I read an interesting piece of critique yesterday by Paul Romer. The core of his critique is that modern macroeconomics (~Dynamic Stochastic General Equilibrium models) has drifted during the past 30 years or so into a form of Post-Real Models, like in theoretical physics, where one has M-theories, string theories and so forth. In essence, mathematical consistency and structure with little touch with reality I suppose. I have to say that as a practising professional economist, I must agree with Dr. Romer. Modern macroeconomics most likely does more bad than good in terms of understanding how the economic machine works. I do understand the need for microfoundations, but given that most DSGE-models do not cater for bank leverage, financial markets and credit/leverage cycles in general, I do not see any intellectual interest in these models currently. Moreover, modelling banks as mere intermediaries is plain wrong. Reading Irving Fisher, Friedrich Hayek, Knut Wicksell, Joseph Schumpeter and J.M. Keynes is far more useful (and rewarding).
Economics and the scientific method
Nevertheless, I do not accept the position of dismissing economics altogether. Economics is a set of ideas. Even though economics is not a natural science, economics, should, in my opinion follow the scientific method, i.e. the method of doing economics should follow the rules of:
- Characterising the phenomenon
- Formulating hypotheses
- Making predictions
- Experimenting –falsification and verification
- Evaluation and making improvements
Most likely the main problem with DSGE -models and modern macroeconomics is with the first step. A model of an economy should not be based on a simple linear optimal control problem with stochastic shocks. One should model the economy more like some artificial economies are modelled in computer games (Simcity, Civilization, Democracy). Physicists model complex phenomena using for example the Ising model, why not to bring this paradigm into economics as well? Optimal control is a good tool for controlling ICBM-missiles, but not so much for complex, human economies. If the model does not work and is ontologically unrealistic, I don’t see the benefits of having microfoundations.
What is good economics ?
There are a lot of good theories and models in economics. The following list is a list based on my experience, on what is important.
- Economic efficiency
Given that resources are scarce, economics has quite nice formalisations for using resources efficiently. Usually we have the following vector program, where one needs to solve:
The objective vector could represent a set of utilities for individuals in a society, or profits for a firm, or a utility function for a central bank. What is important that efficient use of resources comes through a (convex) optimisation program. Pareto efficient solutions can be desciribed for partially ordered sets. Of course one of the triumphs in welfare economics are the theorems that codify the invisible hand of Adam Smith: competetive market allocation is Pareto-optimal (conditions cannot be improved for some, without affecting negatively others). Kenneth Arrow and Gerard Debreu were the pioneers in these issues.
2. Public choice
Economic methodology married with political science. Assuming voters, bureaucrats and politicians to be selfish utility maximisers, one can deduce interesting things. My favourite is the median voter theorem, which basically says that the median voter preferences are represented by a majority rule.
3. Economics and social welfare -veil of ignorance
This is close to ethics and moral philosophy, especially Rawlsian moral theory but maybe surprisingly economics has something to offer for social welfare theory as well. One of the most interesting results is by John Harsanyi (1955), which says that under some mild assumptions, social welfare is a weighted sum of individuals’ expected utility. Expected utility was by the way developed by the genius John von Neumann. The veil of ignorance -concept was as well originally developed by Harsanyi well before John Rawls.
4. Cournot competition and game theory
Cournot competition (1838) is an equilibrium concept (Nash equilibrium), where two firms choose optimal levels of production in a strategic setting, where individual firms have market power. Firms have what is so called “best response functions” and the intersection of these curves define the equilibrium. Very intuitive, and profound. Again, while John von Neumann was not busy discovering quantum mechanics and the first computer, he developed game theory further as well. Minimax theorem and of course later John Nash’s equilibrium theorems are important.
4. Portfolio theory
Choosing an optimal allocation for an investment portfolio. Harry Markowitz saw the light and formulated it as a quadratic program. One has the covariance matrix of stock returns C, the return vector R, risk parameter lambda and the weights w. The program is simply:
Capital asset pricing model is kind of funny, because it seems trivial (as a projection in the Hilbert space) but it can be deduced also from utility maximisation. What CAPM says is that there is a price of risk, which can be calculated based on the covariance with the market risk.
6. Options pricing and the risk neutral probability measure
Heat flow and financial derivates benefit from the same underlying structure, which is the continuous Brownian motion that models white noise in the continuum limit, Black & Scholes (1973) realised that using the Ito lemma and Fourier transforms, one could derive excplicit pricing formulae for European options, stemming from the following parabolic PDE:
The link between the heat equation and BS equation comes from the fact that the generator of the stochastic process in stock markets and diffusion processes is a second order Partial differential operator.
The risk neutral probability measure is an imagined probability measure that equates the price of an asset with the expectation of its present value payoff. In this way, all the assets can be priced using only expectations.
7. Covered interest rate parity
The determination of exchange rates in relation to interest differentials is of course interesting, at least for traders. So here we have it
In other words, the ratio of forward FX rate to spot FX rate is the ratio of the interest rates in the respective currencies.
8. Comparative advantage (international trade)
‘Focus on where you’re good at’ -David Ricardo
9. The Trilemma (international finance)
It is impossible to have the following things at the same time: fixed exchange rates, free capital mobility and an independent monetary policy.
10. Tragedy of the commons
We have an incentive to defect, when we have common property –>private property.
This is the list. As you can see, no DSGE models.