Sunday, September 27, 2009

The great divide in macroeconomics

Paul Krugman's NYT Magazine essay on a crisis facing modern macroeconomic theories has provoked an intense debate in the blogosphere. Mark Thoma has captured this in two exhaustive lists here and here.

The ongoing economic crisis has exposed the inadequacies of both the saltwater (who, in Keynesian tradition, think that monetary and fiscal policy interventions can mitigate or even prevent recessions) and freshwater (who, following Milton Friedman and the Chicago School and Real Business Cycle school, claim that recessions cannot occur without government interventions and even if they occur monetary policy is adequate to generate recovery) schools of economics. Apart from this, it has also exposed the sharp cleavage between these two, with the former having supported the aggressive monetary and fiscal policy actions of the last twelve months and the latter continuing to believe that economies work best when markets operate freely, with limited government participation.

More fundamentally, neo-classical and Keynesian economists differ about how quickly and whether wages and prices adjust in response to demand-supply mismatches. The former have built their macroeconomic models on the assumption that wages and prices are flexible and that prices 'clear' markets, and balance supply and demand, by adjusting quickly. In contrast, the latter believe that markets do not clear automatically and can remain stuck in a deficit equilibrium and also argue that the neo-classical models cannot explain the peristent and repeated short-run economic fluctuations.

A bit of wikipedia to fill up details (and refresh my own memory!) of the great divide that came to the fore sharply in the early seventies. Keynesian economics, that gained ascendancy in the aftermath of the Great Depression, argued that fluctuations in the business cycle due to a decline in aggregate demand (due to a fall in money supply or economic recession) would lead to a (not temporary) fall in employment and output and it requires Central Bank and government interventions, through monetary and fiscal policy actions, to stabilize the cycle. In contrast, the neo-classical school had focussed on market-based (through supply-demand balancing) determination of prices, output, income distribution and resource allocation, arrived at by the interaction of individuals who have rational preferences, maximize utility and act independently based on full information and firms whose objective is to maximize profits. They emphasize equilibria as solutions of agent maximization problems, and see limited role for governments in addressing economic fluctuations and crises.

The neo-classical assault on Keynesian economics in the early seventies, led by Milton Friedman's monetarism, put monetary policy at the center of the debate ("inflation is always and everywhere a monetary phenomenon") on addressing macroeconomic crisis and general deviations from the potential output. It rejected fiscal policy and other demand management policies as leading to the crowding out of private investments and inefficient allocation of resources. This was reinforced by the rational expectations theory, pioneered by Robert Lucas, which assumes that individuals take all available information into account in forming expectations and that though agents' expectations are wrong at every one instance, but is correct on average over long time periods.

Therefore, for example, attempts to lower unemployment through expansionary monetary policy economic agents will fail (after an initial fall in unemployment) because people and firms will anticipate the effects of the change of policy and raise their expectations of future inflation (and interest rate hikes) accordingly, which in turn will counteract the expansionary effect of the increased money supply. Stagflation will result and the events of late seventies and early eighties appeared to vindicate this.

And this meant that anticipated policies (monetary and fiscal) have no effect on employment and cannot stabilize the economy, as expectations nullify the gains from expansionary policies. Only surprise changes in, say, the money supply matter. This Edmund Phelps-Robert Lucas analysis appeared to have weathered the Keynesian attack on the neo-classical paradigm by explaining the evidently Keynesian behaviour of the economy in terms of imperfect information and the rational expectations formed from that. However, as Paul Krugman has argued, such rational expectations based models (which assumed that people had to form expectations about the future based only on price signals) "broke down as soon as you let people have access to any other information – say, by looking at interest rates, or reading a newspaper".

This limitation of the rational expectations led neo-classical revival along with the Lucas critique, led to another split, between the New Keynesians and the Real Businsess Cycle (RBC) schools, which has since manifested as the freshwater-saltwater divide. The Lucas critique argued that it is naive to try to predict the effects of a change in economic policy entirely on the basis of relationships observed in historical data. Models constructed based on historical information may not yield accurate predictions because the predicted co-relations can be expected to change when new policies are introduced. Instead, Robert Lucas claimed that it was important to model atleast the fundamental parameters that govern human (and firms) behaviour - preferences (objectives of agents), technology (productive capacity of the agents), and institutional (mainly resource) constraints - in response to (and in anticipation of) policy changes and then aggregate the effects of the resultant individual decisions to calculate the macroeconomic effects of the policy change.

All this encouraged the Keynesians to revisit the microfoundations of their models. The New Keynesian school that sought to build microfoundations for Keynesian economics, agree with the neo-classicists about the rational expectations among businesses and individuals but differ in that it provides for a variety of market failures (atleast in the short run) - prices and wages are "sticky"; monopolistic competition; credit market imperfections; co-ordination failures among market participants, leading to aggregate demand multipliers and possible multiplicity of equilibrium; unemployment caused by moral hazard and matching frictions etc. Since these failures (especially that relating to price and wage stickiness) lead to an economy stagnating at less than full employment, they argued that macroeconomic stabilization by the government (using fiscal policy) or by the central bank (using monetary policy) can lead to a more efficient macroeconomic outcome than a laissez faire policy would.

The New Keynesian models with 'sticky' wages and prices rely on the stickiness of wages and prices to explain why involuntary unemployment exists and why monetary policy has such a strong influence on economic activity.

However, the New Keynesians have focussed attention mainly on the role of monetary policy, especially when both inflaiton and output are down, in which case expanding the money supply (by lowering interest rates) helps by increasing output while stabilizing inflation and anchoring long term inflationary expectations. They have paid limited attention to the role of fiscal policy, though they acknowledge its utility in stabilizing output.

The neo-classicists responded to the deficiencies pointed out by Robert Lucas and Co by formulating their RBC Theory models in which business cycle fluctuations, to a large extent, can be accounted for by real (in contrast to nominal) shocks. They see recessions and slowdowns as the efficient response to exogenous changes in the real economic environment and hold that the level of national output necessarily maximizes expected utility. Accordingly, they feel that governments should concentrate only on the long-run structural policy changes and not intervene through discretionary fiscal or monetary policy designed to actively smooth economic short-term fluctuations. In other words, unlike other theories of the business cycle (including Keynesianism and monetarism), the RBC school see such cycles as "real", in so far as they do not represent a a failure of markets to clear but reflects the most efficient possible operation of the economy, given its specific structure.

Further, in response to the Lucas critique, both the aforementioned groups went about constructing their own versions of macroeconomic models (different from the earlier static and deterministic general equilibrium models) based on the micro-foundations of agent behaviour. Since macroeconomic behavior is derived from the interaction of the decisions of all these players, acting over time (dynamic), in the face of uncertainty (probabilistic) about future conditions, these models are classified as dynamic stochastic general equilibrium (DSGE) models. The DSGE models accordingly attempts to explain aggregate economic phenomena, such as economic growth, business cycles, and the effects of monetary and fiscal policy, on the basis of macroeconomic models derived from microeconomic principles.

The RBC theory based DSGE models, formulated by Fynn Kydland and Edward Prescott assumes flexible prices to study how real shocks to the economy might cause business cycle fluctuations. In contrast, the New Keynesian DSGE models assume that prices are set by monopolistically competitive firms, and cannot be instantaneously and costlessly adjusted.

To locate the debate about the state of macroeconomcis in the present generation and the latest research, Justin Wolfers draws attention to a paper by Narayana Kocherlakota who analysed the works of the latest generation of tenured macroeconomists in the top 15 American university economics departments and is cautiously optimistic about the future of macroeconomics. He does not find any saltwater-freshwater divide among them and finds that they do not ignore heterogeneity nor frictions nor bounded rationality, and their models incorporate a role for government interventions. However, as Wolfers points out, unlike their predecessors, being more interested in economic theorizing they have not been active participants in current economic policy debates nor are they engaged with issues of the real world.

The ongoing crisis appears to have decisively, atleast for the time, shifted the upper hand to the Keynesians. Paul Krugman has been at the forefront of those leading the backlash agianst the free-market monetarist ideologues who opposed the government fiscal stimuluses and aggressive unconventional monetary policy responses by the Federal Reserve during the present crisis. He has even lamented that it is the "Dark age of macroeconomics"!

All this has expectedly generated a very shrill reaction from Chicago School, with John Cochrane leading the charge. However, as Bradford Delong illustrates, with this excellent compilation, the Chicago School is clearly on weaker wicket in this battle, atleast for now. They appear more like Ostriches burying their head in the sand when their favorite theories and holy cows seem like crumbling in the face of the overwhelming evidence of reality, instead of rationally and objectively responding to their critics. See also this, this and this by Paul Krugman, Nick Rowe, and Mark Thoma respectively.

Update 1
What's wrong with modern macroeconomics? - Conference papers by Mark Thoma.

Update 2
Alan Kirman has this critique of modern macroeconomics and financial market theories and argues that the economy is complex adaptive system and advocates making the analysis of the structure of interaction between agents more central to our models.

Update 3
Charles Kindleberger's anatomy of a bubble, which draws heavily from Hyman Minsky is available here. John Quiggin on what next for macroeconomics.

Update 4 (4/9/2014)

Excellent essay by Olivier Blanchard cautioning against the dark corners of the economy, where danger lurks and where all prevailing macroeconomic models fail. He narrates how the emergence in the seventies of the rational expectations belief and the development of techniques to solve macroeconomic models that assumed rational expectations came to shape the future of macroeconomics for more than three decades. The rational expectations belief necessarily meant a backward-looking and linear model of the economy. Though it was subject to constant shocks and resultant fluctuations, it generally recovered quickly (the central claim of the neo-classical RBC school).

While economists acknowledged the possibility of small shocks having large effects - bank runs, sudden stops, liquidity traps etc - it was thought that sound macroeconomic policies could help avoid them and also these challenges had been conquered in developed economies. The Great Moderation only served to reinforce this belief, only to be jolted by the Global Financial Crisis and the resultant Great Recession. They have drawn attention to the grave dangers faced when the economy reaches the dark corners - liquidity trap, high debt levels, financial market opacity etc.

Noah Smith has two nice parables that explain the RBC and New Keynesian theories. See also this profile of Ken Arrow


Gedi said...

The Gedi thinks that Economists do not understand national accounts and hence mess up.

Anonymous said...

You wrote: This Edmund Phelps-Robert Lucas analysis

This is not quite fair to Milton Friedman. With his 1967 Presidential Address to the AEA, he presented a very similar model, so it is usually referred to, AFAIK, as the Phelps-Friedman model, not Phelps-Lucas; see Krugman's account in the NYRB in 2007 (Feb 15, 2007, v54, #2 in case the link does not work). Lucas formalized the model in a paper published in JET 1972. said...

I firmly believe that the monetary and fiscal policies that the world is embarking on have created liquidity induced securities speculation (all over again) and have engendered inefficient resource allocation. You can't erase trillions in bad loans and poor capital investment decisions just by throwing around some "stimulus" funds and lowering interest rates to 0%.

READ - The Resurgence of Keynesian Economics and Interventionism