well-known in France. Most importantly, the paper seems to omit a fairly important regulation for companies with 50+ employees, namely mandatory profit sharing.

    Up to and including 49 employees, French firms can optionally engage in profit-sharing (so-called intéressement); this is not required, though, and subsidized by the government through tax incentives. Starting at 50 employees, profit-sharing becomes mandatory, is based on a defined formula and is now called participation.

    Also, the effect is not that firms “avoid growing too big”. Instead, companies get split into multiple smaller companies when that is less onerous than complying with labor laws and profit sharing requirements for companies that are 50+ employees strong. This is why there are so many firms that “happen” to have exactly 49 employees.

    1. Didn’t you just confirm Matthew’s point? Namely that the intended benefits don’t materialize and that instead you get this awkward organizational structure. Or is your point that it’s the profit sharing regulation that’s the primary driver for this phenomenon?

      1. “Namely that the intended benefits don’t materialize”

        How about we
        (a) ask WHAT those intended benefits were and
        (b) PROVE that they have not materialized
        before we drive this rather rickety contraption constructed of a whole lot more theory than evidence any further?

        1. Also need to discuss whether firms in the 20-49 range are really “awkward”. Many businesses grow larger than they need to be (their optimal size) due to momentum and/or because that’s what their owners think they should do. Taking a pause at 40 and determining a solid strategy for growth isn’t necessarily a bad thing; if the nature of the business is that it should grow to 100, 200, 400… then it will regardless of any perceived regulatory banner.


          In my experience while owners of growing businesses are necessarily quite attuned to cash flow and P&L, no one – large business or small – sits around obsessing with the minutia of what next year’s tax rates or HeritageCare/RomenyCare regulations are going to be (much less the RomneyCare regulations in 2020). There is no way to know, the regulations are going to affect your entire market consistently, and you would simply drive yourself insane trying to do so.

      2. My point is that while I do not disagree with Matthew that size-based regulations can create perverse incentives, it is important to look at the actual incentives to judge their effect and not to simply conjecture that businesses will stop growing once they approach a regulatory size barrier (not all disagreement has to be of the “LOL you’re completely wrong” variety).

        You will note that if you google “france 50 employees” or something like that, you’ll get a great many hits describing this phenomenon. This issue has gotten a lot of coverage not because it is typical, but because it is fairly unique [1]. It also shows that it’s well-known and not something that the paper suddenly discovered.

        The reason why the effect here is so prominent is that in France, companies with 50+ employees not only get hit by pretty onerous regulations compared to companies with less (such as the difficulties involved in firing 10 or more employees for economic reasons within a month), but that other regulations fairly directly affect their bottom line. And thus, given that it is possible to circumvent the law, it’s not really a surprise that French firms do circumvent it when circumvention becomes cheaper and less onerous than compliance [2].

        Most importantly, circumvention even then does not take the form of “ceasing to grow”. You cannot run a business under that assumption; assume that you have 49 employees, and suddenly you need a further employee or two to satisfy increased customer demand. Are you going to hire or tell your customers that you aren’t interested in their business, potentially losing them to competitors? No. If it looks like you’re getting close to a regulatory threshold (be that size or revenue or something else), you start making plans for how to minimize your costs without stopping to grow, e.g. by splitting part of your company off into a wholly owned subsidiary.

        All that said, the PPACA is a completely different beast from French regulations. To be clear, I expect it to have some effect: for example, it is likely to have a substantial effect on American sweatshops. Now, I happen to think that it would be a good thing in general if they couldn’t externalize their social costs anymore and have the general public pay them [3], but I also wouldn’t be surprised if increased operating costs for them would have consequences, up to and including bankruptcy if they rely too much on exploiting cheap labor.

        That said, most reputable American businesses already provide health insurance to their employees; it’s subsidized through a tax exemption, and you’ll have a hard time attracting skilled labor if you cannot provide your employees health insurance. For those businesses, very little will change. And for some smaller businesses, it will actually become easier to offer competitive benefits to it employees, making it in turn easier for them to compete with big firms [4]. I.e., for most regular firms, very little should change.

        [1] This is not to say that other size-based regulations always have zero effect, but the effect is generally much milder or not even observable, as can easily be seen by studying the graph in the paper at the transitions for 10, 11, and 20 employees, where other size-based regulations kick in in France.

        [2] Another example of a size-based regulation that failed was an old German law where maternity leave for businesses up to 20 employees was paid through public health insurance, but larger businesses had to pay themselves. That made it sufficiently unattractive for those larger businesses to employ women that some avoided it, up to and including eating a fine rather than hiring a female employee. Eventually, the German Constitutional Court declared the law unconstitutional for its discriminatory effects and a new law was passed.

        [3] Obviously, just because these employees do not receive sufficient health benefits (or none at all) does not mean that they do not get sick, and in the end, the cost is paid by the insured through ER visits or tax payers via Medicare.

        [4] Even so, big firms still can have an advantage. For example, Google started offering five months of fully paid maternity leave to its employees; not out of the goodness of their hearts, but because it helped them retaining skilled employees, reducing attrition by 50%.

        1. (not all disagreement has to be of the “LOL you’re completely wrong” variety)

          Yeah, I didn’t think you were on that road. Thanks for expanding your thoughts. I agree that looking at the specific context has more explanatory power than attempting to understand the observation by simple “algebraic” formulation. Your sentiments on PPACA seem reasonable to me as well.

  7. Considering that my little firm never got beyond *four* persons, even though we were successful and growing, getting awards and gaining more contracts all the time — but I had to close up shop when I couldn’t find health insurance coverage — I think the 49- person limit is a minor problem.

    Furthermore, I think we will see a small but noticeable jump,in business creation stats when the full ACA goes into effect January 1, as people who have great ideas and business plans, but who are currently stuck in “job lock” because of health coverage through an employer, start out on their own. Since most health-insuring employment is with mature and hidebound firms, or units of government, the innovation rate and entrepreneurship will go up as creative and entrepreneurial employees are released from insurance constraints and start firms or join small, dynamic firms.

    (“Job creators” and ” job creation”)

    1. Betsy,

      Good point.

      It is a mistake to complain about the effects Kahn discusses without also mentioning the benefits of workers not having to worry about health insurance. It’s not just the original entrepreneur who is relieved of the worry. So are prospective early hires.

      1. We’re not supposed to think about the millions of people without insurance.
        Instead, we’re supposed to fret about the outrageous imposition on the rights of business owners
        to provide crappy jobs without benefits.

        1. As opposed to not providing crappy jobs WITH benefits, because it would cost more than the labor they’d be buying was worth.

          I guess it’s viewed as a win-win for the government: Either you keep your job, and are grateful that the government got you more bennies, or you lose your job, and are grateful to the government for the dole.

  8. What is your real point here, Matthew?
    You seem to feel we should be outraged by something: maybe how employers are “gaming” the system by limiting themselves to 49 workers, or (more likely) how the state is suppressing unemployment by making it harder to hire that 50th worker. But I’m having a hard time getting worked up about either of these.

    As is ALWAYS the case when we get this sort of post, the underlying assumptions appear to be totally ignorant of both history and actual human psychology.
    These laws are (for the most part) not in place because society as a whole has a death wish, or because of regulatory capture; they are in place as a response to very real ills of the past. And the cutoff at something like 50 people is not some bizarrely unfair strangeness, it
    (a) limits the social damage that can be caused by the rar(isn) pathological employer AND
    (b) (probably more importantly) in small enough firms, all people interact with each other, and the standard social cues we are born with kick in, limiting just how awfully we will treat each other. Even HITLER extended special protection to Eduard Bloch!

    Yes, societies make it harder to create impersonal massive corporations, the sorts of entities that can (deliberately or just by mistake) easily destroy the lives of many people with a flick of their wrist. Societies make it more difficult for me to get a license to drive an 18-wheeler full of hazardous materials than to drive a car, which is in turn more difficult than a license for a bicycle. I fail to see why this is a problem or anything other than common sense.

    As an argument against laws that protect employees, I’m afraid this is pretty pathetic.

      1. That was me, posting as the Cookieless wonder.

        But as long as I’m back to clarify my identity, let meal so belabor Mr. Handley’s point by noting that societies actually make it artificially easy in many cases to create larger firms, by granting corporate entities endless life, certain immunities, and other artificially created bennies.

        1. “…certain immunities…” Say what you mean, Betsy. They are above the law. Now unpack how that is ethically unacceptable, how swiftly and how comprehensively it corrodes the foundations of civilization, and how there is no incremental remedy for it.

  9. This might actually be a good thing in one way. As any organization (machine) grows, its total internal friction grows faster than the number of components (people, moving parts). “Coase’s ceiling” (Ronald Coase, “The Nature of the Firm”, 1937; incredibly, he is still alive and working at the age of 102!) is the name for the effect at the margin where the organization has grown to a size where essentially all of its potential output goes to overcoming internal friction and there is no longer any real work being done. This is, of course, a qualitative rather than quantitative threshold, but empirical observation implies that it is never higher than about 100 people and often much lower.

Comments are closed.