The unhelpful lionization of small business

By Felix Salmon
October 24, 2011
Jared Bernstein has the wonky version in the NYT, but Jim Surowiecki has the soundbite:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Jared Bernstein has the wonky version in the NYT, but Jim Surowiecki has the soundbite:

Small businesses are, on the whole, less productive than big businesses, and though they do create most jobs, they also destroy most jobs.

Both of them are making the important point that high-flying rhetoric about the importance of small business is much better politics than it is policy. We’ve been hearing a lot about the individual 99% of late; here’s the corporate 99%, from Bernstein.

New research by the Treasury Department finds that small businesses — defined as those with income between $10,000 and $10 million, or about 99 percent of all businesses — account for just 17 percent of business income, and only 23 percent of them pay any wages at all.

The facts of the matter are stark: larger businesses are more productive (this will come as a shock to anybody who spends most of their life in meetings, but it seems to be true), and they even create more jobs, once you control for firm age. Or, to put it another way: it’s not small businesses which create jobs, it’s startups. Here’s the chart, from this paper by Haltiwanger, Jarmin, and Miranda (HJM):


The statistic that small companies create the most jobs comes from the purple line. But firms mean-revert when it comes to size: as they fluctuate in size over time, they tend to add jobs when they’re smaller, and lose jobs when they’re bigger, and even if they add no jobs overall, that still makes it look as though smaller companies add more jobs.

If you control for mean-reversion effects and look at a firm’s average size, the effect seen in the purple line becomes much less pronounced, and you get the green line instead.

And then look what happens when you add age controls — that is, when you control for the fact that younger companies are more likely to create jobs than older companies. A small family business which has been around for decades is unlikely to be an engine of job growth; meanwhile, large young companies (think Groupon) can hire extremely quickly.

Once you adjust for company age, you get the blue line in the chart — small companies actually lose jobs, on average, and it’s not until you get to about 500 employees that they manage to create any jobs at all.

Here’s Bernstein:

The big story here is that startups—which can only grow at first but which also have high death rates—play an important role in these dynamics. They’re small at first, and many perish—about 40% of the startups’ jobs are lost through firm death after five years. But if they survive, they will generate significant job growth (HJM: “conditional on survival, young firms grow more rapidly than their more mature counterparts”).

This finding and the flip of the lines in the figure when the proper controls are applied have important policy implications. Once we account for the startup effects, small businesses, per se, are not the engines of job growth they claim to be.

The policy implication here is that if you want to create jobs, you’re much better off encouraging startups than you are bending over backwards for small businesses, most of which are reasonably long in the tooth. And the interests here do diverge, although of course there are overlaps.

For me, the most important difference is the degree to which the two groups are reliant on a social safety net. Precisely because most startups fail, the founders and employees at such shops need to be able to know there’ll be someone to catch them if they fall. The success of Silicon Valley can be attributed in large part to a culture where people who have worked and failed at a startup are extremely employable. But on a national level, there are good reasons why we would want to move towards the Norwegian model.

Finally, it’s worth resuscitating this classic Kinsley column from 2008: the fact is that some of the richest members of the 1% are small business owners, while many of the hard-working 99% are in fact large business owners.

Big businesses are not owned by big people, and small businesses aren’t necessarily owned by small people. The typical shareholder in a big business is a worker in some other big business whose pension fund has chosen to invest in that company. Or a retiree who has bought this stock as his or her nest egg. Or it’s somebody’s 401(k).

So the next time a politician lionizes small business owners, remember that you are in fact a large business owner. And that small business, while it sounds romantic, isn’t nearly as important as the political rhetoric tends to make it out to be.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Your quote from the Kinsley column is misdirection, as you’re intentionally confusing publicly held companies with hundreds of millions in shares with a small private company that is owned by one or a few people.

Despite that, it’s a valid point. I have created several LLCs over the years, and they will never employ more than me. But it is fair to say that small businesses are the most dynamic, and best reflect the concept of both job creation and destruction. And large mature businesses certainly have an anecdotal record of mostly destroying US jobs – IBM, GM, GE (home of our jobs czar), to name a few.

Posted by Curmudgeon | Report as abusive

“larger businesses are more productive”

Don’t confuse economies of scale with being more productive. Large companies have more overhead, more management, and more rules than small companies, so I’d like to see some of those facts of the matter (evidence) about them being more productive.

Big companies like to boost profits, and when they can’t grow revenues, they cut expenses, which usually means jobs (that’s how they increase that totally meaningless productivity stat, revenue divided by payroll). New companies (and even small ones) can usually only grow profits by growing revenues, and to do that, they usually have to add employees. Both classes of companies do what they need to survive and neither should be placed on a pedestal, but that is virtually guaranteed to happen when politicians look for demographics to pander to.

Posted by KenG_CA | Report as abusive


you’ve replaced one form of confusion with another by suggesting a difference between “start-ups” and “small business.” Most start-ups are started up to become the small businesses that employ few people, are low on the productivity scale and survive for only a short time. Only a a small percentage of start-ups intend to grow to any significant size and only a very small percentage of them succeed.

The policy questions here aren’t about start-ups vs small businesses. They are:

* Can anything be done to encourage the forming of more start-ups which intend to grow?
* Can anything be done to increase the success rate of start-ups that intend to grow?

Right now the answers are: “We have no idea” and “History suggests, ‘no’”

Posted by timothyogden | Report as abusive

TimothyOgden, I disagree with your answer to your first question. There are lots of things we can do to encourage the forming of more start-ups, but they are rejected because they are viewed as “industrial policy” and picking winners. The recent bankruptcy of Solyndra, which received over $500M in government guaranteed loans, will make it even more difficult for governments to create policies and an environment that fosters the creation of more start-ups.

Politicians usually focus on the complaints of executives from large businesses about what they perceive to be obstacles to creating new jobs, but these are rarely the same impediments for starting new businesses. They demand lower taxes and less regulations, claiming this will help job creation, but those are not the kind of obstacles would-be founders face. As someone who has started or invested in more than a dozen start-up tech companies, I can say that those two issues were NEVER a factor in the companies success or failure.

If we want to help more companies be formed, we need to ensure a good supply of capital, talent, and infrastructure. That, unfortunately, isn’t going to happen in the current political climate.

Posted by KenG_CA | Report as abusive

There are both pros and cons of small vs large. Contrary to popular beleif large companies pay much better wages and benifits than small ones do for employees. My brother works at Home Depot and they pay only half of their employees a livable wage and offer a weak benifits package. Contrast that grim picture with the reality that the locally owned hardware store pays non-owner family employess even less and offers no benefits at all.

Large companies also pay a much larger share of their revenues in income taxes. This is also contrary to public opnion. Litterally millions of small businesses scrape buy ever year showing little to no net income, yet they allow their owners to take out substantial cashflows due to the accelerated depreciation of financed equipment.

The area where small busineses come out ahead is the distribution of income. One Home Depot displaces 5 local hardware stores. The 5 owners all made about what the 1 home depot store manager makes so in the most oversimplificed example there is a 75% drop in upper middle class incomes when large businesses displace small businesses.

We need both small businesses and large businesses to thrive for the economy to recover. Felix is spot on though that small businesses are put on a pedistal while large businesses are vilified. The reality is somewhere in the middle.

Posted by y2kurtus | Report as abusive

“The typical shareholder in a big business is a worker in some other big business whose pension fund has chosen to invest in that company. Or a retiree who has bought this stock as his or her nest egg. Or it’s somebody’s 401(k).”

More Kinseyian B***S**; you should know better, Felix. The 401(k) shareholder is no more an “owner” of the firm than Curmudgeon’s LLC is NFLX–er, Olympus–er, GM–er…GRPN?

The few hundred–or few thousand shares–are given, at the discretion of the real owners, some scraps from the table in the form of dividends. Or not, if the stock price is appreciating, or at least not dropping, so much as the rest of the market.

The choices of a 401(k) are choices of who might give you the most crumbs. And even if all of those 401(k) shareholders were aware, and worked together–which they won’t–they still hold too little of the company to be able to fix it, the way an owner would.

The great myth of the past thirty years has been the “ownership” society actually encourages–or even allows–small investors to have some “control”: over their investments (generally limited choices offered by the company, including the horrible autocorrelation of company stock), over the way firms are run (look! I filled out a proxy statement telling DIS I don’t want them to give away 1/3 of the company to “go”! Certainly they listened!), and over their own retirement (well, your investments lost 40% of their value–and you cannot write off any 401(k) losses on your tax form, the way real investors may–but look at the recovery the market has made while you’ve been eating dog food and being foreclosed!).

The 99% are more dependent than ever on the 1%. And Kinsey likes it that way; pretending slaves are owners requires overseers, a job he’s perfectly happy to do.

Posted by klhoughton | Report as abusive

Commuting distance, time and rising cost of employees to reach the bigger employers must be having its effect on employees of the big box stores and other larger firms.

I live in a rural area where all the major big boxes are located in two of three regional cities that have small to middle sizes populations (30,000 to under 200,000). From where I live, commuting distance to the nearest is about a 50 mile round trip. To the farthest, it is about 150 miles RT. That eats into a low wage job quickly.

The town I live in is comprised of mostly small businesses but they are accessible within ten miles. There are only two large employers. One is a chain supermarket and the other is a subsidiary of a European manufacturer. It pays starting wages somewhat higher than locals, and as Felix notes, the locals are not large employers or good paying either. The super market put several smaller groceries out of business over night. The big manufacturer pays some benefits but isn’t generous. I think only management gets any benefits in the supermarket and the pay for most in near minimum wage. The local small businesses seem to hire mostly part timers.

There are several small scale manufacturers within a fifty mile commute south of here. Some of them have been around for decades and some for at least 50 years, I think. And there are several smaller manufacturers of lumber products, sand and crushed stone, some pre-cast concrete products and related building materials in the next town.

What difference does it make to lionize small firms of not? The fact on the ground say that the small employers are more numerous but no amount of praise is going to make them grow. But now, if either of the two big employers go under, the lower wage employees and a large part of the town would be out of luck. The town would have to live of the property tax from some well-heeled homeowners. Its small “downtown’ intersection is made of of four building and only one is occupied.

If fuel costs go up dramatically, the economic sense or even the possibility of working for the distant big firms is going to wither and there will be no room for those job seekers locally. Even the big town employers could face problems with long distance employees who will find their budgets erode just to get to work.

This country was designed for the private automobile and cheap fuel. High fuel costs will have disastrous consequences, especially for outlying suburban and rural areas. And there will be no cheap fuel no matter what anyone does. Cheap fuel makes the distances between towns and cities seem smaller and expensive fuel will have the opposite effect. There are no “emergency management” plans for that problem because I doubt anyone has given the possibility that fuel costs could cripple the country gas could much thought. All the wars in the ME are doing is transferring the costs of fuel security to the federal level and raising the price at the same time.

A possible RX for this difficulty would be to adopt some of the principles of what has been called “the New Urbanism” that tires to create mixed-use neighborhoods that incorporate residential, employment and institutional uses within walking distance. But the real estate collapse will make that a difficult fix now. Many of those most in need of more sensible housing within an easy commute of employment, will not be able to sell their underwater homes without a loss.

Posted by paintcan | Report as abusive

whoah this blog is excellent i love studying your articles. Stay up the great paintings! You know, many people are hunting around for this info, you can aid them greatly.

continuously i used to read smaller posts which also clear their motive, and that is also happening with this paragraph which I am reading at this time.

Political deaths aren’t any anathema to Kerala, however killing in Chandrasekharan had been a resonating logo from the collected consciousness in the community. She seemed to be stabbed in the face 49 periods, additionally, the destroying seemed to be carried out not even by motivated cadre nevertheless by appointed murderers. Thus lie the tentacles associated with a unsettling communal trend.