Robert Solow on 'Why Economies Grow'

As a follow-up and companion piece to my previous post, I decided to publish a transcription of a lecture on economic growth by Robert Solow that I transcribed originally as an aid for friends and colleagues who were studying economics. Although the lecture was given by Prof. Solow a few years ago during the height of the financial crisis, it contains loads of timeless insights, some of which is useful to be reminded of in the current situation, as discussions about the output gap resume in the next few years (see chart).

However, it's extremely important to keep in mind that in our current predicament as a result of covid potential GDP will also likely take a huge hit, as businesses and employees require some catching up in terms of business practices (misaligned with changing consumer preferences) and job training (due to skills entropy from employees being on furlough), to name only a few aspects that are likely to be impacted. In many ways, the post-covid period will bring us back to the type of economic analysis that used to occur a long time ago when natural catastrophes had significant and frequent impacts on economies' productive capacities.

The video of the lecture is included down below, though the sound quality is very bad, which is why I recommend reading the transcription instead (and you'll get through the transcript much faster by reading it).

Key insights are highlighted in bold font. Enjoy!

The business of this course is the long run. What are the sources of economic growth in the national economy or in the larger economy? Where does growth come from? And the policy implication – well, not implication, but policy question – is ‘How do you get an economy to grow rapidly and to have that growth widely shared in the nation?’
But there is a problem – it is a problem that appeared in the slides that Prof Newstone showed. It is a problem about getting there from here. So I’m going to start by talking a little bit about right now – this is not going to be the usual stuff about the financial crisis and all that – I have something else in mind.

There is something very odd about our economic situation in the US today. I read just recently an estimate from the Federal Reserve that about $7 trillion worth of wealth has been destroyed in the last year or year in a half (in 2008-2009). The country, so to speak, is $7 trillion poorer than it was.

When I wasn’t having a conversation with Cathy in the car, I was trying to divide 7 trillion by 300 million--the population of the US--in my head. It comes to about $23,000 for every man, woman and child in the country. Some, of course, have lost more, some have lost less.

What I want to point out is how strange that is: $7 trillion of wealth has gone down the drain but the productive capacity of the US economy – the capacity of our system to produce goods and service for its people – hasn’t diminished at all. In fact, it is undoubtedly higher than it was a year ago or 18 months ago: the labour force is a couple percent larger, the skills and education and training of the population is certainly not deteriorating and have probably gained. The net investment in capital has been positive – it’s been declining – but has been positive.

So we have a bigger stock of productive capital in the economy now than we did a year ago or 18 months ago. So the productive capacity of this economy is bigger than it was, despite of this $7 trillion of disappearance of wealth. If you are thinking of buying the US economy as a gift for your boyfriend or girlfriend, it would be worth just as much as it was worth – you know, like a used car – it would be worth just about as much as it was worth a year ago.

So in that sense we haven’t lost anything at all. But, of course, the point is we are in a recession. It is one year old according to pundits. And according to other pundits, or the same pundits, it’ll continue for at least until the second half of this year and maybe beyond. And the point is we are not using the productive capacity that we have.

You saw the unemployment numbers that Professor Newstone showed you. It is a lot harder to measure excess capacity in industry than it is to measure unemployment, but there are such figures, and they show an increase in unused capacity. So we have this machine for producing the goods and services for the population and we are not making full use of it. And that under-use of economic capacity, of productive capacity will go on for a long time. Even if the economy turns up in the second half of this year we will undoubtedly finish 2010 still with some slack in the economy because the slack disappears only gradually. 

So if you are interested – now, this is the point, this is why I started this way – if we are thinking about the long run growth of the economy (which means the long run growth of its capacity to produce), it’s not a separate but it’s an analytically slightly different problem to make sure that that capacity is used.

As long as we are not using all of the capacity that we have, the economy and the decision-makers in the economy are not likely to be motivated to do the things that increase potential output, that increase the productive capacity very rapidly.

So the short-run order of business – policy business – for us and every other rich country in Europe or Asia right now is to close that gap or narrow that gap between productive capacity and actual output, which means fundamentally trying to increase the demand for goods and services. And to do that in a way that at least doesn’t create obstacles to the long-run growth of the economy once the gap is closed, and maybe does some things that will help it.

So, imagine it is now January 2011 and the American economy and the economies of the other rich countries – developed countries of the world – are prospering reasonably well, are using their capacity, have closed that gap. Then the question is: What makes them grow? What economic activities that take place have the effect of increasing the capacity of the economy to produce useful goods and services? 

Now, you won’t be surprised – in fact, I’m staring at this monitor here and it says: so what determines the rate of economic growth in the economy? And that’s the question that I want to come to now, and it becomes relevant after we have done the short run task of closing that gap. There isn’t any one word or two word answer to that question. 

And I should make it explicit that I am thinking now about what determines the rate of economic growth in a rich economy, in an advanced industrial economy. I am not thinking about developing economies where the answers are related but the answers are somewhat different.

And the truth is that for an advanced economy the answers to that question – what are the sources of growth of national output, of productive capacity – are really the usual suspects. They are things we have known about now for quite a long time. And basically, what matters is what you might describe as investment in a very broad sense. I have to emphasize “in a very broad sense”.

What increases the productive of an economy like ours is investment in physical capital, in machinery, in computers and all the rest of that, investment in what economists call human capital, meaning skills and capacities of workers and people who work in the economy, and investment in new technology.

And here there is a slight difference between the US and even most of the countries in Europe. Not quite across the board but in most branches of industry the US is the technological leader. The gap was very big at the end of the Second World War and has closed considerably. But still, if you look at sector by sector, with some exceptions, the US is the technological leader.

Other countries of the world, that were even fairly rich countries have the luxury of being able to acquire technology by innovation, essentially by adopting, using what is already known. This country (i.e., the US) is in the position of having – so to speak – to invent its own future.

So basically, if we are looking now at the US, the things we have to look after in order to have a successful fairly high rate of growth (we can talk about the equity issues later) are a high rate of savings and investment in plant and equipment. I’d rather have the saving done here than abroad so that, in effect, the capital equipment that is built by investment in this country is owned in this country, and the returns to it stay in this country. It’s not necessary but it’s probably desirable. 

We need an extraordinary amount of emphasis – and we’ll talk more about this later – on investment in human capital, on producing the labour force that has the skills that are necessary to successfully operate that plant and equipment. And that is especially important because a country like this also has to invest in new technology. There is no place it can copy from – it has to in most cases create it itself.

Now, when I say new technology, the phrase tends to have a “high tech” air about it. But I don’t mean it that way.  New technology needn’t be high tech. It turns out that – in many ways – the most important contributors to productivity in the US over the last decade or two have been the application of information technology to wholesale trade, retail trade and financial services.

In fact, there are studies trying to understand why the major, big European economies, Germany, France, UK and Italy have lagged behind the US in productivity terms, general productivity terms. And the common answer seems to be that they have been slow to adapt the information technology to the service sectors. In manufacturing, there is very little gap, if any. But the gap is in the service sectors. 

So, this is extremely important. And I want to emphasize it, even at the risk of some repetition. One of the standard, valid, almost universal generalizations about the way people behave economically is that technically the income elasticity of the demand for services is high. All over the world, as incomes rise, personal incomes rise, people want to spend, [and] choose to spend a larger fraction of that income on services rather than goods. And you can understand why that should be so.

So this means that most of the rapidly growing advanced economies grow more rapidly in the service-producing sectors than in the goods-producing sector. There are exceptions to that. A country like Germany – to a lesser extent Japan, or formally Japan, not so much anymore – has a strong bias toward trying to make its living from simply exporting high quality manufactured goods. You notice I said exporting because the population of Germany, like the population of anywhere else, wants to consume services as it gets rich, not goods.

So those are the things, the essentially important things that a country like the US needs to do to generate long-run growth of productive capacity. 

I should say, in terms of policy, that you should beware of any universal advice like “well, the market will take care of that”. You know, if the alternative to the free-market economy is some kind of central planning, there is no question to where the advantage lies. But there is absolutely no evidence in the historical record of the advanced economies that zero regulation or weak regulation of industry is somehow conducive to rapid growth, or that minimal involvement of the government in the economy is conducive to rapid growth.

The functions of the government in terms of long run growth are just what you would deduce from what I have already said: promoting research and development, providing incentives for investment when they are lacking, taking care of education, and looking after mobility. By the way, it is probably also true that a country – there is less evidence for this generalization, but it’s probably also true – that business cycle instability is bad for economic growth.

For countries that are given to wide fluctuations like the ones we were looking at a few minutes ago, that’s not helpful for long-run growth because it adds to uncertainty. The likelihood of broad fluctuations adds to uncertainty is bad for all forward looking activities, like investment, like mobility, like education.

I wanted to say one more thing about the issue of mobility. When I say mobility, I mean industrial mobility and occupational mobility. In a rapidly growing, technologically-based economy, people have to change the nature of their jobs frequently and capital has to flow freely from obsolescent industries to new industries.

It is very important when you come in this course to talk about issues of equity. I think it is very important to find ways so that the burdens that are associated with necessary mobility don’t fall on workers and other people who are ill-equipped to prepare them [for that eventuality].

Dislocation and sometimes dislocation is probably an inevitable part of fast, mainly technologically-based growth. But it is the task of economic policy to find ways of combining that with income security, up to now, where it’s mostly below the median for incomes.

The macroeconomics of Robert Solow: A partial view

During a hearing before a Congressional Committee on Science and Technology, Robert Solow (MIT) described himself as a "generally quite traditional, mainstream economist".

In my view, either Prof. Solow is unaware of who qualifies as a "traditional, mainstream economist" these days or the definitions of the words "traditional" and "mainstream" need to be completely changed!

Consider, for instance, his views on the notion of "expansionary fiscal consolidation":

[H]ow does a human race with limited intelligence...deal with situations in which the short run need for policy are quite different from the long run need for policy? The feeble-minded, it seems to me, attempt to solve this problem [by asserting] that fiscal consolidation is really expansionary in the short run. I have never been able to understand the mental processes that underlie that statement. But I will take it seriously only -- only -- when its protagonists faced with a situation of clear excess demand propose fiscal expansion. Because if fiscal consolidation is expansionary then fiscal expansion must be contractionary. I don't believe that would happen. So I don't take that argument seriously at all. I think it's cooked up to make a real difficulty go away. (The Feasibility of European Monetary and Fiscal Policies: Rethinking Policy from a Transatlantic Perspective)

...on the supposed lack of microfoundations in Keynesian economics:

You know, there is something a little ludicrous in the belief that microfoundations for macroeconomics were invented some time in the 1970s. If you read Keynes's General Theory or Pigou's Employment and Equilibrium (or many lesser works) you will see that they are full of informal microfoundations. Every author tries to make his behavioral assumptions plausible by talking about the way that groups or ordinary economic agents might be expected to act...But you can recall Keynes's argument that the marginal propensity to consume should be between zero and one, or his discussion about whether the marginal efficiency of investment should be sensitive to current output or should depend primarily on "the state of long-term expectations". Those are microfoundations. (2004, p. 659)

...on the claim there is a connection between the money supply and price level:

[T]he financial press sometimes writes as though there is some special direct connection between the money supply and price level. So far as fundamentals are concerned, monetary policy works through its effects on aggregate nominal demand, just like fiscal policy, in the long run, too. The only direct connection I can think of is itself the creation of pop economics. If business people and others become convinced that there is some causal immaculate connection from the money supply to the price level, completely bypassing the real economy, then the news of a monetary-policy action will generate inflationary or disinflationary expectations and induce the sorts of actions that will tend to bring about the expected outcome and thus confirm the expectations and strengthen the underlying beliefs. (1998:4)

...on the problem with Milton Friedman's reliance on correlations between and M and other variables to infer policy conclusions and the assumption of an exogenous money supply (with John Kareken):

The unreliability of this line of argument is suggested by the following reducto ad absurdum. Imagine an economy buffeted by all kinds of cyclical forces, endogenous and exogenous. Suppose that by heroic, and perhaps even cyclical variation in the money stock and its rate of change, the Federal Reserve manages deftly to counter all disturbing impulses and to stabilize the level of economic activity absolutely. Then an observer following the Friedman method would see peaks and troughs in monetary changes accompanied by a steady level of economic activity. He would presumably conclude that monetary policy has no effects at all, which would be precisely the opposite of the truth. (Karaken and Solow, 1963, p. 16)

...on choosing the right model in macroeconomics:

[I] believe rather strongly that the "right" model for an occasion depends on the context --  the institutional context, of course -- but also on the current mix of beliefs, attitudes, norms, and "theories" that inhabits the minds of businessmen, bankers, consumers, and savers. (2004, p.xi)

...on the problems with the DSGE model:

I do not think that the currently popular DSGE models pass the smell test. They take it for granted that the whole economy can be thought about as if it were a single, consistent person or dynasty carrying out a rationally designed, long-term plan, occasionally disturbed by unexpected shocks, but adapting to them in a rational, consistent way. I do not think that this picture passes the smell test. The protagonists of this idea make a claim to respectability by asserting that it is founded on what we know about microeconomic behavior, but I think that this claim is generally phony. The advocates no doubt believe what they say, but they seem to have stopped sniffing or to have lost their sense of smell altogether. (For more, see here and here)

...on the difference between budgetary and real resources costs:

The trouble is that the great world -- including a large part of the intellectual world -- has lost sight of the fundamental difference between budgetary costs and real resource costs. An unemployed worker and an underutilized or idle plant is not something we're saving up for the future. Today's labor can't be used next year or the year after. And the machine time in a plant that's down can't be redone two years from now or three years from now. Three years from now we hope that the plant will be running for current uses. So there's that important sense in which idle resources are almost - and maybe literally - free to the economy. The problem is to get them used in a reasonable way. (see 22:00 here):

...on the importance of fiscal policy for stabilization purposes:

I start from the belief that non-trivial imbalances of aggregate supply and demand do occur in modern industrial capitalist economies, and last long enough that public policy should not ignore them...When such imbalances occur, fiscal policy is a useful tool. The single instrument of monetary policy can not do justice to the multiplicity of policy objectives; and the Ricardian equivalence claim is in practice not nearly enough to convince a realist of the ineffectiveness of fiscal policy. The real obstacles to the rational conduct of fiscal policy are the uncertainties about the proper target for real output and employment, and the tendency for stabilization goals to become inextricably tangled in and distracted by distributional and allocational controversy. (Is fiscal policy possible? Is it desirable?, p. 23)

...on the long run potency of deficit spending financed by bonds vs. deficits financed by money creation, and on the contractionary nature of open market purchases of government bonds (with Alan Blinder):

[N]ot only is deficit spending financed by bonds expansionary in the long run, it is even more expansionary than the same spending financed by the creation of money. [Foonote: An interesting corollary of this is that an open-market purchase, i.e. a swap of B for M by the government with G unchanged, will be contractionary!] (Blinder and Solow: Does Fiscal Policy Matter? 1973)

...on how statements by a central bank can influence how the public translates relative price changes into expectations about the consumer price index:

There are various interest groups in the economy: bankers, investors, savers, lenders, borrowers, buyers and sellers and what not. There is no reason for them to react in the same way. How does one aggregate expectations?

...on the use of "expectations" to explain macro policy outcomes:

[T]o rest the whole argument on expectations -- that all-purpose unobservable -- just stops rational discussion in its tracks. I agree that the expectations, beliefs, theories, and prejudices of market participants are all important determinants of what happens. The trouble is that there is no outcome or behavior pattern that cannot be explained by one or another drama starring expectations. Since none of us can measure expectations (whose?) we have a lot of freedom to write the scenario we happen to like today. Should I respond...by writing a different play, starring somewhat different expectations? No thanks, I'd rather look at data. (1998:93)

...on the claim of self-correcting markets and the role of aggregate demand in causing output fluctuations:

Capitalist economies do not behave like well-oiled equilibrium machines. For all sorts of reasons they can stray above or below potential output for meaningful periods of time, though apparently they are sightly more likely to stray below than above. Even apart from considerations of growth, macro policy should lean in the general direction that will nudge aggregate demand toward potential, whenever a noticeable gap occurs. The relevant point is that this strategy is also growth-promoting. Whatever the level of real interest rates, excessively weak aggregate demand -- and the prospect of weak and fluctuating aggregate demand -- works against investment. Few things are as bad for expected return on investment as weak and uncertain future sales...Successful stabilization contributes to growth too. (Role of macroeconomic policy, p. 301)

...on the need for public policy to address the unemployment of unskilled labor:

It needs to be insisted that the root of the problem lies in the enormous range of earning capacities generated by the interaction of modern technology (and other influences on the demand for unskilled labor) with the demographic and educational outcomes on the supply side of the labor market. There is no really good way for a market economy to deal humanely with that spread. (Too Optimistic)

...on the fallacy of self-correcting markets and the limits of monetary policy during deep recessions:

One important lesson that I hope we have learned from the crisis and the deep recession still going on is that economies like ours can experience uncomfortably long intervals of general excess supply or excess demand. Of course, we -- economists and interested civilians -- used to know that. But it was widely forgotten during the Great Moderation and the accompanying optimism among economists and civilians about smoothly self-correcting markets. The general belief than was that monetary policy was an adequate tool for taking care of any minor blip. During long and deep recessions, however, it has become evident that monetary policy may reach its limits without being able to generate enough aggregate demand to close the excess supply gap. (IMF Talk: Macro and Growth Policies)

...on the problems with Ricardian equivalence:

What might interfere with [the claim that it is optimal for households to save a tax reduction]? Any number of things: if households had been unable to consume as much as their optimal plan required because they lacked liquid assets and could not borrow freely, then the added liquidity provided by the tax reduction would enable them to consumer more now. If the Treasury were a more efficient, less risky, borrower than many households, then the appearance of some new public debt would also affect real behavior. And, of course, if consumers do not look ahead very far or very carefully, if they give little weight to the interests of descendants, or if they tend to ignore or underestimate the future implications of current budgetary actions, then Ricardian equivalence will fail, and tax reduction financed by borrowing will indeed be expansionary. All those "if" clauses strike me as very likely to be real and quantitatively important, and that suggests that Ricardian equivalence is not a practically significant limitation on fiscal policy. (Is fiscal policy possible? Is it desirable?, p. 12)

...on the problem with the natural rate of unemployment hypothesis:

Let me try to explain what nags at me in all this...We are left here with a theory whose two central concepts, the natural rate of unemployment or output and the expected rate of inflation have three suspicious characteristics in common. They are not directly observable. They are not very well defined. And, so far as we can tell, they move around too much for comfort -- they are not stable. I suspect this is an intrinsic difficulty. I have no wish to minimize the importance of, say, inflationary expectations. But we are faced with a real problem: here is a concept that seems in our minds to play an important role in macro behavior, and yet it's very difficult to deal with because it escapes observation and it even escapes clear definition. 

On the natural rate of unemployment, I think the behavior of the profession exhibits problems. In order to make sensible use of this kind of theory, you want the natural rate of unemployment to be a fairly stable quantity. It won't do its job if it jumps around violently from one year to the next. But that's what seems to happen. We, the profession, are driven to explaining events by inventing movements of the natural rate, which we have not observed and have not very well defined. The issue came up first in the passage of the big European economies from 2 percent unemployment, on average, to 8 or 9 percent unemployment, on average, within a few years. The only way to explain that within the standard model is to say that the natural rate of unemployment must have increased from something like 2 percent to something like 8 or 9 percent. The actual facts that could account for any such dynamics never seemed to me or to any critical person to be capable of explaining so big a change. So we are left with inventing changes in the natural rate of unemployment to explain the facts, and it is all done in our heads, not in any tested model. I regret to say that you often find this kind of reasoning: the inflation rate is increasing because the unemployment rate is below the natural rate. How do you know that the unemployment rate is below the natural rate? Because the inflation rate is increasing. I think we are all good enough logicians to realize that this is exactly equivalent to saying that the rate of inflation is increasing, and nothing more. 

It seems to me that we ought to be thinking much more about the determinants of whatever you choose to call it. I hate to use the phrase "natural rate" but of course I do. It was a masterpiece of persuasive definition by Milton. Who could ever want an unnatural rate of unemployment? (Fifty years of the Phillips Curve: A Dialogue on what we have learned, p.84)

...and more on the natural rate of unemployment:

There is nothing like an adjustable, unobservable parameter to keep a theory afloat in rough seas...I think the doctrine [of the natural rate of unemployment] to be theoretically and empirically as soft as a grape. To say that in the long run the unemployment rate tends to return to the natural rate of unemployment is to say almost nothing. In the long run the unemployment rate goes where it goes. You can call where it goes the natural rate; but unless you have a more convincing story than I have seen about the length of the long run and the location of the natural rate, you are only giving a tendentious name to a vague concept (1998, pp. 9, 91)