A new report from the Center for Economic and Policy Research in Washington DC tries to work out whether the IMF has really changed its thinking in response to the global economic crisis and the general perception that countercyclical responses (rather than belt-tightening austerity) are the right way to go in a recession.
After a (fairly polite) public row with Fund staff at a panel on the crisis (which you can watch on the 50 minute youtube below), the CEPR team went back and analysed IMF agreements with 41 countries. (For Fund geeks, these include Stand-By Arrangements (SBA), Poverty Reduction and Growth Facilities (PRGF), and Exogenous Shocks Facilities (ESF)). What did they find?
‘31 of the 41 agreements contain pro-cyclical macroeconomic policies. These are either pro-cyclical fiscal or monetary policies – or in 15 cases, both – that, in the face of a significant slowdown in growth or in a recession, would be expected to exacerbate the downturn. In some cases, the Fund subsequently relaxed the original conditions; sometimes (as in Hungary, Latvia, Republic of Congo, and Haiti) this appeared to be the result of social unrest or other pressures on the borrowing government.
In many cases the Fund’s pro-cyclical policies were based on over-optimistic assumptions about economic growth. 
The IMF has a history of over-optimistic projections in many countries. So it is not so easy to separate forecasting errors from an underlying bias toward overly restrictive fiscal and monetary policies.
Over the past year or two the IMF has been a strong supporter of the use of government fiscal stimulus to counter-act the world recession, and it has long supported expansionary monetary policies, e.g. in the European Union, as well. It may then seem paradoxical that so many of the IMF’s agreements concluded during this recession have been pro-cyclical. But there has long been a double standard for low-and-middle income countries, in that Fund policy does not allow or encourage the same types of expansionary macroeconomic policies as it recommends for the high-income countries.
The economic argument for this double standard is that developing countries face a much more binding foreign exchange constraint… Of course some countries headed into this downturn with unsustainable fiscal or current account deficits; but even in these cases there should be a strong bias towards waiting until the world recession has passed before attempting to adjust these deficits. Other countries may have an unsustainable debt burden; in these cases there is an argument for more and speedier debt cancellation in the near future, rather than trying to improve the fiscal balance while the economy is crashing.
In some countries the rationale for tightening macroeconomic policies during the current downturn has been to restore confidence as a result of capital flight. In these cases, the IMF should be more open to capital controls, which it does not recommend in any of the agreements, and in some cases (e.g. Pakistan) it opposes these measures.’
The report was rebutted by the IMF in another one hour panel discussion (that should test your youtube stamina), and the CEPR has responded to the rebuttal with an eight page response. Also worth checking out is ‘Doing a Decent Job‘, a Eurodad report on the IMF’s alleged conversion that goes into more depth at what is actually happening at country level. This one could run and run….