Dr. Donal Donovan:

First, I thank the Chairman and other members of the committee for the opportunity to appear here today. I am happy to provide them with my thoughts on the role played by the IMF in its relationship with the Irish authorities in the years prior to our economic and financial collapse.

I should mention that my observations are offered in a personal capacity. As members of the committee may know, it is relatively rare for IMF staff to be assigned to work in their own countries. I was not involved in the work on Ireland during my IMF career which spanned almost three decades. My thoughts are based on a detailed evaluation of the extensive documentation published by the IMF on Ireland in the build-up to the crisis, viewed in the light of my own experience with the preparation of similar documents for many other countries.

The independent evaluation office of the IMF, which, as its name implies, is an independent outfit and usually staffed by external persons, is currently undertaking an extensive assessment and investigation into the role of the IMF vis-à-visthe euro area in the years both prior to the crisis and during the crisis itself, with special focus on the experiences relating to Ireland, Greece and Portugal. Since I have been asked to work on this project, I thought I should mention it in passing.

By way of background, it may be useful to explain briefly some elements of the IMF's surveillance process over its members. Normally a team from the IMF visits each country for a period of about two weeks each year to discuss the economic situation and outlook and to review and assess the authorities' policies. A report is then prepared, called the Article IV consultation report, which is subsequently discussed by the IMF's executive board on which all countries, including Ireland, are represented. The report is prepared by the staff, not under the direction of the executive board. The report, together with a shorter document called the executive board assessment, is then published. This report is not subject to review or approval-clearance by the country authorities prior to issuance, except for checking for factual inaccuracies. Before being published, authorities may request deletion of what is called market sensitive information. However, this aspect is interpreted rather narrowly and would not, for example, cover deletions of IMF views that might be quite critical of a country's policies.

Between 2000 and 2010, with the exception of 2008, an important point which I will come back to later, consultations took place annually with the Irish authorities. In addition, in 2006 as part of the financial sector assessment process, FSAP, a separate specialised report was prepared and published on the state of Ireland's financial sector, and this report was discussed alongside the normal Article IV report.

In assessing all of this material, I believe it is helpful to distinguish between the period 2000 to 2007, that is, before the crisis broke, and the 2008 to 2010 period when the crisis began to emerge in full force and the content of the discussions was substantively very different. As an overall comment on 2000 to 2007 period, I think it is widely accepted, and it would be my personal view, that the IMF's surveillance process failed in Ireland.  Although, as I will discuss, some vulnerabilities were noted, the assessments by the IMF staff gave no inkling that a major disaster could be in the making. Adjectives such as "exceptional", "remarkable" and " highly impressive" were used throughout the first seven years of the decade to describe Ireland's overall economic performance.

Such overly positive assessments by the IMF were not confined to the case of Ireland and there is no doubt that the organisation did not see warning signals in many industrial countries in the build-up to the global crisis. However, the extent to which a country's economic and financial situation deteriorated so sharply and dramatically with minimal prior anticipation by the IMF, that is, in the Irish case, was to my knowledge probably unprecedented in the history of IMF surveillance. To help understand why such a major misreading occurred, four key areas covered by the consultation reports for Ireland need consideration: the properly market, the financial sector, the budget, and the macroeconomic linkages between these sectors.

First, in the case of the property market, in view of what ultimately transpired, it should be noted that the IMF staff did devote considerable attention to the residential property market. Key indicators, such as the movement in prices relative to rents, the increasing proportion of buy-to-let investors, the historical evidence from other countries' experiences over long periods, and the distorting role of the fiscal incentive regime applicable to property, were all highlighted repeatedly. The thrust of the staff view was that for most of the 2000 to 2007 period, house prices were somewhat overvalued and a speculative element was increasingly present.

Until around 2007, the Irish authorities did not accept the IMF's assessment, arguing that so-called fundamental factors such as the continued rapid increase in personal incomes given Ireland's high growth prospects, inward migration that would boost housing demand, prospective low and stable interest rates, and prevailing low levels of household debt justified the price surges that had occurred. It appears that, in essence, the IMF staff and Irish officials implicitly agreed to differ over this question. In addition, IMF staff had consistently urged from the beginning of the decade the introduction of a property tax and a reduction in mortgage interest relief to help counter a strong pro-house ownership bias. However, in quite candid exchanges in 2004 and 2006, Irish officials, as quoted in the staff reports, referred to "political, like insurmountable difficulties" associated with taking action on these fronts.

Although the message of likely property overvaluation was cited throughout, the IMF staff did not go so far as to suggest the existence of a possible bubble and did not speculate on the size or timing of a possible drop in prices. However, in 2007, the staff report explicitly endorsed the authorities', that is, the Central Bank of Ireland's, view that a soft landing was the most likely outcome. The CBI's judgment was rightly described later by the Honohan report - I should mention again, as the Chairman did, that I was a member of the Honohan inquiry team - as a "triumph of hope over reality". This cautious stance by the IMF may have been due to several factors, some of which may have operated subconsciously: difficulties in challenging definitively the Irish authorities' fundamentals argument; a fear of being seen to cry wolf too often, since the overvaluation thesis had been raised every year since the start of the decade; and also, perhaps, concerns about adverse market and perhaps public-political reactions.

Finally, a major weakness of the IMF's analysis was the neglect of the commercial property market. This aspect was not really dealt with in the CBI's financial stability reports either.

However, it turned out to be a crucial, perhaps the most crucial, element explaining the extent of the banks' later losses. My own jointly authored book on the Irish crisis with Professor Antoin E. Murphy, to which the Chairman kindly referred, pointed out that the surge in lending to developers was actually substantially greater than that to householders, especially when lending by foreign subsidiaries was included, and this latter category of lending was highly significant when, in the end, the bill came to be presented to Irish taxpayers. As an eminent scholar of financial crises, Charles Goodhart, once pointed out, "Banks are international in life, but are national in death". When the bill has to be paid, everything comes home to the country where the bank is located.

The second area is the state of the banks. Analysis of the financial sector always featured in the IMF reports but was given much more intensive attention during the 2006 FSAP update exercise. This involved a team of financial sector specialists undertaking a separate, more in-depth, look at the state of the Irish banks. Somewhat sad to say, their overall assessment was reassuring, if not indeed rosy. While, as with the property market, vulnerabilities were mentioned, the typical message was that the outlook for the financial sector was positive and that banking sector profitability and capitalisation were strong. The was no hint given as to the possible occurrence of a major problem, let alone of the potential for the crisis that eventually unfolded. As Governor Honohan has rightly stressed in his report and elsewhere, this clean bill of health from the IMF came at a particularly inappropriate time as it undermined whatever concerns might have been voiced internally.

Why did the IMF team get things so badly wrong? There are several explanations. First, the analysis was grounded largely on acceptance of the CBI view favouring the soft landing outcome for house prices. While some stress tests involved larger price falls, the possible rise in non-performing loans, NPLs, was, it appears, capped at 5% of all mortgage loans. Second, the neglect of commercial property lending was a crucial omission, Third, while the increased dependency on wholesale external funding by the banks was noted, no one considered what might be the catastrophic effect of a worldwide drying up of liquidity as actually occurred in 2008. Finally, the IMF team appeared to have had no inkling of the indecisiveness and lack of firm engagement underlying the detailed interaction between the Financial Regulator and key individual institutions, problems that were uncovered only much later by the Honohan inquiry.

These were serious shortcomings to which can be added the general approach at the time that favoured so-called principles-based, what is sometimes called light touch, financial regulation. The IMF, being a creature of its member countries, was undoubtedly heavily influenced by this prevailing philosophy. That said, in my view, it would have been incumbent on the FSAP report, as a minimum, to have posted a health warning and cited more explicitly the limitations that underlay the positive conclusions it presented. The Irish experience suggests that the absence of such warnings can seriously undermine the credibility of the IMF's work.

The third area is the budget. The last major area covered by the consultation reports was the budget. Until 2008, Ireland had been running small overall budget surpluses. However, the IMF staff generally urged that these surpluses be increased somewhat, both to counteract what was thought to be overheating, often described as using contra-cyclical policies, and to build up a reserve against future unknowns. By and large this recommendation fell on deaf ears.

A far more serious shortcoming in my view was the conclusion by the IMF up to and including 2007 that the underlying, that is, cyclically adjusted, fiscal balance, CAB, was in approximate balance throughout 2007. This CAB measure attempts to strip out so called temporary factors, such as higher than average growth or transitory revenue flows, that mask the true underlying fiscal picture. In Ireland's case, the IMF, together with the Department of Finance, went along with a common EU methodology used to calculate the CAB.

The problem was that this methodology assumed that the high output levels reached by Ireland in the first half of the decade of the 2000s, which in turn reflected the massive reliance on the construction sector, were permanent structural features. The same assumption applied to the artificial boost in revenues associated with the property boom. Using more technical phraseology, it was assumed that actual output was close to potential output, but the reality was the other way around. Irish output throughout the latter years of the boom was far above sustainable potential. After all, as was pointed out by very few at the time, there was a limit to how many homes people can actually live in. By 2009, the assumptions underlying the earlier IMF calculation of the CAB were clearly untenable. In a quite dramatic reversal, the 2009 IMF report re-estimated the CAB for earlier years using a quite different methodology. For example, the CAB for 2007 turned out to be a deficit of 8.7% of GDP compared with an originally estimated surplus of 0.7%, a change of more than nine percentage points in GDP for the same year. Seldom has the picture of a country's fiscal health deteriorated so sharply and so quickly. The question can be asked if, starting from 2009, a far more appropriate methodology was used, why this was not done in earlier years or at least presented as a variant of the standard approach that had been uncritically accepted.

The last area concerns the overall macroeconomic interlinkages. The various IMF reports did point out to some extent the vulnerabilities associated with particular sectors, that is, property, financial and fiscal, but they explore the dynamics of a possible downward self-reinforcing spiral such as eventually ended up happening. At best, some first round effects were considered. While precise quantification of a worst case scenario would have been very difficult, some key elements could have been addressed more explicitly. It is possible that the IMF reports did not go down this route because of the somewhat speculative nature of what was involved. It would likely also have been seen as highly alarmist, and provoked strong negative market and, one assumes, political and media reaction. Nevertheless, that complication could have been dealt with by raising the issues confidentially with the Irish authorities as opposed to including a discussion in the published staff report. However, there is no evidence available indicating that such discussions occurred. There could have been pressures both within the IMF andvis-à-visthe Irish authorities to dismiss the possibility of such very negative outcomes. The consensus, reflecting perhaps strong elements of groupthink, was to stay with the soft landing hypothesis and to hope that perhaps, in the end, our luck would hold out.

I will turn briefly to the second period, 2008 to 2010, when, of course, things were quite different. The 2007 report was the last rosy IMF report on Ireland. By the time of the 2009 consultation two years later, the picture had changed dramatically. The property market was in free fall, the budget deficit had exploded, unemployment was soaring and the full extent of the banking disaster was starting to emerge. However, no consultation took place in 2008. Normally, the consultation would have taken place as scheduled, unless the authorities indicated a desire to postpone it. The reasons underlying this hiatus are not in the public domain. The postponement meant that during this critical two-year period, from mid-2007 to mid-2009, there was no formal dialogue between the IMF and the Irish authorities. This must be considered a significant flaw. If IMF surveillance is to be meaningful, there should be at least the opportunity for timely inputs from the IMF at a time when, amidst global financial disarray, many key policy options were being considered on the Irish side.

In particular, there are no indications, at least on the public record, suggesting that the IMF staff provided input on the end of September 2008 bank guarantee, either before this decision was taken or afterwards. The 2009 consultation report described the guarantee but did not offer any views as to its appropriateness or otherwise. It did, however, contain a useful table summarising the key features of guarantees provided by various other countries during the past 30 years. This table is summarised as part of chapter 10 on page 214 of the book Dr. Murphy and I wrote. The chapter deals extensively with the guarantee decision. Perhaps contrary to what is sometimes said, it appears from this table that the coverage of the Irish guarantee was not that radically different from that contained in several other earlier guarantees.

Overall, it seems that around this critical time, the IMF did not provide sufficient timely professional technical advice to the Irish authorities. Whether this was primarily a supply side issue, perhaps because the IMF was busy elsewhere, or reflected demand side factors, because perhaps the authorities preferred not to hold the consultation in 2008, remains an important question to which I do not have the answer.

I thank the Chairman and committee members for their attention. I am, of course, very happy to answer any questions they may have on the foregoing or on any other related aspects covered in our book or elsewhere. I received a list of possible questions that committee members may raise and I am certainly happy to do my best to try to answer them as far as I can.