This is an unedited version of my article for Village Magazine, February 2014
The Global Financial Crisis and the Great Recession are actively reshaping the public discourse about the ways in which we analyse social phenomena, and how our analysis is shaping public policies choices.
In many ways, these changes in our attitudes to social inquiry have been positive. For example, more critical re-appraisal of the rational expectations-based models in macroeconomics and finance have enriched the traditional policy analysts' toolkits and advanced our understanding of choices made by various economic agents and governments. Shift in econometric tools away from those based on restrictive assumptions concerning underlying probability distributions and toward new methods based on more direct integration of the actual data properties is also underway. The result is improved analytical abilities and more streamlined translation of data insights into policy famework. The launching of the public debates about how we teach economics in schools and universities and how economic parameters reflect social and cultural values (as evidenced by the ongoing debate at the OECD and other institutions about introducing measures of quality of life and social well-being into economic policy toolkits) are yet more examples of the longer-term positive change. Absent such discussions, the entire discipline of social sciences risks sliding into complacency and statism.
However, in many areas, changes in our approaches to social studies have been superficial at best, and occasionally regressive. And these changes are not limited to economics alone, spanning instead the entire range of social sciences and related disciplines.
For the sake of brevity, let me focus on some comparatives between economic analysis and one other field of social policies formation: environmental policy. The same arguments, however, hold in the case of other social policy disciplines.
Prior to the crisis, environmental sciences largely existed in the world of mathematical modeling, with core forecasts of emissions paths and their effects on the environment relying on virtually zero behavioural inputs. These technocratic models influenced both public opinion and policies. The proverbial representative agent responsible for production of emissions, was not a human being requiring age, gender, family, income and otherwise differentiated supplies of energy, goods and services. In a way, therefore, environmental policy was further removed from the realities of human and social behaviour than, say, finance, monetary or macro economics. Where economists are acutely aware of the above differences as drivers for demand, supply and valuations of various goods and services, environmental policy analysts are focused on purely aggregate targets at the expense of realism and social and economic awareness.
The same remains true today. Over recent years, the thrust of environmental policies has drifted away from local considerations of the impact of pollution on quality of life and economic environment considerations. As the result, environmental policies and programmes, such as for example wind energy development or localised incineration of waste, are becoming more orthogonal, if not outright antagonistic to the interests of consumers. Rhetoric surrounding these environmental policies considerations is also becoming more detached from the demos. For example, Ireland's attempt to make a play at European wind energy generation markets, replete with massive wind farms and miles of pillions, is now pitting our imagined (or mathematically-derived) exports potential, fuelled by nothing more than massive subsidies and consumer rip-off pricing for electricity, against all those interested in preserving the countryside's natural amenities, cultural heritage and other economically and socially meaningful resources.
Whereby behaviourally-rich analysis is now moving into the mainstream in finance and is starting to show up within the macroeconomic models, it is still wanting in the environmental policies research. The result is distortion of public responses and reshaping of political landscape around the environmental movements.
In most basic terms, there are three core problems with the current state of social sciences and policies formation mechanisms. None of these problems are new to the post-crisis world or unique to economics. In fact, in many case economics as a discipline of inquiry is years ahead of other social sciences in dealing with these shortfalls. In summary the core problems are: insufficient modeling tools, poor data, and politically captive analytics and decision-making.
The first problem is the lack of rigorous modelling tools capable of handling behavioural anomalies. Put differently, we know that people often make non-rational choices and we occasionally know how to represent these choices using mathematical models. But we are far from being able to incorporate these individual choice models into macro-level models of aggregate behaviour. For example, we know that individually people often frame their choices in the broader context of their own and collective past experiences, even when such framing can lead to undesirable or suboptimal outcomes. Yet we have few means of reflecting this reality in economic models, although we are getting better in capturing it empirically. We can model habitual and referenced behavior of individual agents and we can even extend these models to macroeconomic setting, but we have trouble incorporating this behavior into explicit policy analysis. We also face mathematical constraints on our ability to deal with the more advanced and more accurate models extensions.
The problem of insufficient tools is often compounded by the problem of over-reliance on technocratic analysis that marks our policy formation. Put simply, we live in the world dominated by policy-making targeting aggregate performance metrics (such as global emissions levels or nation-wide GDP growth rates). This implies that we often aim to create policies that are expected to deliver specific and homogeneous outcomes across a number of vastly heterogeneous geographies – physical, cultural, political, social and economic systems, nations and societies. The only feasible approach to such policymaking is via technocratic reliance on ‘hard’ targets, often with little immediate connection to everyday life, and prescriptive policy designs. The core pitfall of this approach is that when a harmonised policy fails, it fails across all heterogeneous locations and environments. There is nothing more erroneous from risk management perspective than attempting to introduce a harmonised response to such systemic failures. Yet this is exactly what the policymakers strived to achieve in the setting of the euro area crises. The more reliance we place on technical models-driven solutions being right all of the time in all of the locations, the more harmonised and coordinated our responses to shocks are, the higher is the probability that a policy failure will be systemic, rather than localised.
The only alternative to this fallacy of reliance on technical analysis and hard targets-based modeling is to permit local innovation and differentiation. This historically-validated approach of the past, however, is not en vogue in the world where global institutions and aspirations dominate local objectives and systems, and where pseudo-scientific fetishism for technical knowledge dominates social sciences and policy making.
Beyond technocratic fallacy of over-reliance on mathematical models and the shortage of some key tools looms an even larger problem.
Consider the most recent example of a systemic failure by the economics profession to predict the current financial crisis. Instead of tools shortfalls, this failure rests with the problem of analysis and policy capture by political and economic interest groups that firstly determine the agenda for policy analysis and research, then define parameters and scope of such research and, finally, set bounds for measuring, monitoring and actioning data on policy outcomes.
With the onset of the financial crisis, economists working outside regulatory offices, ministries and central banks have gotten a much greater access to data than ever before. Still, even with data in public domain, analytical resources come at a cost premium, as anyone attempting to compete with, say the Department of Finance, finds out very quickly. By the time it takes an independent analyst to compile and analyse data, the Department of Finance can deploy dozens of staff to flood the media and public domain with own reports and papers. The asymmetry of resources drives the asymmetry of power in analysis and this fuels the asymmetry in policymakers’ perception of data insights. For example, lone voices of dissent or single pieces of contrarian analysis are pushed aside by the sheer magnitude of consensus, often representing little more than one agency replicating the insights of the other agency.
We might be able to produce better insights into the workings and risks of the banking sector today than before the crisis, but this does not mean that the actions of regulators and Governments are going to be any better informed or better tailored.
Even when independent analysis and scrutiny are available, regulatory and policy responses largely ignore empirical insights. In a recent study, myself and a co-author looked at asset prices across the number of advanced economies prior to and after the crisis. Using a very simple econometric model, we showed that data prior to 2006 was providing clear and loud signals as to the emergence of a number of crisis-level risks. However, to derive this result we had to calibrate the model using a parameter that was set at ten times the levels assumed to be likely by the banking regulators. Thus, by regulations, by own governance and remuneration standards, our public servants simply were not required to do this analysis. As the result, regulators around the world sleepwalked the entire financial system into the latest crisis and found themselves utterly unprepared for the fallout.
This is not unique to our study conclusions. Back in 2005-2006, inside the Irish civil service there were several senior voices raising concerns over the direction of our economy. These were echoed by a number of research papers and analysts warnings coming from the ranks of independent and academic economists. They were ignored not because they lacked empirical basis, but because the policymakers were captive to consensus view aligned with their own political objectives.
Nobel prize winners, Robert Shiller (2013), and Edmund Phelps (2006) economists such as Nouriel Roubini, Roman Frydman and Michael D. Goldberg repeatedly warned about systemic problems in the US property and financial markets back in 2004-2007. The NYU Stern School of Business research centre did the same for the banking sector. Last, but not least, in academic economics, research into non-rational, non-representative agent models has been on-going since the start of the 1990s, largely unbeknown to the general public and politicians. In fact, since the mid-1990s, majority of the Nobel Memorial Prize awards in economics went to researchers who pushed aside the bounds of rational expectations and/or representative agent frameworks.
Still, the problem of policy capture by the often poorly informed adherents to specific schools of thought is hardly unique to economics. Let's take two examples of policies that have seized public imagination and policymakers' attention, while sporting only tenuous empirical foundations.
One is wind and wave energy. Although it appears that there is a near-consensus in academic and policy circles that these two sources of energy offer preferred alternatives to traditional fuels, in reality, such consensus can and should be questioned. The latent energy stored in water and wind is huge. However, wind energy harvesting is also subject to own externalities. One key one is the transfer of cost of pollution abatement from the commons relating to energy production and use, to the commons relating to land and natural amenities use. This externality was already mentioned above and its discovery credit goes to economics, not to environmental sciences. Another one is the transfer of the cost of energy-related pollution to consumers. In the real world, different consumers access energy through different channels. Some channels offer energy users a subsidy over the other. Some channels come with a choice that a consumer can make to substitute between different service providers based on environmental and economic costs considerations, other channels do not. Again, credit for pointing this out goes to economists; environmentalists are all too often simply opt to ignore these realities in pursuit of aggregate emissions targets over and above the consideration of their feasibility or their effectiveness in the face of social, cultural, political and economic realities.
For example, state-owned public transport is commonly priced differently from the privately-owned public transport and both are priced distinctly from private transport. Unless use of energy is explicitly and uniformly priced across all modes of transport and unless all modes of transport are perfectly substitutable, some consumers of public transport will receive subsidies at the expense of others and majority will be subsidized relative to private transport users. Thus, a suburban family is likely to pay a higher price for pollution per mile travelled than an urban one. The fact that in many cases a suburban family might have been forced (by planning, zoning, pricing and other systems operating in a heavily distorted markets) to make a choice of living outside the areas with dense cover by transport alternatives does not enter into the determination of pollution-linked taxes and prices. Any decent economist can be expected to understand this much. Yet the simplified worldview that public transport subsidies and private transport taxes are always good persists among our policymakers and within environmental lobby.
Another example of the policy that is empirically shoddy, yet politically heavily supported is electrification of transport. Recent research shows that in the US, even if electrical vehicles made up over 40 percent of passenger vehicles in the, there would be little or no reduction in the emission of key air pollutants. Now, consider the case of Ireland, where ESB has been running multi-billion euro investment programme aimed at developing EVs networks since the early days of the financial crisis. Just as the value of private sector investment shot through the roof, Irish semi-state sector, encouraged by policymakers and subsidized by high prices on consumers, launched into a major investment programme based on questionable benefits to the economy and society at large. The Government of the day even announced back in April 2010 (with the country rapidly hurtling toward an IMF-led bailout) a EUR5,000 grant to EVs buyers. That Ireland’s electricity supply comes from environmentally damaging sources does not phase the environmental policy advocates.
The debates about the current state of economics and social sciences in general are a welcome departure from the pre-crisis status quo, where such discussions primarily took place in the marbled halls of academia and beyond the scrutiny of public attention. However, it is worth remembering that the core problems faced by social policies analysts today are the ages-old ones problems of insufficient modeling tools, poor data, and politically captive analytics and decision-making. We might be able – with time and effort – to fix the first one. Fixing the other two will require a paradigm shift in the ways we collect and publish data, and in the ways our political and public service elites approach policy formation. Two thirds of economics and social sciences problems are political, not scientific.
