It is now widely recognized that GDP has many shortcomings as an indicator of social welfare. It is time for macroeconomists and policy makers to finally take action.
Despite the careful documentation of theoretical and empirical shortcomings of the gross domestic product (GDP) as a measure of social welfare and progress, its use in the course of time has not changed. Information about GDP growth has a significant influence on decision-making by individuals, businesses and governments. In fact, politicians and macroeconomists become noticeably nervous when there is little or no GDP growth. Unfortunately, the use of GDP must be regarded as a serious form of information failure.
The most important criticisms of GDP can be summarized as follows (Van den Bergh, 2009). GDP is actually an estimate of the costs rather than the benefits of all market-related economic activities. Moreover, it does not capture various social (including external) costs and basic needs such as community, serenity, clean air and direct access to nature. Absolute individual income is not a suitable proxy of individual welfare as it also depends on relative income and various income-independent factors. The aggregation of individual absolute incomes in GDP terms is therefore not a robust indicator of social welfare. Furthermore, individuals get used to changed circumstances, including rises in income, which cause well-being to return to its baseline level after a temporary change. GDP does not capture this ability to adapt. GDP (per capita) also ignores changes in the distribution of income. Since the marginal utility of income is decreasing in income, this means GDP will deviate from real welfare. In addition, GDP does not capture income comparisons or rivalry through the purchase of status goods. As status is scarce, rivalry for it is a zero-sum game. GDP growth neglects the effects of rivalry, which contribute to overestimating welfare growth.
Another shortcoming is that GDP only covers transactions that have a market price and neglects informal activities that take place outside markets, even though they may contribute to individual welfare. This is relevant to both developed (child care, voluntary work) and developing (substance agriculture) countries. GDP thus tends to overestimate the impact on welfare of fundamental changes involving transitions from an informal to a formal economy. An important subcategory of unpriced effects relates to external environmental factors and material resources taken from nature. As a result, current market prices insufficiently reflect social costs, causing the calculation of GDP to deliver at best an inaccurate proxy of social welfare, even when all relevant categories of activities, goods and services have been included (which is not the case). In addition, whereas pollution damage is not factored into GDP, cleaning up that pollution will increase it. Finally, GDP does not capture natural capital depreciation, including environmental (notably climate) change, depletion of resource supplies and biodiversity loss.
While criticism of GDP is hardly original, over time it has been refined, partly due to empirical research into human behaviour and happiness (Layard, 2005). Many former GDP or growth critics of the 1960s and 1970s made what were at the time intuitive statements about the role of status seeking and adaptation to changes in income, which in the meantime have been empirically confirmed. Nonetheless, it seems that criticism of GDP has had a difficult time establishing a foothold in the minds of economists, educators, policy makers, politicians and journalists.
Ideally, criticism of how GDP information is used should be directed at politics and society. However, non-economists cannot be expected to properly assess the accuracy, relevance and policy implications of GDP data as long as economists do not adopt a clear position on the GDP critique. Indeed, statements about GDP uttered by the vast majority of journalists and politicians, regardless of their political affiliation, are decidedly uncritical. Even lead editorials in quality newspapers frequently use the terms ‘prosperity’ and ‘GDP growth’ as if they were synonyms. Ironically, whereas GDP (per capita) is applauded in empirical macroeconomic circles, it plays no serious role – normative or otherwise – in macroeconomic theory.
Two common responses to GDP criticism
Despite severe criticism, the use of GDP has not changed at all. This paradox is the result of an ambivalent attitude among many economists. While they accept the criticism, they claim it is irrelevant. Such denials take on two basic forms. First, there is the belief that GDP information has only a modest impact on economic reality (which makes it difficult to object to disregarding GDP information in the public sphere). And second, there is the belief that despite all its shortcomings, GDP still provides useful information.
Regarding the first point, all signs indicate that GDP significantly influences the economy. For banks and financial markets, GDP is a core indicator. For companies, GDP growth indicates a favourable investment climate. GDP growth expectations even affect consumer confidence. Low GDP growth in particular worries politicians, who are constantly haunted by the spectre of re-election. GDP influence is bolstered whenever influential research institutes and advisory councils attach great value to GDP growth – notable international examples include the International Monetary Fund and the Organisation for Economic Co-operation and Development (OECD). GDP also has a pro-cyclical effect. In other words, as long as everybody believes that it is having a significant, tangible impact on everyday life, then like a self-fulfilling prophecy, it ultimately will.
Regarding the second point, the question is whether GDP information has advantages that outweigh its shortcomings? One possible benefit of GDP growth is that it creates confidence and economic stability. But the downside is that a decline in GDP generates negative expectations that reinforce the decline. Moreover, GDP per capita is widely seen as a useful measure of productivity. But the less frequently reported GDP per hour worked is a much better and internationally comparable indicator of productivity, especially since working hours greatly fluctuate between countries. Since an increase in labour productivity is ultimately not the goal, however, GDP per hour is not a useful proxy of social welfare either. Another common argument in favour of the use of GDP information is that the international standard for national accounts and GDP enables a comparison of GDP (per capita) to be made across countries. While necessary, this is not a sufficient condition for a meaningful indicator.
Policy relevance: climate and economic crises
Ignoring GDP information means removing systematic and cumulative errors resulting from individual and public responses to it. It creates a new perspective for judging alternative public policies. For example, suggestions have been proposed to institute regulatory taxes on overtime and the consumption of luxury goods, as well as restrictions on commercial advertising and cutbacks in temporary contracts (Robert Frank, 2004 and Richard Layard, 2005). While such policies may seem unattractive from the perspective of GDP growth, they are more appealing from the perspective of real social welfare.
Policies promoting social welfare at the expense of GDP growth would receive less resistance if there was less impetus for unconditional GDP growth. Take the economic assessment of policies designed to avert climate change. Most economic studies view the problem as a trade-off between a policy’s benefits and its costs as measured in terms of reduced GDP growth, as if the latter is a good measure of lost prosperity (happiness). This approach is essentially irrelevant though because this trade-off is likely to cover a time span of a hundred or even several hundreds of years. The GDP (per capita) in rich countries will have risen far beyond any welfare-maximizing level, in which case less GDP growth would not impact welfare.
Indeed, from a happiness perspective the real social cost of such policies is lower, while the costs of climate change, notably for poor countries, are more serious than when expressed in terms of the effects on GDP (Van den Bergh, 2010). In fact, this is not surprising since the bill for climate change or climate policy in terms of GDP is paid primarily by rich countries, where the correlation between GDP and happiness is small or non-existent. Stringent, safe climate policies aimed at precaution – i.e. avoiding serious climate change risks – would therefore be much more welcome from the perspective of happiness or real welfare. In effect, this means GDP growth would carry less weight in the evaluation of climate policies.
The economic-financial crisis
Ignoring GDP information would also influence our response to the current economic and financial crisis. Crisis management should focus on two main aims. First, minimize unemployment (’crisis malaise’), since it entails a huge loss of welfare for affected individuals and families. And second, restore economic confidence. Continued GDP growth is not a prerequisite for achieving either aim. In other words, employment and confidence do not necessarily require a minimal rate of growth. The aim of unconditional growth merely limits our ability to find solutions. Economic growth is no guarantee for reducing unemployment. For example, in an open economy growth can be easily sustained by outsourcing to low wage countries. Moreover, following a crisis growth is typically associated with the ‘creative destruction’ of specific sectors and corresponding jobs. As a result, training and experience are temporarily less well matched with available jobs. While lower unemployment is likely to increase GDP, a higher GDP does not necessarily lead to more jobs.
What is likely to lead to more jobs is cleaner production and less environmental degradation. According to Roefie Hueting, ’A given amount of production and consumption requires more labour with environmental conservation than without.’ 1 In this respect, the substitution of labour over time by scarce environmental resources and by manufactured capital causes pollution during the use and production phases. Cleaner production or production that exerts less pressure on natural resources is therefore generally associated with lower labour productivity, which requires more working hours.
Instead, economists should invest more intellectual energy in studying full or nearly full employment – and they devise policies for achieving this – in an economy that isn’t constrained by the onus of ‘constant growth’. Constant media reports about disappointing GDP growth merely serve to reinforce a negative spiral of confidence. The pursuit of growth consequently becomes an obstacle to finding a way out of the crisis. Needless to say, this obstacle needs to be dealt with. This also means defining ’crisis’ in terms of unemployment and not in terms of low or negative rates of economic growth. Unfortunately, this is contrary to the beliefs of modern economists and is unlikely to be welcomed by the current generation of economists. It would require that some influential economists (e.g. Stiglitz, 2009) set aside old dogmas.
These two examples illustrate that debating the GDP indicator is not a strictly ’ivory tower’ affair. Putting all our eggs into the GDP basket unnecessarily blinds us to what might effectively solve the most worrying problems of our time. More generally, GDP growth fetishism, i.e. the relentless pursuit of growth, constrains us in our attempts to find ways of improving human welfare.
It should be stressed that ignoring GDP information does not imply being against growth. The message is more subtle. Ignoring GDP in public policy implies that governments have to be neutral or indifferent to economic growth. If a government had no GDP information and could not gauge whether its economy was growing – nor by how much – its only possible response would be to show no concern about growth at all. Instead, the focus should be on real changes in welfare. If such changes go hand in hand with growth, some people might be happy, but those in the know would not give it a second thought. Growth should not be the ultimate aim of our society, nor an intermediate one.
Obviously, one has to accept that many individuals pursue the growth of their personal incomes. Despite the fact that people do not always strive for things that make them happy, acknowledging this does not mean that society should follow similar motives. For example, increasing individual incomes may generate status and therefore individual happiness, but at a societal level status seeking is a zero-sum game since it is an extremely scarce commodity.
Indifference towards growth does not imply a belief in indefinite sustainable growth. Being either rigidly for or against growth is not necessarily the best thing for social welfare. In other words, there are equally valid reasons for being against both growth fetishism and degrowth fetishism. The notion of degrowth has become fashionable recently. Like the pro-growth ideology, it attaches too much importance to the GDP indicator. Degrowth proponents also seem to misunderstand the basic causes of the issue. If we implement good environmental policies – and most importantly, stringent climate policies – this may well lead to negative growth, i.e. a decline of GDP, especially since growth sectors are relatively high polluters. But this should not be confused with the notion that degrowth will solve our environmental problems, or that it is the best (most effective and/or efficient) way to solve them. Degrowth is simply too blunt an instrument or strategy. We should not worry about the growth resulting from good policies; rather we should take it in stride. We need to be reminded of this regularly, since the relentless pursuit of economic growth (‘growth fetishism’) has unfortunately become the instinctive response of most economists, journalists and politicians.
Why does support for GDP continue?
GDP was never intended as a measure of welfare, but it has come to be used that way almost by default. Most economists, journalists, investors, public servants and politicians seem completely unmoved by GDP’s shortcomings. They consider irrelevant arguments that point out its failings. But the truth is that supporting the GDP indicator is a bad habit, nourished by the uncritical treatment of GDP information in economics courses and in the media.
Most economists believe that we should not abandon GDP until a suitable alternative has been identified. Some proposed alternatives represent a considerable improvement as an indicator of social welfare (Van den Bergh, 2009), but none of them is perfect. Median income would already be a great improvement over average income (GDP per capita) as it gives more weight to income distribution (inequality). Other than that, it has all the shortcomings of GDP per capita. The Human Development Index (HDI) is often mentioned as an improvement, but it is unable to differentiate between rich countries (this may just reflect that these countries have already attained a social welfare threshold). The best alternative may well be the Index of Sustainable Economic Welfare (ISEW) developed by Herman Daly and John Cobbin 1989. It suggests that the social welfare of OECD countries has stabilized since the 1980s (the USA even earlier) despite continued GDP growth.
But regardless of the availability of a credible alternative, if the use of any indicator represents a serious information failure, as is the case with GDP, this practice should be altered as soon as possible. Removing information failure is entirely consistent with mainstream economic theory. It is this argument, if any, that may persuade economists of the need for an improved indicator.
Macroeconomists in particular, and development economists to a lesser extent, seem reluctant to abandon GDP until a suitable alternative has been found. They cling almost instinctively and dogmatically to GDP. It seems that GDP information is so central to their training and work that it is emotionally difficult for them to criticize or detach themselves from it. To do so would be tantamount to questioning the relevance of their own studies and publications, and perhaps even the relevance of the broader field of economic research.
There are many other economists and non-economists, however, who are much more willing to take GDP information with a grain of salt in light of its many shortcomings. Macroeconomists need to adopt a more critical and honest attitude. They need to seriously consider whether the current body of macroeconomic knowledge really justifies assigning any kind of normative role to GDP. If not, there is only one course of action. Drop the habit. Ignore GDP as a measure of social welfare.
Daly, H.E. and Cobb, J. (1989) For the Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future. Beacon Press.
Frank, R.H. (2004) Positional externalities cause large and preventable welfare losses. American Economic Review 95(2): 137–141.
Hueting, R. (1996) Three persistent myths in the environmental debate. Ecological Economics 18(2): 81-88.
Layard, R. (2005) Happiness: Lessons from a new science. Penguin.
Stiglitz, J.E. (2009) GDP Fetishism. The Economists' Voice 6(8), article 5. http://www.bepress.com/ev/vol6/iss8/art5
Van den Bergh, J.C.J.M. (2009) The GDP Paradox. Journal of Economic Psychology 30(2): 117–135.
Van den Bergh, J.C.J.M. (2010) Safe climate policy is affordable – 12 reasons. Climatic Change, forthcoming. DOI: 10.1007/s10584-009-9719-7
Photo credit main picture: Hollandse Hoogte / David H. Wells
Hueting, R. (1996) Three persistent myths in the environmental debate. Ecological Economics 18(2): 86.