A few months ago, disappointed with the measured growth in US employment, the current occupant of the White House fired the head of that country’s most important statistics agency. Financial markets reacted measurably to what was perceived as an increase in uncertainty about the future quality of official statistics. An ingenious study published last week puts the estimated social cost of this event at 20 times the total budget of that agency.
In Ireland, despite living in one of the most hard to measure and volatile of Europe’s economies, we have sensibly avoided shooting the measurer. For example, the sharp reported declines in Gross Domestic Product (GDP) in the past two quarters have, for good reason, not disturbed the body politic, busy as its members are instead downloading fragments of family history gleaned from the newly released 1926 Census returns.
Here it seems to be generally understood that accurate information is essential for good policy choices. The Central Statistics Office (CSO) has been happily free of political interference.
But statistical agencies face other challenges: for example the diminishing willingness of households and firms to respond to official surveys, and a growing sense that many of the concepts that have been used for decades to measure economic activity and wellbeing don’t capture the economy of today.
READ MORE
It’s not just the particular data distortions associated with globalisation and international tax competition that have made GDP irrelevant and indeed misleading for Ireland. As Diane Coyle illustrates in her new book The Measure of Progress, it’s a worldwide problem: less and less of the economy is easily measured by established methods. There are, to be sure, new sources of information that can be scraped from the internet, but most of them do not map easily into the standard concepts.
The first director of the CSO back in the 1950s was Roy Geary, a somewhat eccentric genius who is being remembered next week by the Statistical and Social Inquiry Society of Ireland, a long-established body of academics and public servants interested in quantitative approaches to economic policy.
Geary was an academic manqué. Despite his day job as a civil servant in the Statistics Branch of the Department of Industry and Commerce, he contributed an impressive sequence of papers to the most important international statistical journals, solving technical problems that had puzzled leading global experts. How he managed to do this while holding down the more mundane day job of assembling and updating Ireland’s official statistics is difficult to imagine. These were not casual opinion pieces, but involved closely argued mathematics: one of his papers contains no fewer than 88 equations in just 19 pages.
In a more practical way, Geary came to believe that official statistics should come with a health warning flagging the uncertainty around how good the measure was: A for very accurate, D for a best guess. Perhaps this idea is worth reviving.
Even if there were a good measure of GDP, it is increasingly evident that it has become less appropriate as a single indicator of economic activity or wellbeing. Well-known deficiencies, such as the omission of pollution and carbon emissions, have become more salient.
[ History signals more hawkish central bank response to Iran war price shockOpens in new window ]
Neglect of the degree to which economic activity is unequally distributed is unsustainable in a world of unprecedented concentrations of wealth and economic power. And theoreticians are still debating how to deal with new types of economic activity (for example, “free” internet searches, AI), for which there is no settled conceptual measurement framework.
The CSO obviously stays abreast of the international debates on these matters, but the fact remains that you can be quite misled if you don’t realise the shortcomings. This becomes increasingly important at a time of rapid population growth when big decisions need to be taken to ensure public infrastructure and services can meet the additional demands.
Collecting relevant data in a consistent and credible manner is not all that is needed. The data needs to be interpreted and interrogated. That’s where the analysis of think tanks, public bodies and universities becomes crucial. In Ireland, for example, the Fiscal Advisory Council has been assiduous in warning the Government not to take misleading GDP figures as a reliable guide for budgetary policy.
The first such think tank in Ireland, the Economic Research Institute (now the ESRI), began as an initiative of the Statistical Society mentioned earlier. Recognising the need for such a body, it was the civil servant TK Whitaker (who is also being remembered on the 50th anniversary of his retirement), who got the society to secure seed funding, and persuaded Geary, who had taken early retirement from the CSO and was leading the United Nations’ work on GDP statistics, to come back to Ireland in 1960 to become the institute’s first director.
Geary was a vocal critic of economic theory, especially of the free-market variety, seizing every opportunity to castigate the profession’s members (Nobel laureates included) typically in his booming voice from the front row of a seminar’s audience. He wanted practical measures to answer important social questions and he would have been an active participant in the current debates about economic data and measurement.
Population growth was one of his favourite topics. Before the second World War, with Ireland’s population continuing to fall, he made a bold prediction that this would not always be so, reinforcing his view with a statistical model. Fifty years later, his relative optimism proved to be justified. He would have relished updating such an exercise today.
For all his quirks, Geary had the right idea about independent data analysis. Not all think tanks strive to be data-driven. In the United States, many such entities are funded by commercial or political interests with an axe to grind. They engage staff who begin with the desired policy recommendation and work back from that. Worldwide, political choices are increasingly being influenced by superficial “truthy” populist nostrums based on false data analysis. Let it not be the case here.
Patrick Honohan is former governor of the Central Bank of Ireland
