Introduction

Alan Greenspan confessed in 2008 to recognizing a ‘flaw’ in mainstream approaches to how financial markets work (Congress, 2010). Donald Kohn (2009), as Vice-Chairman of the Federal Reserve, admitted:
“It is fair to say . . . that the core macroeconomic modelling framework used at the Federal Reserve and other central banks around the world has included, at best, only a limited role for . . . credit provision, and financial intermediation. . . . asset price movements and the feedback among those movements, credit supply, and economic activity were not well captured by the models used at most central banks.”
The research agenda which culminated in macromodels without banks (Walsh, 2003), without monetary aggregates (Woodford, 2003), or without a financial sector (most real business cycle and dynamic stochastic general equilibrium models), has not been successful.
Macroeconomics has proceeded down the wrong path. In order to find the right one, we need to return to the crossroads at which it was taken. In the 1980s, the prevailing approaches faced deep problems. At the time, classical, many neoclassical, Keynesian, monetarist, post-Keynesian and many eclectic approaches shared one fundamental pillar linking money (M) to the economy: the ‘equation of exchange’ or ‘quantity equation’:
MV = PY (1)
For monetary policy purposes, to explain and forecast GDP (Y) and prices (P), as well as to estimate the money demand function, a stable velocity V was required. This stability was the thread on which macroeconomics hung. It held up for a while, but increasingly a ‘velocity decline’ was observed: nominal GDP did not grow as much as the money supply.

While the velocity decline has accelerated in many countries since, economists have turned away from the problem—not because a solution had been found, but because they had given
up searching for an answer. This was precisely the time when the moneyless economic models became increasingly attractive to economists. They were a form of escapism:
instead of rising to the empirical challenge to improve on their understanding of the monetary sector, economists chose to pretend that money and banks did not exist and had no influence on the economy. This was possible thanks to the prevalence of the hypothetico-deductive approach, which places little emphasis on empirical facts.

The new moneyless economics however merely produced further empirical puzzles and ‘anomalies’ which by today have also discredited it: ‘The notion that there is something about banks that makes them “special” is a recurrent theme’ (Blanchard and Fischer, 1989, p. 478), one that is empirically well supported (e.g. Peek and Rosengren, 2000; Ashcraft, 2005; Werner, 2005; see also Fama, 1985). However, economists failed to identify what makes banks special. So the recurring banking crises have remained unexplained as well—a highly visible and embarrassing refutation of moneyless and bank-less macroeconomics.

There is a future for macroeconomics, if it can explain the many and major empirical puzzles. Below I suggest a solution that has performed well empirically. It also has the advantage of being simpler: the principle of parsimony holds that a simpler explanation, relying on fewer assumptions, is preferable to a more complex one. Finally, it is not based on the deductive methodology, which was instrumental in leading macroeconomics into a cul-de-sac.