Remarks by Patrick Honohan at an event marking the retirement of Professor John FitzGerald

22 October 2014 Speech

‘Carelessness or misfortune?

Lessons from Our Generation’s Two Macroeconomic Collapses’

Two deep national macroeconomic crises, each following suddenly after a change of exchange rate regime, have defined the working life of John FitzGerald’s generation of Irish macroeconomists. Separated by a decade and a half of growth and rapid recovery, the two crises are often spoken of in the same breath, sometimes to say that they were but variants of the same problem; sometimes to highlight the differences.

So which is it: should Ireland, once bitten, have been sufficiently shy to avoid the second, or was it really quite different? And even if there are similarities, is there a danger of neglecting important differences that, if understood and learnt from, would help improve the recovery process that is still under way and help avoid an early recurrence?

In some important respects, it is the similarities between the two episodes that jump out: spending patterns that exceeded the economy’s ability to generate output that could pay for what was spent; policy errors based on exaggerated reliance on bowdlerised versions of simplistic paradigms (a naïve caricature of Keynesianism in the first instance, a credulous adoption of intensive financialisation in the second); an almost reckless disregard for the fiscal consequences of policy decisions. And yes, the experience of the past decade has been yet another example of the start-and-stop pattern of the hare-and-tortoise metaphor used more than a decade ago by some other old-hands to describe the previous half-century.

But the dynamics of the two crises differed a lot. The second crisis happened faster and resulted in a deeper loss of output and of employment, especially of men but with women also affected much more and for longer than in the earlier episode. Public debt jumped to a roughly equivalent proportion of GDP, but did so from a much lower base and much faster in the recent episode. Asset prices moved sharply in both cases, but were far more important in the more recent episode.

The boom before each new crisis tends to display subtle differences from what has gone before. It is these differences that can blind policy makers to the vulnerabilities that are building. In the recent crisis, over-simplified approaches to assessing fiscal vulnerability, and a formulaic and credulous approach to bank solvency and bank credit risk, created risks that are still being worked through. Although the denouement revealed the source of the problem to be relatively obvious, it has been less easy to pinpoint the behavioural flaws (in the behaviour of lenders, regulators and borrowers) that prevented enough of them from foreseeing the risks and mitigating them.

Some of the differences are attributable to the prevailing international economic environment—always crucial for as open an economy as is that of Ireland. Some are due to the degree to which the pre-crisis economy had become distorted before the crash. The distortions were surely much greater in the recent case because of

(i) the heavy concentration in the boom phase on the construction sector (reflected in the sharp fall in male employment);

(ii) the skewed reliance of the public finances on volatile sources of tax revenue (this resulted in a sharp and unforeseen loss of tax revenue when the boom ended); and

(iii) the grossly imbalanced private sector portfolios, which triggered a wave of progressive losses of confidence, one of the earliest of which triggered the serious unforced policy error of enacting an all-embracing blanket banking guarantee, and a later one ensured that the authorities had no alternative but to seek funding from official sources including the IMF.

While the macroeconomic adjustment of the economy in both cases shared common features: re-balancing of the public finances using both tax increases and spending cuts, the adjustment in the more recent case began sooner and was being accomplished more quickly. This may be attributed to the loss of international confidence forcing the Government’s hand in line with the adjustment programme agreed with the official lenders. Wage competitiveness was also a central element in the adjustment process. This was helped in the 1980s by the new weakness of the currency regime relative to that of the largest trading partner (the UK), and by the ability, exercised in 1986, to make a significant but one-off devaluation, which was an option that was not available in the euro area.

An additional feature of the later crisis, less significant in the earlier case, is the impact of the boom and bust on wealth and debtor positions. Part of this is widely discussed, namely the widespread presence of households with negative housing equity, the financial and operational crippling of the banks, and the assumption by the Government of a large block of banking liabilities. Another part is less easy to measure or even to unravel conceptually, namely the sizable consequences of the debt-financed property purchases on spending and wealth accumulation: I hope we will soon be in a position to throw more light on the magnitudes involved.

I. Causes

One should always distinguish between proximate and underlying causes in considering major conjunctural disturbances.

(a) Proximate causes

For proximate causes we can easily accept the important commonality and the main distinction between the two episodes. Both were caused by the emergence of multi-year overspending: the distinction lying in the fact that the 1980s followed a period of mainly public sector overspending, whereas it was private overspending that teed-up the recent crash.

The government over-spending of the late 1970s is generally and rightly traced back to the winning election manifesto of 1977. This heralded a post-austerity splurge, following the sharp fiscal contraction that had corrected emerging imbalances in 1975. The expansive policies of 1977-81 set in place pay rates and spending programmes which were politically difficult to dismantle when they proved unaffordable.

(b) Underlying causes

But why did these policy errors emerge? Here it is useful to recall Keynes’ observation that policymakers are often in thrall to some defunct theoretician. In both the 1970s and the early 2000s, Irish governments were misled by bad ideas falsely distilled from simplistic versions of prevailing orthodoxies.

Thus, in the 1970s it was a bowdlerised deformation of Keynesianism that was used to justify a government spending-led growth spurt which was intended to generate self-sustained revenue expansion. The implicit parameterisation of this idea was wildly wrong.1

In the 2000s, in contrast, the Government was lulled into complacency by an apparently crass belief in untrammelled financialisation. Indeed, Ireland’s experience in this crisis differs from the experience of several other financialised economies mainly in scale. Of course a well-developed financial system is an essential contributor to balanced economic growth. But it was excessive reliance on financialisation that allowed private spending (including by developers abroad) to create balance sheet vulnerabilities threatening the stability of the state, and deluding it into thinking it was running a sustainable fiscal policy, when in fact it too was building up unsustainable commitments financed by insecure and transient revenue sources

II. Contrasts in the evolution of the main macro aggregates

Both recessions were extremely deep and protracted. But there are interesting differences in timing and dynamics. Take just four indicators: employment, unemployment, asset prices and government debt.

(a) Employment: Faster and deeper fall and for both men and women in the 2000s

Here the interesting contrast is across the gender divide. In the 1980s, the adverse shock to employment produced an overall fall of about 7 per cent over 8 years. The decline in female employment was much smaller. The big movements were in male employment, which fell further and did not fully recover until 1996.

There was a huge fall of 20% in male employment between 2007 and 2012; female employment also fell sharply in this period (8½% decline peak to trough). These were faster and deeper falls than in the 1980s. The employment recovery in both male and female employment started in 2012, but total employment is still about 12½% below the peak at the time of writing.

(b) Unemployment: Tracking the UK in both periods: but went much higher in the 1980s

With migration as an important balancing mechanism, trends in unemployment in Ireland do not always closely track those in employment. One astonishing regularity, which has persisted through both crises, is the close manner in which the dynamics of Irish unemployment has tracked that in the UK (and not that in the euro area).

The surge in unemployment rates was higher in the 1980s, despite a smaller dip in employment. Unemployment had already been high during the 1970s, and the late 1970s recovery was moderate. Rates surged to record levels against the background of the contemporary recession in UK. Indeed, there was a double dip in unemployment rates into 1992.

In contrast, the 2008 crash had been preceded by a 6-7 year period of full employment (measured unemployment rate around 4%). Despite the contemporary global financial crisis and the euro area crisis, Irish unemployment rates peaked below 1980s levels and, helped by stronger labour market developments in UK, started to turn down from 2012.

(c) Asset prices: Much larger swings in the recent crisis

Here there is a big contrast. The early case did not show huge shifts in asset prices. 1980 marked the peak of a house price boom; equities also peaked in that year, but these peaks were small relative to what happened in the 2000s. Real house prices hardly moved up before 1977 and then only by about 25% total. Likewise they didn’t dip much below the 1970s median.

In the 2000s, the house price bubble (even though starting from a much higher plateau in real terms than that of the 1970s) was very large and was accompanied by record levels of construction. Interestingly, five years after the crash, house prices had turned well before reaching the historical average in real terms. (Has the low euro nominal interest rate regime permanently lifted equilibrium house prices? Are real costs of construction permanently and unavoidably higher? Or are prices still elevated relative to their likely long-term trend? I won’t speculate further.)

Equity prices also fell sharply (especially banking equities on which some older middle-income people were relying for dividend flow to sustain living standards). The bursting of the bubble has had other long term wealth distributional effects.

(d) Public Debt: Slow build-up in the 1980s, sudden jump in the 2000s

Here the contrast in dynamical sequence is quite interesting. The 1980s crisis was characterised by a slow but steady accumulation of debt during the crisis containment phase until the political logjam was broken and decisive corrective fiscal action was taken. Indeed, debt had not yet got seriously out of line when the crisis began to be tackled, albeit too tentatively, in 1981. It is relevant that most debt was domestic and in local currency. This contributed to limiting fears of default, 2 and indeed permitted successive governments to defer decisive adjustment until the debt ratio crept up to the vicinity of 120% of GDP.

In contrast, in the more recent case, public debt was falling until just before the crisis was identified and containment began. Then it started growing rapidly (even before the banking debt was taken onto the books). The debt at that stage was almost all held by non-residents and, of course, none in a “domestic currency”. The sudden and belated international awareness of the scale of banking losses being crystallised onto the Government’s debt during the third quarter of 2010 contributed strongly to the rapid loss of international market confidence in the weeks that followed, leading to the application for financial assistance from the IMF and EU

III. Concluding Remarks

Every crisis is different, and we have shown that the two boom-and-bust cycles that have bookmarked the middle part of John FitzGerald’s career as Ireland’s pre-eminent macroeconomic modeller display, as we have seen, striking differences between them both as regards their causes and the dynamic evolution of macroeconomic variables.

But after every crisis, home-spun truths reassert themselves. Living within one’s means, exercising prudence, not running before you can walk. In the end, Lady Bracknell would have been justified in censuring Irish policymakers as she censured her prospective son-in-law Ernest.

Curing the current crisis has been in some respects more challenging than the last time, given the absence of some previously available tools and because of the greater complexity of resolving the indebtedness problems that have been much more to the fore this time around.

Policy has been tackling the challenges and they are now evidently beginning to be brought under control and subdued. Lessons were learnt from the earlier crisis, as exemplified by the more speedy fiscal policy response. The extent to which private debt issues were much more important this time was also recognised and resulted in sweeping legislative changes on insolvency and progressively more assertive regulatory intervention to ensure more pro-active and effective management of the problem by lenders. Thus the public authorities have been acting on the basis of lessons learnt from experience.

We are still learning. In particular, for example, the case-by-case resolution of situations of chronic over-indebtedness has proved difficult. And a full analysis of the distributional consequences remains elusive3 as does a convincing assessment of the longer term consequences.

But what is undoubtedly true is that our national ability to achieve improved policy responses now – and in the future – has been greatly enhanced by an improved understanding of the macro-economy to which John FitzGerald’s tireless and immensely well-informed research has greatly contributed.


1 See Ó Gráda’s 2012 Whitaker lecture for a vivid table showing how badly the plan for 1977-81 missed its unrealistic targets.

2 The distinguished commentator Rudiger Dornbusch was one of the few to suggest that a default could be needed, but already by the time he was writing (1988) the corner had been turned and growth was beginning to take care of debt sustainability concerns.

3 The evidence (some of which I have reviewed elsewhere) is that the impact of boom and bust on the distribution of income seems to have differed less as between the two episodes, after account is taken of the redistributive effect of fiscal policy.