I prepared this note on the contribution of Joseph Stiglitz of the same title to an important symposium in Oxford Review of Economic Policy, Volume 34, Numbers 1–2, 2018 (pp. 70–106). It is for economists. A shorter version is published in the AUT Briefing Papers series.
Stiglitz’s text is in italics:
I believe that most of the core constituents of the DSGE model are … sufficiently badly flawed that they do not provide even a good starting point for constructing a good macroeconomic model. These include
(a) the theory of consumption;
(b) the theory of expectations—rational expectations and common knowledge;
(c) the theory of investment;
(d) the use of the representative agent model (and the simple extensions to incorporate heterogeneity that so far have found favour in the literature): distribution matters;
(e) the theory of financial markets and money;
(f) aggregation—excessive aggregation hides much that is of first order macroeconomic
significance;
(g) shocks—the sources of perturbation to the economy; and
(h) the theory of adjustment to shocks—including hypotheses about the speed of and mechanism for adjustment to equilibrium or about out-of-equilibrium behaviour.
I cannot review in detail all of these and other failings … so I am selective. So am I.
Many of these are related. For instance, the presence of imperfect and asymmetric information leads to credit and equity rationing. Thus, individuals in maximizing their lifetime utility have to take into account credit constraints and … this gives rise to a markedly different problem than that analysed in the standard DSGE model.
One of the reasons that the representative agent model doesn’t work well is that some individuals are credit constrained, others are not. Moreover, numerous studies … have emphasized the importance of debt for aggregative behaviour; but in a representative agent model, debt (held domestically) nets out, and therefore should have no role. At least at times, short-run to medium-term macroeconomic analysis needs to debt and real debt-dynamics … institutional details can matter.
More generally money is not an integral element of a DSGE model. In the traditional Arrow-Debreu GE model there is no substantive money. No transactions occur out of equilibrium so there is no need for a medium of exchange; barter is sufficient. Insofar as there stock of money, it is like any other commodity; presumably the market participants think it is, in some sense, fully back. There may be a standard of value, a numeraire, but it is of no great significance.
Stiglitz adds to this by observing if there are only representative agents in the model – as applies for DSGE models – noone can be in debt to anyone else because all agents are identical. (This applies for a closed model. In an open model each agent may in debt offshore – by the same amount.)
Debt raises another issue with which GE models cannot deal. Recall that all prices in a GE model are non-negative. It follows that nobody can have negative equity in their financial balance sheet. Were that true there would be no financial crashes.
For a theory which prides itself on its microeconomic foundations, the DSGE model’s are remarkably inadequate in regard to the role of money. Given the money is at the heart of macroeconomic behaviour – pre-monetary economies suffered only from external shocks – the DSGE model did not just fail to predict the Global Financial Crisis. It was not designed to predict a financial crisis any more than it was designed to predict the weather.
So are DSGE models are of any use? Stiglitz says acerbically ‘[f]rom a social perspective, whether the economy grows next year at 3.1 per cent or 3.2 per cent makes little difference.’ Perhaps for forecasting purposes they should be pitted against auto-regressive models. The structural vs auto-regressive forecasts is an old debate. The last time I looked at it, the auto-regressive forecasts were superior except where there was a novel structural change (such as an oil shock) as you could adapt a structural model using more economic analysis to allow for the circumstances outside the model’s range of experience. I suspect that still holds, but even so that only works if the structural model is not too rigid and the shock exists within the theoretical compass of the model. For instance most do not have balance sheets which play an integral part in the financial crises.
At this point we could dismiss DSGE models as irrelevant but I suggest that Stiglitz’s point is not about DSGE models per se, but the economics underpinning them and which underpin many macroeconomists’ thinking. So let us go on with his critique. Just to remind you, the ones I have yet to cover are:
(a) the theory of consumption;
(b) the theory of expectations—rational expectations and common knowledge;
(c) the theory of investment;
(d) the use of the representative agent model (and the simple extensions to incorporate heterogeneity that so far have found favour in the literature): distribution matters;
(e) the theory of financial markets and money;
(f) aggregation—excessive aggregation hides much that is of first order macroeconomic
significance;
(g) shocks—the sources of perturbation to the economy; and
(h) the theory of adjustment to shocks—including hypotheses about the speed of and mechanism for adjustment to equilibrium or about out-of-equilibrium behaviour.
Stiglitz makes numerous reference to expectations in his paper but it seems to me his basic point is captured by ‘analyses of expectations were (correctly, in my view) not based on what those might be if it were assumed that individuals had rational expectations, or acted as if they did, but on survey data of what expectations are and have been.’ Resorting to survey data is an admission that economists do not have a useful theory to predict actual expectations or how they are formed. He rightly, in my judgement, rejects the theory of rational expectation. It is a strong theory with precise predictions which are frequently wrong to the point of useless.
I was disappointed by Stiglitz’s paper’s discussion on the potential of behavioural economics. Perhaps there was not space, perhaps he thinks the theory is yet too primitive to play a major role in macroeconomics. I am more optimistic.
I think his point about investment is that, like consumption, it suffers from credit rationing. This seems to have been particularly relevant because financial institutions seem to have been reluctant to fund business investment , since the GFC.
His doubts about aggregation on the production side are summarised as:
Long ago we learned the difficulties of constructing an aggregate production function. The ‘putty-putty’ model provides great simplification, but one should not claim that any analysis based on it is really ‘microfounded’. While earlier analyses provided a critique of the use of the standard model for equilibrium analysis, e.g. when there is production of commodities by means of commodities or when there are production processes involving capital goods of markedly different durability; the use is even more questionable for analyses of dynamics: the dynamics of putty-clay models and vintage capital models, for instance, are markedly different from those of putty-putty models.
Capitalism is fundamentally about turning putty into clay and being rewarded for the loss of flexibility by a higher return.
Even more important is perhaps the aggregation of the whole economy into a single sector, particularly when the underlying stress on the economy is one of structural change, requiring the movement of resources from one sector to another (say agriculture to manufacturing). I, of course, heartily agree with this last point. Single commodity thinking has been the bane of too much of New Zealand economic analysis.
He goes on [m]onetary policy is typically presented as an efficient tool. But monetary policy has disproportionate effects on interest-sensitive sectors, thus inducing a distortion in the economy that simply is not evident in a one-sector model. I would add that one of our most interest-sensitive prices is the exchange rate.
Stiglitz is also concerned with aggregation of households which raises parallel mathematical problems. Assuming a representative agent goes too far, because it eliminates any possibility that distribution matters. Again I add ‘hear, hear’.
He has a lot about shocks. I suspect much of this is an attack upon real business cycle theory. Stiglitz says that in (most) DSGE models, downturns are caused by an exogenous technology shock. In agriculture, we know what a negative technology shock means—bad weather or a plague of locusts. But what does that mean in a modern industrial economy—an epidemic of some disease that resulted in a loss of collective knowledge of how to produce? By contrast the shocks giving rise to economic fluctuations in many, if not most cases, is clearly endogenous. The 2008 shock was endogenous, caused by the breaking of the housing bubble. This is really an argument about the meaning of exogenous and endogenous. ‘Exogenous’ means outside the model where it is unreasonable for the modeller to deal with but. Endogenous are things that are – or should be – inside the model. Real business cycle theory said that endogenous shocks did not exist or were unimportant. I agree with Stiglitz that economists ought to be able to analyses endogenous shocks like those which generated the GFC, even if (by definition) they cannot predict them with any precision.
The final point in Stiglitz’s list is that he has doubts about the theory of adjustment to shocks. I found the section both rich, dense and lacking detail. Here are some extracts from the key section. One of the reasons that downturns with high levels of unemployment persist relates to the process of adjustment. … the super-smart individual simply thinks through the consequences of choosing any other set of current wages and prices [but] it is not apparent how that is to be done in the context of a world without common knowledge. If there were a full set of markets extending infinitely far into the future, the problem … would not occur. But there are not—this is one of the key market failures. …. [The models assume] a decentralized process of wage and price adjustment, with wages and prices in each market responding to the tightness in that market (in the labour market, that is the simple Phillips curve, asserting that wages rise when labour markets become tight). In the short run, … adjustment processes may be disequilibrating: the fall in wages as a result of unemployment may result in a decrease in aggregate demand, increasing the level of unemployment. … What matters is, of course, real wages, and that depends on the adjustment of wages relative to prices. Wages and prices may both be falling at the same rate, resulting in real wages being constant, a kind of real wage rigidity. … Moreover, the deflation itself has a depressing effect, since it increases the real interest rate (holding everything else constant). Similarly, adjustments of prices have balance sheet effects of the kind already discussed, with large macroeconomic consequences.
My selection may have made Stiglitz’s argument even less transparent than he presented it. Let me put it this way. Should not the econometrics in a DSGE model get the right adjustment path – after all it claims to be ‘dynamic’? (There is an implicit assumption that the adjustment path is short, which helps the econometrics.) He says DSGE models … simply assume that the economy jumps to the new equilibrium path. However if adjustment processes are asymmetric (wage and price rigidities are an example) then current econometric techniques will lead to poor representations of actual adjustment paths. If the adjustment processes take a lot of time they will be poorly estimated. Add in this all Stiglitz’s other criticisms and the so-called ‘dynamic’ DSGE models are not going to have very good dynamics.
There are other criticisms of DSGE in the Stiglitz paper – most are elaborations of what has been already discussed – but space demands I do not detail them. I finish with two general points.
First, most of Stiglitz’s criticisms – of assumptions which are not robust – were well-known shortly after the flowering of the first developments in GE theory – the Arrow-Debreu model. Stiglitz reminds us of them, for most economists seem to have forgotten them. Instead they have been ignored in the main analysis in their economic paradigm without the caveats which weaken the robustness of their models. It is true that Stiglitz refines and sharpens our understanding of the caveats. In my case I am now more alert to the importance asymmetries, the complication of heterogeneity in the household sector (I was well aware of it in the production sector) and I am even clearer about the role of money (or the lack of the role) in general equilibrium theory.
Second, while Stiglitz paper focuses on the use of the limited paradigm in DSGE models, he clearly intends the criticism to apply more widely across economics. There are many economists who struggle with general equilibrium theory or who have only the vaguest notion of DSGE modelling but whose own account of the economic world is subject to the criticisms in the Stiglitz paper.
In the last half of the paper – there is not the space to discuss it here – Stiglitz proposes an advance on the DSGE standard model which deals directly with the criticisms earlier in the paper. There is both a verbal exposition and mathematical one. The latter is but a set of equations which are not solved and offer no new insights. They may well be to complex too do so. (For example, it would be nice to know whether his model has a unique equilibrium which is another common GE assumption which wrecks havoc if it is not true.)
I finish with one of my heroes comments of mathematics in economics. Alfred Marshall wrote
[I had] a growing feeling in the later years of my work at the subject that a good mathematical theorem dealing with economic hypotheses was very unlikely to be good economics: and I went more and more on the rules – (1) Use mathematics as a shorthand language, rather than an engine of inquiry. (2) Keep to them till you have done. (3) Translate into English. (4) Then illustrate by examples that are important in real life. (5) Burn the mathematics. (6) If you can’t succeed in (4), burn (3). This last I did often.
Paul Krugman, who also gives a paper in the symposium, illustrates this style well. He is one of Marshall’s great successors.