One of the few uncontroversial statements in economics is that long-run income growth is only possible as a consequence of the development and application of new knowledge. To continuously raise incomes over long periods of time, it is not enough to simply bring more resources to bear in the production of goods and services. You must also or instead develop ways to do more with less, and that necessarily involves learning and discovery.
To be more specific (and possibly to begin chipping away at our consensus) the speed limit on growth in per capita incomes is determined by the pace at which a society deploys useful applications of new ideas. That is, it’s not enough to come up with a brilliant new idea. The brilliant new idea must then be translated into practical mechanisms which allow people to do more with less. Those mechanisms must then be deployed across an economy in order to generate the capacity for faster growth in per capita incomes—which might nonetheless go unrealized thanks to other constraints on economic activity, like chronically weak demand.
Which of these components of the innovative process is most important? It’s kind of a weird question to ask, like wondering which of the wheels, engine and fuel is most critical in getting an automobile moving. Yet when it comes to research on the topic of “innovation”, economists tend to focus disproportionately on the way in which incentives facing innovators or innovative firms affect their innovative output—which as a general but unsatisfactory rule is measured by patents and patent citations. So you very commonly see research results showing, for example, that higher income tax rates, on both the personal and corporate side, are associated with less patenting. Or, more egregiously, papers showing that if you assume that ideas drive growth, and that the reward for coming up with new ideas is a high income, then you discover that for the good of everyone it is best if tax rates on top incomes are very low. To be clear, there are other lines of economics work which focus on very different things: like the factors (including income inequality) which influence the supply of entrepreneurs, or the obstacles to diffusion of new technologies and techniques.
But more often than not, economic discussions about innovation tend to presume that incentive structures are highly important in influencing the pace of innovation, and that inequality is in some sense a necessary component of innovation and thus of long-run productivity growth. Now, I don’t doubt that incentives matter, and that the prospect of a high income is a motivating factor for many inventors and entrepreneurs. But when we think about the application of those ideas to the much broader process through which an economy raises its productivity through the exploitation of new ideas, it’s really not clear how much they matter, or whether they’re an issue of primary importance at all when it comes to trying to achieve faster growth.
It’s useful to think about actual periods of rapid growth in productivity and incomes. Take the quarter century after the second world war, for example. How do we account for that burst? Charles Jones, the author of the paper mentioned above about idea-driven growth and top incomes, allows that this era poses a challenge for the notion that high or rising tax rates are bad for innovation while low or falling ones are good. Perhaps low tax taxes encourage innovation, he says, all else equal. But all else was not equal; he speculates that growth may have boomed despite high tax rates because the government was pouring money into basic research. He doesn’t elaborate on the idea, because that’s not the focus of his paper. But think about that! From the 1940s through the early 1960s, the top marginal income tax rate was above 90%! Now I suppose that one could argue that during the 1950s growth in total factor productivity would have been, oh, 8% per year rather than the never-matched average annual rate we actually got, of about 2%, if only tax rates had not been so high. But if that were the case and high taxes were really quite a big drag on innovation, you’d have to anticipate that tax cuts in the early 1960s or early 1980s would lead to a massive jump in growth. Of course, they did not. And so one is left feeling that income tax rates are at best a fourth- or fifth- or sixth-order influence on innovation and growth if in fact they affect it much at all. Or more pointedly: tax the rich all you damn want so long as you plow money into research. Strangely enough, this is not something that one tends to read in many economics papers.
Or what if we turn back to the period between 1870 and 1950, when the American economy caught up to Britain in productivity terms and then proceeded to build an enormous productivity lead over every other advanced economy? How to account for that? Well, consider a classic paper on the experience, by Moses Abramovitz and Paul David. The authors credit America’s massive resource advantages relative to European rivals, massive investments in education and infrastructure to facilitate the exploitation of those resources, labor scarcity and high wages, and a domestic market which was both large and homogeneous (which is to say, less riven by class and inequality than Europe). Again, it is absolutely possible that tax regimes had an effect on innovation and growth, over time and across countries. There’s just no way to arrive at a place where that’s a central part of the story.
Or think about the 1990s. Hard to credit tax cuts for that, given that tax rates went up early in the decade. I think many people see the burst of productivity then as an inevitability: something that was bound to happen once computers were powerful and ubiquitous enough and companies had spent enough time figuring out how to use them to increase efficiency. And maybe it was. But maybe there’s more to learn from that experience than we recognize.
Personally, I think big investments in research, infrastructure and education (all sorts!) are critical contributors to technological progress and productivity growth. But perhaps we can go further than that. Here, for discussion purposes: Two speculative propositions about growth.
Running an economy hot is essential to realizing rapid productivity growth
There is a literature on the role of what economists call aggregate demand externalities in generating technological progress (see this, for instance). Basically: growth in one part of the economy generates increased demand for other sectors in a virtuous cycle of expanding output and incomes. And there is work that’s been done on investment complementarities: when one firm’s investments become more productive than they otherwise would be in the presence of complementary investments by other firms. If you’re producing powerful chips, big investments in capacity and in new techniques are more likely to pay off when other firms are investing heavily in applications for those chips, while others are investing in production of goods that make use of both the chips and the applications. There’s a high-growth equilibrium and a low-growth equilibrium, and it’s hard to move from the one to the other when everyone is struggling in the face of weak demand.
But we can make it simpler. If you want people to do more with less, it helps for there to be scarcity. If you want workers to be more productive, it’s better if firms have to work hard to find labor. When labor markets are tight, we can expect there to be more investment in both labor-replacing and labor-augmenting technologies, as well as in good old fashioned training. Plus, workers may feel more comfortable taking risks, and may be more likely to find good job matches. Abramovitz and David reckoned that labor scarcity mattered. Full employment was among governments’ top priorities in the years after the second world war. And it seems at least a little telling that the one period of rapidly accelerating productivity growth we’ve had in the last half century occurred when the Federal Reserve decided to bet that letting growth rip wouldn’t lead to soaring inflation because productivity growth would rise. It’s not a perfect theory, but I’d take it over the arguments for lower taxes.
Reducing inequality is essential to running an economy hot
The basic logic here is simple. The richer you are, the more you’re able to save. Increases in incomes for poorer households translate into bigger increases in demand than comparable increases in incomes for richer households. It is not a coincidence that interest rates have fallen steadily over the past forty years as inequality has risen. We’ve seen this year, in the effect of the CARES Act, how powerfully an economy responds when low-income households are given a boost, even when the economy is dealing with a major pandemic. Of course, lower inequality alone will not guarantee a faster pace of demand growth; you need economic institutions like the Fed to refrain from choking off demand growth out of a fear of looming inflation. And so an important part of capitalizing on (2) is persuading economic policymakers of (1).
How do you achieve (2)? You can raise taxes on top earners; we know that won’t matter much in terms of its direct effect on innovation but it would compress the income distribution. You could raise the minimum wage, establish wage boards, or raise the real incomes of the poor by providing them with high quality, low cost public services. You could also just give people money. Give everyone a taxable universal basic income and call it an innovation dividend. There are lots of ways to come at this.
More broadly, it seems worth reflecting on the possibility that weak demand and high levels of inequality have seriously constrained growth in productivity and real incomes. I’m sure some people will disagree, but it seems to me a far more plausible account of America’s productivity doldrums than the argument that rich individuals and firms are being taxed too heavily.
Readers: what’s your crazy theory of economic growth which is at least as compelling as “keep marginal tax rates low”?
didn't countries with less inequality e.g Northern European social democratic states also have productivity slowdown?
Large oligopolistic firms should be coerced into higher levels of research&investment. Large monopolistic firms have some of the funding/time horizon advantages of governments, while also having some advantages in terms of their executives' authority over work arrangements and investment decisions being stronger than elected officials' authority over state bureaucracies.
Forms of consumption/investment such as apartments and condos should be legal in many more places.
Much more paternalism about alcohol would be helpful. More beer gardens, maybe, and 1-3 beer limit drinking at corner stores but way less liquor. Maybe tax liquor heavily to fund after-school programs or something
Much more mask-wearing to reduce flu, cold, and other sicknesses
Revisit some of the regulation of stocks that the US did after the Dotcom bust. Maybe firms are going public too slowly now.
Policies that encourage more family formation, child-rearing.
Having a big military budget but not killing a bunch of people.
Aussie-style immigration system but weight it towards bringing in a bunch of Chinese and Pakistani women to shrink our enemies a bit
Protectionism
Giving everyone medical insurance via government should improve general creditworthiness if medical problems are as big a cause of bankruptcy as is often claimed
Subsidize coffee and black tea more
Try to disperse poor populations by, say, issuing many more section 8 vouchers and making rules limiting concentration
More regulation of entertainment media, internet media, porn. Why should there be rap radio stations if the highest quality rap music promotes violence, criminality, emotional abuse, and narcissm? Why should there be porn if it's addictive, distracting, and unnecessary for masturbation? Social media depressed and distracts people, too. More regulation of machine gambling. Youtube, Facebook, etc should be letting people know about good job opportunities