Prosperity for All was published in 2017 by the Oxford University Press. It is 291 pages, and is broken into 11 chapters. Although there are some technical appendices (which were added in response to academic peer reviewers), the book is written for a general audience (that has an interest in economic theory).
This article is essentially focused on the question: is this book worth reading? Professor Farmer's ideas are distinct, and would take some time to discuss in detail. Later articles will return to the meat of his analysis and proposals. In other words, this is the first part of a multi-part review.
What is in the Book?I would split the contributions of the book into three parts.
- A potted history of the development of mainstream macro, in particular dynamic stochastic general equilibrium (DSGE) models.
- Farmer's discussion of his variant, which is known as the Farmer Monetary Model.
- Farmer's policy discussion, which culminates in his discussion of the sovereign wealth fund.
I was mainly interested in his theories (point 2), and in my view, make the book worthwhile. It might be easier for some to start with his academic articles (or blog posts), but I prefer to start with book-length introductions.
The history of DSGE macro (point 1) may be of interest to readers who are new to the "macro wars" and want to know what they are about. Farmer's views on model construction are not radically different than the mainstream; he just proposes swapping out one very unsatisfactory part of the models. His complaints line up with my practical criticisms of DSGE macro -- the models are completely unable to model business cycle dynamics, as they have obliterated the importance of the business sector.
For readers who are not aligned in the macro wars, or are of a mathematical bent, Farmer's criticisms may be easier to swallow than the post-Keynesian critiques. That said, the post-Keynesian criticisms of "mainstream" macro are powerful (although I am unconvinced by how they are often presented). Farmer's suggested fix for DSGE macro models might not be enough to save them from post-Keynesian critiques. Since I now have an unusual bias to prefer models that are useless in practice (yes, seriously), I am not particularly perturbed about many arguments about the theoretical approach.
If there is a weakness in Farmer's approach, it would show up in policy recommendations. Since I believe his policy proposals would fail, this reflects on the interpretation of his theory.
Animal SpiritsThe use of "animal spirits" is why I distinguish Famer's theories from DSGE macro. Business cycles are about investment. Firms invest when they are confident about the prospects for future profits. Meanwhile, fixed investment are a source of profits in the Kalecki Profit Equation. The implication is that there is a self-reinforcing dynamic at work. Capitalist economies are unstable -- they normally explode upward. Unfortunately, the self-reinforcing cycle can go the other way, as many found out the hard way in 2008.
(Another major distinction between Farmer's work and most DSGE macro is the treatment of unemployment. Standard DSGE models suggest that there is a single equilibrium unemployment rate, which converges to NAIRU. Farmer's models have an infinite number of equilibria, and there is no convergence to NAIRU. Although there appear to be theoretical differences, this ends up with some similarities to post-Keynesian views about the labour market.)
This is in line with other Keynesians -- other than New Keynesians (DSGE believers). If we were to take DSGE models literally, recessions happen because people suddenly decide to simultaneously take a vacation.
Although "animal spirits" sounds squishy, one could operationalise the concept in mathematical models. We can imagine it to be a time series, and it would have a well-defined role within a mathematical model. "If animal spirits rise from 7 to 8, fixed private investment will rise by 5% in the next period." We could attempt to estimate this variable with the dreaded Kalman filter -- exactly like potential GDP, or the natural rate of interest (or r*). The only unusual thing about "animal spirits" is that it is not tied directly to a measured variable, and so its level is purely the result of convention. (For example, a neutral animal spirits could be 0, or 100.)
I will write technical articles later discussing the concept -- they will be part of my next book on business cycles.
Fiscal Policy DismissedThe traditional Keynesian way to damp down the business cycle is to use fiscal policy as well as monetary policy. Professor Farmer dismisses fiscal policy, arguing that expenditure multipliers are low.
He concedes that economic techniques like SVAR show that the multiplier is greater than one, but "if the model is wrong, so is the inference about the size of the multiplier." I have no objection to that argument. Instead, he cites the research of Valerie Ramey who uses a "narrative approach" to the topic. The idea is to look at discussions of fiscal policy, look at the amount of spending, and then examine the growth impact. Using that approach, she found a multiplier less than one: government spending causes some contraction in private activity.
However, Farmer analytically shoots himself in the foot by admitting that tax cuts do have a stimulative effect. The most effective role of activist fiscal policy is to fill in demand when a recession hits, and that is almost always done with a tax cut. Tax cuts can be implemented almost as fast as rate cuts, and can be targeted at areas of weakness. Outside of a recession, even many fans of fiscal policy do not advocate attempting to achieve optimal trajectories by tuning tax policies. (The MMT job guarantee is a passive fiscal policy, that reinforces automatic stabilisers.)
Going back to first principles, I would argue that we should not expect to easily measure the multiplier associated with fiscal expenditures during an expansion. (We should expect almost no effect on growth if the spending increase is offset with higher taxes, which is common practice. If taxes rise to offset greater expenditures, all that has happened is that the government is a greater percentage of GDP. There are arguments whether increasing the size of government as a percentage of GDP slows growth, but those theories are based on trend growth rates, not an immediate shock to growth.)
Using the definitions of dynamic systems, the economy is obviously unstable -- stock and flow variables are growing at exponential rates. When people think of economic stability, they are typically referring to growth rate stability. In order to get growth rate stability -- which we tend to see in the middle of the cycle -- there have to be internal dynamics that dampen growth impulses. We know of many such effects.
- Taxes and welfare state spending stabilise nominal growth.
- Inventories can be drawn down in response to growth spurts.
- Growing faster than trade partners generally results in a deteriorating trade balance, which is a drag on growth.
- Households smooth their consumption.
- If one believes that interest rate policy has an effect on growth, the central bank reaction function would presumably cancel out the effect of the expenditures. (This does not imply that the expenditures had no effect, just that it was cancelled out by another action.)
This means that any impulse to growth is smoothed out, which implies that it is hard to move growth away from its steady state level. Therefore, one-off wiggles are damped out relative to the trend. However, fiscal policy settings help determine that trend growth rate -- taxes are one of the key damping factors for nominal income growth.
Like other conventional economists, Professor Farmer is trapped in the 1960s optimal control world where policymakers attempt to fine-tune the growth path of GDP. Who cares? If we want to fight unemployment, create *@&^%$ jobs, let GDP seek its own level.
I will likely return to this debate other times. I would however note the importance of the mix of spending; pretending dollar amounts measure the effect of a policy is wrong-headed. Many Democrat-leaning economists were deeply concerned about the inflationary effect of the Republican tax cuts, but no such inflation is yet in evidence. Handing money to billionaires and corporations so that they can do buybacks does not do a lot for aggregate demand (although I imagine that prices of luxury goods have been soaring; but those are not in the CPI). I have only just encountered this research, but the paper The deterioration of the public spending mix during the global financial crisis by Debra Bloch and Jean-Marc Fournier discusses indicators based on the mix of government spending.
Targeting equity prices to stabilise the economy?The culmination of Professor Farmer's analysis is the suggestion that governments set up a sovereign wealth fund that attempts to set the level of aggregate equity prices. The fund would be yet another independent body that attempts to fine-tune the economy, this time by borrowing at Treasury bill rates to buy equities. It would go long and short, and try to keep equity prices on some optimal path. The objective would be to stop financial crises from happening, which as Farmer notes, occur after stock market declines.
I used to work at a large (by Canadian standards) sub-sovereign wealth fund, and I am familiar with the environment such funds operate in. If he had showed up at my office with that theory, "polite disdain" would have been the nicest possible reaction I could have given to his suggestions. Any markets professional with even the slightest Austrian inclination would probably hit the ceiling when hearing the plan. This is the Plunge Protection Team turned up to 11!
I will return to this topic later, as it is interesting (unless the euro crisis drags me into commenting on it...). For now, I will outline the various hurdles the suggestion faces.
- Is it possible to stabilise equity prices at all? How are these trades going to be implemented? What is the set of allowable securities? Is management of the basket of securities done by government employees, or is it outsourced? How tightly are prices to be controlled? Is the target public information? What discretion is used in security selection and trading? (Even if you have an algorithm trading, the choice of the algorithm is discretionary.) How will the government vote its shares?
- Is this fund politically sustainable? Although Professor Farmer is correct that a floating currency sovereign has deep pockets and can take the long view, how will politicians react when the fund loses its first trillion dollars?
- Will the markets be destabilised? Price pegs invite the private sector to leverage assets against the fixed price. At some point, leverage may get so large that the interventions required to save the system would cause anyone running the fund to collapse in panic.
- Will this have any effect on the real economy? Yes, there is a statistical relationship between equity prices and the real economy. However, that does not imply a causal link. If we engage in a bit of statistical hyperbole, the yield curve has a nearly perfect forecasting record for modern U.S. recessions. (Disclaimer: the previous statement might be incorrect, that is not my concern.) Nevertheless, nobody sensible believes that by pegging the short and long end of the Treasury curve, we can avoid recessions forever. Professor Farmer argues that the stock market is different; I will let readers form their own opinion. (I will return to this in my planned later article.) Will handing trading profits to the rich trickle down to the low-skilled workers that lose jobs in a slowdown?
Concluding RemarksAs is obvious, this article is only half of a review of Prosperity for All. The second half will be filled in when I write the technical discussions of the theory of the book, and his policy proposals.
Even if one believes that sovereign wealth fund would fail as a stabilisation tool, it is an interesting question to ask why it would fail.
(c) Brian Romanchuk 2018