Professor Philip Lane:

Before I come to my statement, on my own behalf and on behalf of the Central Bank, I want to join in expressing my condolences. Peter Mathews made a valuable contribution to the important public debate about how to resolve the crisis. I can recall that throughout that period, he, along with others, stood up and became actively engaged in working out different strategies for managing the crisis. That speaks to his sense of civic duty. I and the Central Bank want to recognise that important contribution.

I am here today in my capacity as the chair of the economic statistics review group, the ESRG. This group was formed last summer to provide guidance to the CSO on how best to meet user needs for greater insight into Irish economic activity, taking account of the challenges inherent in providing a comprehensive picture of the highly globalised Irish economy.

The membership included policymakers, analysts, regulators, business and trade union representatives, as well as academics. We also had input from the international statistics community, for example, EUROSTAT, the United Nations, the OECD and the IMF. In addition, the ESRG benefited from written comments and presentations made by various members of the group and also by external contributors. All of this material can be found on the CSO's website which includes the report and all of the ancillary material.

The CSO compiles national accounts and balance of payments data in accordance with global standards, which is essential if results are to be comparable across countries and over time. EUROSTAT has also affirmed that the 2015 national income and expenditure accounts comply with these international standards. However, it is increasingly difficult to interpret the complexities of activity in highly globalised economies by reference to the most familiar indicators such as gross domestic product, GDP, and gross national income, GNI. Reflecting the increasingly interconnected nature of business and its growing fragmentation across national borders, it has become much more challenging to fully understand the impact of globalisation on national statistics. We can think of particular issues relating to intangible assets such as intellectual property, the globalisation of production processes and the residential location of the corporate structures of global firms. That is the backdrop. Given all of this, it is critical to generate reliable measures of the aggregate size of the economy. This is important for private sector decisions. It is also essential for fiscal planning and, from the Central Bank's point of view, in assessing the sustainability of private and public debt levels. For macro-prudential policy, the size of the economy is a necessary scaling factor in assessing the sustainability of the level of credit issued by the financial system.

As indicated, the CSO is mandated by international requirements to produce gross domestic product, gross national income and related measures. However, it is the view of the ESRG that supplementary or extra statistics are more appropriate in the measurement of domestic economic activity. These supplementary indicators need to be accessible and publishable in order that the confidentiality of data from individual firms is not compromised. They also need to be sufficiently robust in order that possible future globalisation-related changes will not damage the relevance of the series.

One core recommendation of the group is the development of what we call an adjusted level indicator, GNI*. Let me explain about the concept of GNI*. A basic principle of national accounting is that in order to have a stable measure of economic performance it should be robust to cover alternative accounting approaches by multinational firms. This is at the heart of one of the strands of the work we did. For decades it has been recognised that GDP is an inadequate indicator in Ireland, given the size of measured factor income flows for multinational firms. For this reason, gross national income or its close cousin, gross national product, has been widely employed since the 1980s as an alternative indicator, since these measures strip out net international factor income flows. The profits paid to multinational firms are stripped out when going from GDP to gross national income. However, we now find that even GNI is no longer a sufficiently reliable indicator. In particular, there are two significant measurement issues, one of which is the treatment of depreciation in the foreign-owned component of the domestic capital stock. The second which has been flagged in recent years is the treatment of the retained earnings of re-domiciled firms. These are firms that have located their headquarters in Ireland and moved a lot of activity onto its balance sheet.

In both cases, measuring Irish economic activity on a gross basis as embedded in GDP or GNI now includes sizeable elements which, ultimately, are not accruing to domestic residents but are borne by or accrue to non-residents.

In the context of depreciation on the foreign-owned portion of the domestic capital stock, that is treated as being part of the overall indicators when it comes to GDP or GNI. This is misleading because who bears the cost of that depreciation? It will not be domestic residents but, rather, the owners of these foreign firms. It should not, therefore, affect a measure that is intended to capture the resources available to domestic residents. This is especially the case if the relocated capital assets are not deployed in combination with domestic workers but are used in combination with overseas workers through contract manufacturing arrangements. For example, one could have intellectual property assets in Ireland which are being used for production elsewhere. Essentially, it is just entering Ireland through the accounts of these multinational firms. It is not something that is visible in the Irish economy.

This is why we have recommended the alternative GNI*, or GNI star, measurement, which, in the first instance, would exclude the depreciation of foreign-owned domestic capital assets. This would prevent corporate re-organisations having an impact on measures of domestic economic activity and would recognise the reality that the ultimate owners of these assets are those who bear the depreciation costs relating to those capital assets.

With regard to re-domiciled firms, it is well recognised that issues arise because the accounting treatment of net income depends on whether a firm is classified as a directly owned foreign firm or as a domestic firm headquartered here which, in turn, is owned by foreign shareholders. Substantially, it should make no difference but, because of global conventions relating to the rules in terms of recording capital income, it does. The current position for a re-domiciled firm is that a firm here might accumulate net global income and it is recorded as part of GDP and GNI. Eventually, it will leave the accounts if the firms declare a dividend. However, many corporations are holding on to cash at present. If they hold on to cash they are accumulating cash on the balance sheet here. It is not flowing out of the economy but, at the same time, it is not really a part of the domestic resource base because ultimately the foreign owners will take the money out. This is essentially the second adjustment. We believe that rather than wait for the cash to be paid out, it should be recorded upon accrual if these firms are making profits. We already do it for foreign direct investment, for this second category of re-domiciled firms. Once the earnings are accruing here, they should be recognised as ultimately accruing to the foreign investor and should not affect important measures such as GDP or GNI. Since we cannot affect the measurement of the GDP or GNI what we can do is create an extra indicator, GNI*. That is the core recommendation regarding measuring the size of the economy.

In addition, in order to strategise for the economy and to think about its structure it is important to learn more about the breakdown between the domestic and foreign sectors in the economy. Of course, any analytical presentation must respect the confidentiality of the data supplied by firms to the Central Statistics Office, CSO, which constrains the level of detail that can be presented. In addition, as I mentioned earlier it is important that any presentation is forward looking and is set up in such a way that it does not rule out future revisions by these corporations in how they do business. Given those constraints, it is never going to be the case that one can go below a certain level of aggregation and, therefore, what can be possible is a set of other splits of the data. Incidentally, this is also relevant in thinking about measures such as the current account balance of the economy and the international investment position of the economy. Having more detailed splits of those data is important for analysis.

A third dimension of the report relates to understanding the cycle. My colleague, John Flynn, is involved in our forecasting exercise. We have to work out whether the economy is overheating or under-heating and so on. More data revision is needed in this area. When forecasting a high frequency activity, the focus is on expenditure measures such as investment, consumption and so on. The problem is that at present the data, including the trade and investment data, are distorted by the massive gross flows related to the activity of multinational firms and aircraft leasing. For practical purposes, those of us in the Central Bank and all other professional analysts have to make adjustments to come to something that is more related to the domestic economy. For example, we try to strip out the impact of intellectual property investment and investment in aircraft that are used elsewhere and so on. Essentially, we are recommending that the Central Statistics Office could publish adjusted measures rather than each individual analyst having to make decisions about these things. This would then be the gold standard measures that domestic and global analysts could examine in considering the business cycle in Ireland. Essentially, it involves netting out the impact of contract manufacturing and royalty payments for intellectual property and so on.

We also have some recommendations for conduct and how the CSO manages the national accounts. One relates to communication. The report identifies recommendations about how to enhance user understanding of Ireland's major releases in terms of the presentations, briefings, questions and answers and so on at the time of major data releases. We note that it is essential to build on the ongoing co-operation at international and domestic level. Internationally, the CSO is plugged in to various global initiatives to grapple with globalisation topics. Domestically, we believe there is room for improvement in terms of the joint activities of the CSO and those of us in the Central Bank in respect of collection of relevant financial data.

Essentially, the message of this report is that the headline numbers are not adequate. They are measured correctly, as affirmed by EUROSTAT, but they do not tell us what we need to know. That is why having these extra indicators is important. It is also important to be realistic in terms of what can be achieved in a given timeframe. The most important point is to have a revision in terms of the annual data and adapting the quarterly data at a later period.