There was a lengthy and pointless debate that began in the early 90s over what sized businesses created the most jobs. The original story was that small businesses created the most jobs. This turned out not to be true on more careful investigation, since small businesses also lost the most jobs. There were various twists and turns in the academic literature before most economists came to the conclusion that businesses of all sizes on net create jobs at roughly the same rate.
While this debate kept many economists employed and no doubt helped to boost wages in the profession, it did little to advance our knowledge of the economy. Unfortunately economists learn little from such experiences.
We now have the sequel to this silliness with the claim that it is new businesses that create jobs. This claim emanates most prominently from the entrepreneurially oriented Kauffman Foundation. It was picked up in an Ezra Klein column yesterday. The argument coming from this direction is that all the job growth in the last three decades came from new businesses. Employment in firms that existed in 1980 has just stayed roughly even.
The reason this claim is silly is that the decision of a corporation to expand and open a new division depends to a large extent on the ease with which new businesses can form. If it is easy for start-ups to form, then existing businesses will be less likely to expand their operations by setting up a new division. If they want to get into a new area, then they will just buy a start-up that looks promising.
The logic here is simple. The vast majority of start-ups will fail. However if there are a large number, then certainly some will succeed. The ones that do can then be purchased by existing companies that want to expand. The new jobs can then be attributed to start-ups and not existing businesses, but this is entirely due to the fact that we make it easy for start-ups to form.
An example of this story is Google's acquisition of Youtube. The fact that there was a successful start-up that Google could buy made it easier to enter this market. But does anyone think that Google would not have moved itself in a similar direction had Youtube not existed?
The point is that even if we accept that all net new jobs came from start-ups it does not follow that we necessarily want to do more to encourage start-ups nor adopt any policies that have a negative impact on existing businesses to favor start-ups. The reason why the former is not true is that the vast majority of new businesses fail within a decade. It is reasonable to assume that the marginal start-up (the ones we encourage with our new more start-up friendly policies) will be less successful on average than the current group that did not need this extra boost from the government.
This means that we will possibly be encouraging millions more people to take their life's savings, work ridiculously long hours, usually dragging in other family members, in pursuing a venture that will fail. We will then see the person without a business, without savings and without a job and just a few years left to retirement. That doesn't sound like good policy, nor is it a good use of the economy's resources.
On other side, suppose we tilt the playing a field a bit to favor new businesses at the expense of existing businesses. Well, if we accept the Kauffman analysis, imagine that instead of holding their own existing businesses had lost 5 percent of their jobs over the last three decades. That would give us a really big hole. Would the additional tilt to new businesses fill this gap? We don't know -- at least the Kauffman data don't answer this question.
The long and short is that new businesses are wonderful, but policies that go overboard to push people to start new businesses are likely to ruin many lives and lead their promoters with lots of egg on their face.