Ian Hathaway is founder and managing director of Ennsyte, a San Francisco-based economics and research consulting firm. He is also a visiting scholar at the Federal Reserve Bank of Cleveland and an advisor to early-stage technology startups. His current research focuses primarily on technology, innovation, entrepreneurship, and economic growth.
Here are edited excerpts of our recent chat for my Ricochet Money & Politics podcast on the decline of US entrepreneurship:
Now you’ve written a number of papers on innovation, and entrepreneurship, and what’s going on in the U.S. economy, but I want to start with a paper you co-authored. It came out this summer. It’s called “The Role of Entrepreneurship in U.S. Job Creation and Economic Dynamism.” And let me just read a couple of sentences from that:
The United States has long been viewed as having among the world’s most entrepreneurial dynamic and flexible economies. It is often argued that this dynamism and flexibility has enabled the U.S. economy to adapt to changing economic circumstances and recover from recessions in a robust manner. But evidence along a number of dimensions and a variety of sources points to a U.S. economy that is becoming less dynamic – of particular interest are declining business startup rates and the resulting diminished role for dynamic, young businesses in the economy.
What about Google, Facebook, Twitter, Uber, Airbnb? How can you say that there’s a decline in entrepreneurship and business dynamism when we see all this – you know, all this sort of good churn and creative destruction happening if you just pick up the business pages?
The numbers don’t lie. The data show declining business dynamism overall, which is largely a function of the declining firm formation rate, what we’re calling loosely the entrepreneurship rate. These are very clear trends when looking at economic data over the course of several decades.
Now, you point out the examples of getting pushback because of what’s happening in the technology sector, in particular – and, of course I hear that often myself. I live in San Francisco, sort of at the heart of all of this dynamic entrepreneurship and growth.
But first I would say that’s a very small portion of the economy. And secondly, I would also point out that I think people are just more aware of technology these days. technology has been around for a long time. The technology sector experienced substantial growth throughout the ’80s and ’90s, but it’s just that these products like Google and Twitter and Facebook and so on are just very personal and they’re consuming, you know, a larger portion of our personal lives and our day.
So I think people are just more aware of the small set of firms that are – that haven’t been around for that long. They’ve been around for a decade or half a decade or so but have grown substantially over a very small period of time.
So your key finding is that here aren’t as many startups as there used to be?
So if you look at the startup rate, which we’re saying is – If you look at the number of freshly launched firms in a given year and you take that as a share of all firms, that rate declined from about 15 percent or so in the late ’70s to about 8 percent in 2012, which is our latest data. We actually just had a data release yesterday. So the startup rate has declined by about half over that period.
But that tracks exact same period where we saw Silicon Valley get bigger and bigger. How do you sort of reconcile those two things happening at the same time?
So if you talk about tech again, and in the 1980s in particular, if you look at the startup rate for the technology sector on its own, if you just isolate that group of firms, the entrepreneurship rate was actually surging at that time. And, well, it continues to be much higher than the overall rate, which, of course, includes everything from dry cleaners and restaurants to tech startups. It encompasses the entire private sector economy.
So, actually, during that period, the startup rate in the tech sector was a lot higher but some other research that I’ve done and others have done shows that the tech sector also experiences a sharp decline, but it’s – it’s later. It’s in the 2000s. There’s a marked change around 2000, following the dot-com bust.
Now, of course there’s been a surge in activity in recent years. And, unfortunately what we’re getting with accuracy with our data, which is administrative data that’s calculated by the Census Bureau, and it comes from – it comes from actual tax records from the IRS, so encompassing the entire private sector universe – what we’re getting with accuracy – unfortunately, we have a time lag there. So, as I mentioned, just yesterday, the data for 2012 were released. And we all know that there’s been a surge in tech entrepreneurship, in particular in the last couple of years.
So while I would expect that rate to have gone up in the two years that have passed, the bigger picture here is that there was a persistent decline for three decades. So if we’ve experienced a rebound these last few years, and my guess would be that we have, particularly in certain segments of the economy, but we still have a huge hole to climb out of.
If we’ve seen this decline in entrepreneurship and startups at the same time we’ve seen this explosion in tech startups, then maybe that overall decline doesn’t really matter. Maybe we’ve seen a decline in drycleaners or something, sort of low-tech, low-wage, while we’ve seen a lot more dynamism in the kinds of companies that hire lots of people and create new jobs and new technologies, you know, the gazelles, the fast-growing companies, the companies that are likely to become the next Google.
And if we’ve seen this decline in tech since 2000, isn’t that just explained because 2000 was a peak in the tech bubble?
So I guess my point is that I’d like to see productivity gains from all sectors of the economy, tech or non-tech.
Secondly, there was a recent paper put out this summer by some economists at the University of Maryland and the Census Bureau, in addition to a report from the Bureau of Labor Statistics I believe last year, that shows these high-growth firms create a disproportionate amount of jobs, that their share is also declining. So we’re seeing a decline in high growth entrepreneurship as well.
A related point that I want to make to that is that research that’s been done on who are these high-growth firms, there’s actually a lot more work that needs to be done in this space but it seems a little bit like a random walk. Yes, technology firms are disproportionately likely to be high growth, but they represent by one estimate only 20 percent of the universe of high-growth firms going back a few decades. So the point I want to make is that we really don’t know at the outset if a firm is going to explode and create an entire new industry or significantly disrupt the industry they’re in.
So I guess I would argue, and my coauthor, Bob Litan – I’ll use one of his analogies, just that if we want to score more balls, we just have to fire more shots on goal. So I’m not satisfied with the answer or with the claim that all we need is tech entrepreneurship and everything will be fine.
And the role of the tech bubble in the decline on tech startups?
Definitely some air was let out of the entrepreneurship in tech in the 2000s I expect that future releases will confirm what’s going on now, which is a resurgence in entrepreneurship in the tech sector. And I would also say though that we also know that there’s a lot of consolidation going on in the tech sector. These companies that people are mentioning as “startups,” in quotation marks, are actually huge companies, and they are innovating in large measure by acquiring other companies at earlier stages. So while there has been clearly an uptick in tech entrepreneurship in the last couple of years, I believe there’s also some of this hardening and this consolidation that we’re witnessing in other sectors, it’s also appearing in tech.
You’re talking about a multi-decade trend, when we’ve had Democrats, we’ve had Republicans, we’ve had tax rates which have been kind of high, and they’ve been lowered. It’s been a changing macro-environment.
D.C. policy audience like to latch on to policy explanations, right? You can just pull a simple policy lever and that will fix things. I don’t think taxes are actually a factor in this.
And I guess the only thing I’ll say about regulation is that let’s have the conversation about how specifically does regulation impede firm entry and how does it disproportionately advantage, you know, incumbents versus new entrants? That’s the kind of conversation that I think we should be having.
One example I’ll use from the tech sector in particular – and this has been a hot button issue with immigration – high skilled immigration in particular and H-1B process. There’s a cap on these on these visas, and there’s a mass surge in April to apply and that pool of H-1B visas is exhausted pretty quickly. So you have big companies like Microsoft and so on file hundreds and thousands of applications for H-1B visas. And if they’re allotted let’s say 250, they can say, great. You know, let’s go hire 250 workers. But if I’m a startup, this doesn’t work for me. This process is lengthy, it is expensive, and it’s unpredictable.
So who would you promote more entrepreneurship?
One in the short term is immigration reform. We know that immigrants are twice as likely to launch new firms, and that’s in all sectors, and in high tech it’s particularly elevated, so we know that’s something that will push the entrepreneurship rate up higher.
Longer term, education – it’s one of the factors that in studies of what drives regional variation, entrepreneurship rates, it’s the thing that keeps showing up. And this is at a time when a lot of states have had to cut back on education because of balanced budget requirements and things of that nature. So these are two things that I would advocate for.
Now, going outside of policy for a moment, we’re just kind of kicking the tires on some things that may be linked with the decline in entrepreneurship and as I said, we’re still working on it so don’t hold my feet to the fire too much on this, but I think that there are some explanations that aren’t very exciting.
So, for example, if we look at metropolitan areas and states, there’s a very strong connection between the level of entrepreneurship rate in a region and it being located in the West and the South. Now, we know that entrepreneurship rates increase to a accommodate population growth so it’s very likely that these population – that population growth in the West, particularly in the West and also in some areas of the South had driven entrepreneurship to high rates in the 1970s and early 1980s. There’s also a strong link and decline in entrepreneurship in those regions. So the very reason they’ve had these high rates which, as I said, there’s a geographic pattern to it also on average experienced the largest declines. As local demand grows, businesses have to grow along with it to meet that demand.
So one explanation, which isn’t very exciting, is just that we experienced high growth, population growth in certain regions, businesses needed to be formed to meet that local demand and after that growth slowed down, the firm entry rates maybe came down with it.
Another explanation has to deal with business consolidation. So economic theory would say that the more consolidated an industry is, the higher the barriers are to firm entry. There may be a connection there; others have talked about this. It’s something that we’ve seen in a number of sectors. I have some more research coming out in the coming weeks and months that will address that issue as well.
So we have a population issue going on. We maybe have a business consolidation issue. Anything else?
Starting your own business is really difficult. I mean, it’s – look, everyone has heard about the benefits of working at tech companies, at Google and so on — there’s food and dry cleaning and everything you can ever need. if I have an offer and a real strong salary, why would I even bother to go start a business? It’s just really hard.
And so, in some ways, it’s just that it’s the other side of that equation. How can I compete with Google? How can I outcompete them with innovating a new product if I can create something and beat them, and earn profit for my company, versus the other side of the equation – why would I want to compete with them when I can just go work there and have a cozy life?
So to use the sort of Google example. Why should we really care that there’s been this decline? There seems to be a lot of innovation in the economy. So do you think it’s actually had an impact on economic growth, on innovation, on jobs?
The most clear case is job creation. So we know that without a doubt that young firms are – new and young firms create a lot of jobs. In fact, I would say that they’re the primary source of net job creation. So this is not to say that older and larger firms don’t create jobs, but, on net when you take into account the job that are destroyed from contracting older firms or those that close their doors, on net it’s the young firms that create the lion’s share of new jobs. So that’s a very clear case here. And this has coincided with a period – well, periods of low employment growth.
the other two cases are more interesting and, of course, the research is a little less clear. But there’s innovation productivity. It’s my belief that at least from a couple of lines of research that young firms tend to be more innovative and in particular more likely to create disruptive innovations, life-changing innovation.
I think just our human experience would agree with that, right? So Apple is a significant outlier, but in general the most life changing and most powerful innovations have come from businesses that were formed with the intention of commercializing those creations, those innovations or those inventions.
The second thing – I guess now it’s a third thing – is the productivity side of this. And I think that this is less clear. So, as I mentioned before, although it’s equal, the realloacative process, this dynamism, this churning is really important for productivity. But I also said, on the other hand, it’s been shown that, overall, the Wal-Mart effect has been productivity enhancing.
We’re getting new firms that are generating revenue and have huge stock market capitalizations — but not many people work for them. How many people work for Twitter, Facebook versus the old-line industrial companies? What’s your take on that?
So Nobel economist Michael Spence and UC Berkeley economist Enrico Moretti have two related frameworks for thinking about this. So Michael divides divides the U.S. economy between the segment that competes globally, it sells goods and services around the world versus those that are located only locally and serve local markets. Tradable/non-tradable – Enrico calls it innovative sector versus non-innovative sector. And, basically, what they say is, look, there’s one half of the economy that is going to generate the substantial majority of income growth or value and, at the same time, it’s going to employer fewer and fewer people as a share of that income. And what’s going to happen is that that income is going to get disbursed around local economies and it’s going to support employment. So there’s going to be very little productivity and income growth from that local segment, but that’s where a lot of the jobs are going to come from, which also explains a lot of this – this polarization of wages and job growth in labor markets. And I think that’s absolutely dead on.
Not everyone can work at Google or Apple, I would say that folks that are working in local services and so on – and I’m saying all lines of services. I’m not just saying, you know, barista at Starbucks. We’re also talking about lawyers and doctors who are supporting the local community. I would just say that those people in the Bay Area and D.C. obviously have much greater employment and wage opportunities as a result of those companies that – those innovative companies, those tradable companies that are bringing – that are creating wealth and capturing it from around the country and indeed the world and bringing that to those local economies.
What about more directly trying to teach entrepreneurship, particularly in high school?
So Bill Aulet at MIT has written a great book on this called “Disciplined Entrepreneurship.” Some folks say you can’t teach entrepreneurship. I disagree with that. And I think Bill’s sort of one of the leaders in this space and has a lot to say on that. And my co-author Bob Litan has this great idea of instead of just teaching math and science where students aren’t that engaged, but actually incorporating how that scientific and that mathematical knowledge has been used in the course of business to create goods and services and, teaching the commercial side of this, providing a link to that from young ages so it’s sort of a part of that curriculum and that learning all along throughout the education process.
Follow James Pethokoukis on Twitter at @JimPethokoukis, and AEIdeas at @AEIdeas.