The Ascent of "Datanomics"

The Case of the European Union

No region of the world embodies better the emergence of data-driven societies than one of its most recent political organizations: the European Union (EU). The European Community (which became the EU in 1992) was founded in 1957, based — with good reason — on the mistrust of politics. Rereading French foreign minister Robert Schuman's declaration of May 9, 1950, there is little doubt that this EU founding father saw unabashed political power as a threat to peace. In his eyes, it was necessary to deprive European countries of the means to destroy their fragile postwar peace, even if this meant that democracy had to be constrained. The constraint in question came in the form of economic rules embedded in the Treaty of Rome (1957), which morphed with the Maastricht Treaty (1992) into quantitative criteria used to constantly monitor and evaluate member states from the moment of their admission into the various circles of European integration.

The EU is today largely governed by numbers. The Maastricht Treaty is the economic constitution of contemporary Europe and in it the political power of data is very tangible: ratios govern the admission of countries into the EU and the euro area and are supposed to guarantee the continent's stability, unity, and prosperity. Countries wanting to join the euro area when it was devised in the early 1990s had to conform to five criteria of convergence, three of which still govern the region's cohesion: inflation of no more than 1.5 percentage points above the average rate of the three EU member states with the lowest inflation over the previous year; a national budget deficit at or below 3 percent of gross domestic product; and national public debt not exceeding 60 percent of GDP. These rules have been only marginally modified in the past two decades. The European Central Bank (ECB) was assigned an inflation target of below but close to 2 percent when it was created in 1998. The Stability Pact governing fiscal policy includes a sanction procedure for "excessive" deficits and debt — over 3 percent and 60 percent of GDP, respectively — and forces governments to commit themselves to reaching "the medium-term objective of a budgetary position close to balance or in surplus."

This project is officially motivated by the goal of paving the way for a better common future for European peoples, but the Stability Pact and the ECB statutes give priority to "price stability" and "fiscal sustainability," even if this means reducing governments' ability to deliver economic dynamism and employment expansion. Some indicators have been given prevalence over others out of political choice. But it is more and more clear that the intermediate objectives that prevail (fiscal balance, currency strength, price stability) are at odds in practice with the attainment of the ultimate social objectives (such as employment) that matter the most to populations.

Countries belonging to the euro area, the most integrated part of the European project, have relied on these economic rules to govern their common policies. It can be said that the regulations have not served them well. If they have brought about an apparent culture of discipline, they were not able to create a lasting culture of cooperation: the "great recession" of 2008–2009 triggered a crisis that has, in recent years, turned into a lack of political trust among European citizens. While economic discipline through numbers was supposed to ease peacetime relations, it has instead created divergence among nations and conflicts within their borders.

The early twenty-first century is emblematic in this regard. To date, European countries have been unable to cooperate in an effective and timely manner to build a coherent economic response to the 2008 crisis, straining their monetary union to the point where dissolution became an option. Without the intervention of the ECB, which has implicitly rewritten the European economic constitution and managed to salvage the euro in the summer of 2012 by overreaching its mandate, the European single currency might have imploded. The lack of resolve in 2008–2009 and the return to the enshrined reflex of restrictive austerity policies in 2010–2011 and onward has, in turn, further worsened social conditions in the EU and fueled political distrust.

The evolution of pre- and post-crisis unemployment rates is one possible illustration of the difference between a society governed by pragmatic economic policy and one governed by rigid economic rules. (See figure 1.1.) Both the euro area and the United States started with an almost identical rate in 2009. By 2016, unemployment in the former was twice as high as that in the latter.

These contrasted outcomes, directly related to different policy responses, are anything but coincidental: they are determined by the respective economic constitutions of the United States and the European Union. While the Federal Reserve's core mission is to "conduct the nation's monetary policy by influencing the monetary and credit conditions in the economy in pursuit of maximum employment, stable prices, and moderate long-term interest rates," the ECB must first and foremost "maintain price stability," which is narrowly defined as "a year-on-year increase in the Harmonized Index of Consumer Prices (HICP) for the euro area of below 2 percent." This difference explains why, in the face of the first signs of recession in 2008, the ECB increased its interest rate, compounding the shock instead of buffering it, while the Fed had already started to decrease its policy rate. As we have seen, the same kind of constraint weighs on fiscal policy and explains why the euro area imposed a negative fiscal shock on its ailing economy while the United States was putting in place the largest fiscal stimulus of its history. Because of the power some indicators have acquired on European economic policy, the culture of discipline has stifled the culture of cooperation and Europeans have paid a high price for it.

Of course, certain countries, most notably Germany, have been able to restore some collective protections that have alleviated the ultimate social impact of the macroeconomic shock, but most EU member states were severely hurt by the 2009 recession, which has actually shaken and damaged the very core of European integration. What is more, it also resulted in an increase of support for populist parties openly hostile to European integration, a support that was clearly visible with the European elections of May 2014 and continues to upset political systems all through Europe, as it did in June 2016 when British citizens decided to exit from the European Union. The current wave of European populism is a direct consequence of the distrust toward European integration.

In a context of economic insecurity and identity anxiety, populism is indeed becoming a widespread feature of the European political landscape. From France to Italy, Norway to Greece, Finland to the United Kingdom, populism is not only making direct political gains but also influencing moderate parties in power through agenda-setting on immigration and security. France is a case in point, with right-wing President Nicolas Sarkozy (2007 to 2012) and then left-wing Prime Minister Manuel Valls (2014 to 2016) embracing part of the far-right political agenda on nationality reform and law-and-order policy.

This political crisis has a lot to do with the "automatic pilot" mode dictated by the inscription of economic rules in the EU's fundamental texts. In such a pseudo-federal system, political deliberation should be about how to finance and produce the public goods that benefit all European citizens, not respecting abstract ratios regardless of real economic conditions. Those European public goods are far more important to the welfare of European citizens than the scrupulous observance of doctrinal criteria of budgetary and monetary stability, which, at their best, constitute intermediary objectives for economic policy and, at their worse, prevent the final objectives from being reached (as when inflation is given preference over employment). It is true that member states retain a great deal of power in the EU, more power, in fact, than either the European Commission or the European Parliament have today. But their political bargaining power is bound by the EU Treaties and their quantified commitments.

There is a real European paradox regarding economic indicators since the great recession: on the one hand, the EU has tried to capitalize on the discontent with standard economics and to embrace the "beyond GDP" agenda; on the other, it has become even more rigid in applying its ill-advised targets. The communication "GDP and Beyond: Measuring Progress in a Changing World," released in 2009 by the European Commission, described ways to improve indicators in order to better reflect societal concerns. The commission's intent was to adjust and complement GDP with indicators that monitored social and environmental progress. Yet, so far, the new strategy has not produced tangible results in changing European governance, which still relies heavily on conventional economic indicators. What is more, while the European Union was putting forward these new indicators, it was also coercing the Greek government, from 2010 onward, into reaching European ratios in a time of recession, delaying the country's economic recovery by several years, imposing extremely difficult social conditions on the Greek population, and creating perilous political tensions among member states.

As the European example makes clear, democracy ends up at risk when too much confidence is put by policy makers on too-narrow indicators. We now need to define simple but robust criteria of good and bad economic indicators (and their usage). These will help us to understand why GDP is fast losing its relevance in the twenty-first century.

Good and Bad Indicators

The Case of GDP

How can we know with a fair degree of certainty that governing twenty-first-century economies with conventional economic indicators such as GDP is a bad idea? We first need to learn more about the qualities and flaws of various economic indicators. An indicator is a simplified representation of a complex social reality. It can serve as both a policy input and a policy outcome. It can aid in the design or the evaluation of a given policy, or both. It is used to perform three main tasks: to know and understand, to administer and govern, and to communicate and represent. An indicator should thus possess three fundamental attributes: formal quality, policy purpose, and political impact.

As for its formal qualities, a good indicator must be accurate, timely, regular, comparable in time and space, disaggregative (that is, it should have a certain granularity), and evolutive, to allow for improvements. But it also needs substantive qualities. A good indicator should be purposeful (that is, measure precisely what it was designed for), policy-relevant, to inform on the real complexity of the social world, and socially appropriable, which is to say, both understandable by citizens and subject to public debate.

Obviously, there are trade-offs between these qualities: purposefulness often contradicts availability, appropriability can hinder formal qualities, and regularity can pose a problem for relevance. Two major types of policy indicators offer a glimpse of ways to deal with these dilemmas. Composite or synthetic indicators offer a one-dimensional view of the social world: they are easily readable and understandable and are comparable in time and space. However, they inevitably bring about issues of data aggregation and weighting, both of which relate to the need to combine heterogeneous variables such as income (expressed in monetary units) and health (expressed in years of life expectancy) in a single piece of information. This is by no means technically impossible, but necessitates the use of specific statistical techniques. The second of these policy indicators, the dashboard, allows the user the freedom not to choose between these different dimensions by adopting a multidimensional approach, which retains the maximum possible amount of information for action. Assessing environmental quality, for example, might mean quantifying air quality, water quality, climate, and other dimensions that are not akin to one another. This, in turn, brings about other problems such as data heterogeneity, hierarchy, and comparability in time and ranking in space.

There is, however, less contradiction between composite indicators and dashboards than is often understood. Figure 2.1 shows how dashboards can be conceived as a step in building a composite indicator. In this theoretical example, an environmental quality indicator is designed with only two dimensions, air quality and water quality, themselves aggregating two subdimensions. The key issue in the choice between a composite indicator and a dashboard is the possibility and relevance of aggregation of the subcomponents, as statistical information is being lost in this process and dimensions become reducible to one another. Choosing between a dashboard and a composite indicator is essentially knowing when to stop aggregating data.

This issue is one of the many raised by the gross domestic product, or GDP. From a technical standpoint, GDP can be defined as a composite indicator measuring marketable and monetized economic activity. It uses monetization of its components' value at their market price and imputation of the value of those not sold on markets to aggregate many dimensions of economic activity, some positive (nutrition, health, education, environmental protection), others less so (arms expenditures or prisons, items sometimes referred to as "regrettables"). By conflating positive and negative elements, GDP not only muddles the social picture, but hides from policy view important evolutions in well-being and sustainability by giving the illusion that each of its components is in a good place because its aggregated sum is growing. GDP is thus often confused with well-being and even sustainability, while in reality it says very little, if anything, of either.

The context in which indicators are born is often an indicator itself of their strengths and weaknesses. GDP was first developed in 1934 by Simon Kuznets at the demand of the US Congress, whose members wanted to have a clear and synthetic view of what had happened to the American economy after the 1929 stock market crash. What had happened was the Great Depression. The prehistorical GDP built by Kuznets had the great merit of showing the huge contraction of almost all sectors of the American economy in the first years of the 1930s. Because the shock was systemic, a synthetic indicator was fully relevant and called for a systemic macroeconomic response, something President Roosevelt eventually achieved with the New Deal and the war effort.

Putting the first insights of the nascent pre-Keynesian macroeconomic discipline to good use, Kuznets offered three simple ways to calculate GDP, each time relying on the monetary value of goods and services produced in the country over a given period. The first was through the production approach: the aggregation of the added value of all monetized economic activities in a year. The second, equivalent measure was through the aggregation of all income distributed in the economy (i.e., profits + wages taxes), and the third, the sum of all expenditures in the economy (summing all the components of demand). Although Kuznets invented the process, he did not invent its name: it was Clark Warburton who first used in print the term "gross national product," which Kuznets later adopted.

Therefore, interestingly, GDP was not born as an indicator of development but as a symptom of crisis. Even more telling, Kuznets was keenly aware that GDP relied on debatable methodological choices, writing, "The national income total is thus an amalgam of relatively accurate and only approximate estimates rather than a unique, highly precise measurement." Going further, Kuznets warned of the "uses and abuses of national income measurements":

The valuable capacity of the human mind to simplify a complex situation in a compact characterization becomes dangerous when not controlled in terms of definitely stated criteria. With quantitative measurements especially, the definiteness of the result suggests, often misleadingly, a precision and simplicity in the outlines of the object measured. Measurements of national income are subject to this type of illusion and resulting abuse, especially since they deal with matters that are the center of conflict of opposing social groups where the effectiveness of an argument is often contingent upon oversimplification ...