Tuesday, 11 September 2012

The illogicality of the IMF

The IMF has just produced a new report on Ireland. It is not happy reading. But it is interesting- not least because it says as much about the beliefs of the IMF economists producing the report as it does about Ireland.

The report is structured in four parts:

1. Household consumption, wealth and saving

2. Access to credit, debt overhang and economic recovery: the Irish case

3. Medium-term fiscal consolidation in Ireland: growth-friendly, targeted, sustainable

4. Averting structural unemployment in Ireland

Each of these is in effect a separate report in its own right, and there is no attempt to draw over-arching conclusions from the findings of the four reports. Consequently the recommendations from one report are contra-indicated by recommendations from another report. Really the IMF should do better than this. But that's not my main issue with this report. It's the appalling logic.

Part 1, the Household Consumption, Wealth and Saving section, notes that Irish households are highly indebted and their preference is for deleveraging (saving). This section claims that the main reason for their high indebtedness is the fact that incomes have fallen. Yet in Part 4, Averting Structural Unemployment in Ireland, the figures for reduction in hourly nominal wage rates in both the public and private sector do not indicate this as a primary cause of high indebtedness. Private sector average rates have fallen by 2.25% since 2009 and public sector by 3.25%. This is hardly a large reduction, and Part 4 suggests that employers have resorted to layoffs rather than wage cuts, leading to high unemployment.  Part 3 also notes that there have been large cuts to both in-work and out-of-work benefits. So the true story in Part 1 is not simply that "incomes have fallen". They have, but the main components of that fall are high unemployment (causing catastrophic loss of income), benefit cuts and increased taxes - not cuts in nominal wages. So household debt burdens (as a proportion of income) did increase as a consequence of the financial crash and ensuing recession, but they have been made much worse by the enormous fiscal adjustment carried out by the Irish government since 2008.

This is the trouble when you look at things only from a macro perspective. You can reach the wrong conclusion as to the causes of the problem you identify, and therefore suggest the wrong solution.

However, the writers of Part 1 are not only guilty of failing to observe that the wood is made of trees. They also produce one of the worst examples of bad economic logic I have ever seen. Even their citation does not support them. The relevant bit is this, in section 10 on p.7:
Disaggregating the contributions to the forecast highlights the significance of Ricardian effects whereby households relax savings when the government cuts the fiscal deficit (and vice versa); this effect is consistently found in the empirical literature.
Their citation is Huefner &Koske's 2010 paper "Explaining Household Savings Rates in G7 Countries: Implications for Germany" . But this paper does NOT consistently find Ricardian effects. It said only two countries in its sample - the US and France - demonstrated Ricardian effects: the other countries did not. The authors survey other research that does show such effects, but then specifically say that their findings do not agree with this. And note that this paper is about SAVING. Economically, paying off debt and saving is the same thing - but behaviourally they are very different, as I shall discuss later on.

Despite this, the IMF then go on to say:
Using the coefficient for the fiscal balance as proxy for this Ricardian offset, the ongoing fiscal consolidation in Ireland will contribute most to the estimated reduction in the household savings rate. The reduction in household liabilities or net wealth has a small yet significant contribution, likely a reflection of the common difficulty of estimating the balance sheet effect from empirical data.The decline in unemployment and the increase in the dependency rate play minor roles.
So, further fiscal tightening will force households to reduce the rate at which they pay off debt. It doesn't take a genius to work that out, but what has it got to do with Ricardian equivalence?

Ricardian equivalence essentially says that if government increases its deficit, households save more in expectation of higher taxes later on to pay that deficit. Therefore - the theory goes - government deficit spending is pointless as a demand stimulus, because households hang on to their money instead of spending it.

Here the IMF is using Ricardian equivalence in reverse. It is arguing that deficit reduction raises households expectations of future tax cuts and therefore encourages them to reduce saving. But the equivalence is asymmetric: households may indeed save more in anticipation of tax hikes, but as governments usually prefer to increase spending rather than reducing taxes when there is a surplus, and households know this perfectly well, it would be a foolish household that reduced saving before those cuts are announced (and even then they might wait until the cuts are actually made - governments can change their minds).

But there is a much simpler explanation for reduced savings rates when governments are tightening. And for that we need to look at how people actually behave - which the IMF economists have failed to do.

When taxes are raised and benefits cut, household real incomes fall unless real wages increase to compensate - which has not happened in Ireland and shows no signs of doing so. When household real incomes fall, in most cases this shows itself as pressure on discretionary spending. Saving is a discretionary spending activity: instead of spending surplus money (money I don't need for essential household expenses) immediately, say on a holiday, I put it away for spending in the future. So, if I have less real income because I have higher taxes and/or lower benefits, I have less money for discretionary spending or saving. I may choose to maintain my discretionary spending, in which case my savings rate will fall: or I may choose to maintain my savings rate, in which case my discretionary spending will fall. Or I may cut both. Which option I choose is determined by such things as the level of interest rates (poor returns on savings encourage spending), my expectations of future inflation (if I think my savings will be inflated away I may choose to spend the money instead - or I may save even more to compensate), and my worries about the future. In an uncertain economic environment where jobs are under pressure, I am more likely to maintain my savings rate while I can. But if my income falls to the point where my ESSENTIAL spending is under threat, I won't maintain either my discretionary spending or my savings rate.

In Ireland, households are maintaining high savings rates despite reduction in real incomes, and that is causing serious demand problems in the economy, as Part 2 notes in relation to the difficulties experienced by SMEs. The IMF thinks that if households are discouraged from saving, through fiscal deficit reduction by means of tax rises and benefits cuts, they will be encouraged to spend instead. This is bunkum, frankly. Fiscal deficit reduction can only be achieved by taking money from the private sector, which both discourages saving and dampens demand - hardly what is needed in a distressed economy.

But what is really worrying is the IMF's assumption that Irish households' preference for saving will tail off in the face of further fiscal consolidation. You see, Irish households by and large are not saving for future spending - they are paying for PAST spending. Debt repayments fall into the category of ESSENTIAL household spending (not discretionary) when they include things like mortgages and utility bill arrears, and many households also treat unsecured debt repayments as essential spending, because of fear of the consequences of default. Households will cut both discretionary spending and other forms of saving (such as pensions) to the bone to maintain these.Yes, if people are over-paying their debts - paying off debt faster than contractually required - cutting their incomes may discourage them. But so do very low interest rates, which we have now had in the Western world for a long time and yet people are still paying off debt. Clearly debt isn't something people want to have. If their desire to get rid of their debt is stronger than their desire for consumer goods and services, they will continue to pay off debt in preference to buying things. In which case, for debt payments actually to reduce, incomes would have to fall to the point where people are struggling to buy food - and we would then see increased defaults and foreclosures, not simply reduction in savings.

The IMF report highlights the fact that Irish households are highly indebted, with debts of over 200% of income - yet they recommend reducing household incomes still further, through more tax increases, benefit cuts and nominal wage cuts. They note that this will have the effect of reducing Irish households' debt repayments - but they attribute this to unproven economic effects rather than to the obvious cause, which is simply that people will have less money with which to pay off debt. They correctly identify lack of domestic consumer demand as the single biggest issue facing the Irish economy - but they think that if people are unable to pay off debt because their incomes have fallen even more, somehow they will increase consumer spending. And they also think that further fiscal consolidation will encourage people to spend, even though it will remove money from the private sector. Where is the logic in this?

I suspect that the recommendations of this report are actually driven by two unstated factors that have nothing to do with the realities of the Irish economy. Firstly, the fact that Ireland is currently in receipt of bailout funds, and return of this money is top of the IMF's priority list. And secondly, the almost religious belief among IMF staffers in fiscal consolidation as a cure-all for economic ills - rather as mediaeval medics believed in blood-letting. The IMF's recommendations for Ireland appear to be driven not by logic but by wishful thinking.


  1. The IMF's and OECD's policies for any country involves cutting wages and welfare, it's the standard response and I have written about it several times on my own blog. It's about reducing that part of GDP that goes to workers' remuneration, all the more for the capitalists.

    The other policies the august bodies advocate, reducing deficits at breakneck speeds so as to cause recessions and stagnation, will most likely result in debt deflation. The real value of debt will be maintained or even increase.

    Looking at it through my Marxist sunglasses everything seems to be quite logical from the standpoint of the ruling class. Less money for the proles while making sure you squeeze them.

    Marx actually mentioned this in the Grundrisse in the late 1850s, it was happening even then.

  2. Great post, you've done a good job of disecting the report with a critical eye. I agree that you'd think the IMF would do better than that.

    But i have no faith in the IMF personally. They're using the wrong economic models, and they will right up to the point they're made redundant.