Inventing Economies

Davis A. Smith-Brecheisen

Since its invention in the 1940s, Gross Domestic Product (GDP) has been both lightning rod and bellwether. Two recent books, GDP: A Brief But Affectionate History by Diane Coyle and The Leading Indicators: A Short History of the Numbers that Rule Our World by Zachary Karabell, trace the history of this ubiquitous economic indicator from its wartime inception to its waning efficacy in the twenty-first century. Each of these works traces different ways the institutionalization and evolution of economic indicators has distorted understandings of economic life and social relations. Invented to explain output and preparedness for war, GDP became the primary statistical representation of Western affluence during the 1940s. It served as the conceptual apparatus that underpinned the United Nations’ goals to keep the peace and end poverty, effectively representing the ideological battleground that defined the Cold War. To put a finer point on it, GDP has become a sort of statistical embodiment of the ideological conflicts that have defined the postwar period, both political and economic. As the conceptual apparatus that has defined the second half of the twentieth century, the importance of GDP can scarcely be overestimated. The past few years, however, have not been kind to that holiest of economic indicators. In the wake of the economic collapse of 2008 and the subsequent jobless “recovery,” GDP has proven an increasingly troubled representation of economic health, unable to keep pace with a global economy defined in terms of innovation and creativity rather than productivity and output. The histories of GDP traced in the accounts offered by both Coyle and Karabell are perhaps best understood in two parts. The first part is the inception and rise to prominence of GDP first as a means of defining a crisis-prone and faltering brand of American capitalism before eventually becoming its greatest champion in the postwar period. The second part of the history traces the limitations of GDP following the financial and economic innovations that emerged from the ash heap of the global economy in the 1970s.

Importantly, both of the works reviewed here begin with the assumption that GDP is one of the twentieth century’s most important “inventions.” As Diane Coyle points out, it is “a made-up entity” that does not so much “measure market production” as it “defines” it (139). Created by the state to assess its ability to fight wars and “run its fiscal policy,” GDP has come to shape contours of what is considered “the economy” (15). She contends, “there is no clear definition of ‘the economy’” beyond the statistical representations of it (139). Zachary Karabell goes slightly further, arguing that GDP effectively invented “the economy.” One of his more striking points is the extent to which the invention and institutionalization of GDP in the 1940s “transformed the way people everywhere understood the material ebbs and flows of society” by treating “the material affairs of a nation as a coherent and cohesive subject that could be defined, measured, and tracked over time” (88, 78). And as productivity and output became the terms framing the economy, other questions such as those of welfare and value receded from view. The result of a forty-year history of discussing the economy in terms of output and productivity has led many to believe that this economic abstraction is, as Coyle puts it, a “natural object” waiting to be measured (139). When, in fact, GDP is a concept — one that has shaped the contours of political and economic life since it was invented after World War II.

At the core of both of these works is an abiding skepticism toward the ways in which economic complexity has been rationalized into overly simplified economic indicators by policymakers, economists, and statisticians. And while both books do a perfectly fine job highlighting some of the ways in which the global obsession with economic indicators has created a distorted view of economies, neither work follows through with the entailments of these claims, leading to some glaring missed opportunities and outright omissions. Among these is the extent to which standardizing GDP and turning economic output into the benchmark for progress buttressed the rise of U.S. economic hegemony in the postwar era. And although both books rightly point to the fact that defining the economy in terms of output and productivity has massively overstated the value of commercial banking and debt financing, the problem they are most concerned with is the question of measurement, not the complexity of value itself. And this is the real missed opportunity in both of these works: they accept the logic of statistical representation of economic forces, even as they critique them. Consequently, both Coyle and Karabell stumble over their object of analysis, reinforcing the importance of data and statistical analyses to formulate our understanding of political economy.

In her history of GDP, Coyle marks an important intellectual turning point at this time in debates about economic measurement beginning with Adam Smith, whose definition of “National Income” was the first to distinguish between productive and unproductive labor. Although since at least seventeenth-century Britain, states had developed some form of national accounting structure, the practice focused more on taking stock of assets and is not quite the same as defining output or productivity. By the eighteenth century, defining the economy in terms of productivity and growth had become increasingly important to the maintenance of empire. For Smith, Coyle notes, productive labor included “only those involved in the making of physical commodities, agriculture and industry. […] The provision of more services was a cost to his employer, and did not create anything” (10). Neoclassical economists of the 1890s revised the definition of productive labor to include services and thus greatly expanded the definition of national income. Though the definition of “productive” has changed, Smith’s division between productive and unproductive labor remains central to economic debates over the health of economies globally (Coyle refers to this as the “productive boundary” [38]). Some services might now count as productive according to economists and the Bureau of Economic Affairs (BEA), which calculates GDP, but the question of what constitutes the productivity of “the economy” is obviously far from a settled matter1

Both Coyle’s and Karabell’s prehistories of GDP point to the ways in which states have long grappled with the best ways to define the economic life of a nation. The answer, ultimately, has been to do so rationally and statistically. As Karabell asserts, to do this, the state needed to invent and institutionalize the data. As he also points out, however, developing a statistical representation of the economy invented the very thing economists and statisticians claimed to measure, and this fundamentally reframed the ways in which political economy was understood. Dropping the “political” altogether, what had previously been understood as a branch of “history and philosophy” moved into the realms of “math and science” (53). In the late-nineteenth and early-twentieth centuries, as the “scientific management of society” began in earnest, the definition of national economics became increasingly quantified. Karabell notes the extent to which, despite the limitations of the analysis as it was being undertaken, both financial markets and “politicians wanted a number, and they wanted it to be ‘the truth’” (24). Among other things, this obsession with the “truth” of data marks a qualitative shift in how national economies come to be understood and defined. By the 1930s these numbers were treated as though they presented “a perfect map of the world” (23). Whatever their intended ends, economic data actually came to define the contours of “the economy.” This is to reiterate the earlier point, that economic indicators do not so much measure the economy as produce it. And the vision they produce is a hyper-rationalized one. Although Karabell points to the ways in which this distorts the lived experience of daily economic life, and Coyle argues that this in many ways misrepresents the way economic exchange actually works, we might put the point more strongly: representing “the economy” through hyper-rationalized and statistical methodologies means that complex and often contradictory economic forces were simplified into ostensibly easily digested statistics.

This process of rationalization, by distilling the economy into measures of productivity and “prosperity,” elides the antagonisms of capital and ignores social welfare altogether. And though neither Karabell nor Coyle want to argue this first point, both are deeply invested in the latter. Karabell traces the degree to which the economy is defined by output, not welfare, to the combination of the postwar economic boom and the heating up of the Cold War. He notes that almost immediately upon its invention in 1946, GDP became a way of talking about how “we” were succeeding as a nation. And how “we” were doing was represented purely in terms of output, producing a vision of economic life whereby welfare was yoked to output and consumption. As GDP became, as Karabell notes, one of the “abiding elements” of national life, it became a way of reassuring people that its economy was protecting the nation from the “unruly patterns of boom and bust” (76). However misguided or ultimately fallible GDP has proven as such a measure, it nonetheless meant for the first time people “thought of their nation [and] their society or their own lives in terms of the collective material production of their country” (77). When “the economy” emerged as a feature of national consciousness, output effectively came to dictate the ground of daily life by linking the material well-being of American citizens to the output of the nation rather than, say, inequality. How much could be produced became the benchmark against which progress was measured, even at the level of “consumer sentiment,” which attempted to measure the degree to which individual consumption choices and feelings could predict output.

The fact that GDP became a way of measuring affluence is unfortunate considering that the invention of key economic statistics was “part of an overall movement toward social and political reform” (Karabell 29). Economic statistics were born, Karabell argues, out of “impulse to control” the “unruly beast that was economic life” (70). In the late nineteenth century, labor unions were part and parcel of the social movement to quantify the “unnecessary” hardship of exploitation and unemployment. Karabell notes the extent to which in the early twentieth century, the progressive movement of the 1930s appealed to this same logic: “the drive for more data and new statistics was born…from a strong suspicion that inequality was widening, social justice was weakening, and the pace of industrialization was creating as much harm as good” (24). The statistical insights yielded by economic data were intended to pinpoint crises of capitalism, not to justify exploitation.

Yet this is precisely what they have come to do. Although Karabell wants to argue that these economic indicators have come to serve ends that were never intended, this is not quite accurate. The labor movement has been predominantly interested in making capitalism work better — i.e., producing more affluence. And appealing to the available statistics to resolve the conflict between labor and capital meant that the progressive movement and labor unions adopted the very mechanisms that created inequality in the first place. As they did, the solutions to the contradictions of capital — unemployment and inequality — were framed in terms of productivity figures. Unemployment could be solved if we made more stuff and put more people to work. Likewise, wages could be improved if more money was coming in. The data proved as much. But this only intensified capitalism by further quantifying and thus justifying the drive to produce. Crucially, appeals to statistics could not solve the problem because the indicators themselves defined the terms of the conflict between labor and capital. And labor reform movements, by accepting these terms of exploitation, ultimately made capital more efficient. There were real gains for workers during this period, and that should not be overlooked. The point here, however, is to highlight the degree to which the appeals to statistical modes of representing the economy further entrench the logic of the exploitive structures they were intended to critique. From this standpoint, the transformation of output statistics from a measure of inequality to a measure of affluence seems almost tragically inevitable.

It was not inevitable of course, but it did happen. And both Coyle and Karabell attempt to explain why. Of the two, Coyle is the most interested in defending the reasons why, in the aftermath of World War II, GDP really was a good indicator for the overall economic welfare. Though she suggests that one of the “distasteful” consequences of disaster is “a boom in GDP” (43). Basically, the logic goes: if you destroy everything, you have to pay people to rebuild it. Following World War II, output, productivity, and welfare were closely linked because the war decimated most of the advanced economies in the world. Rebuilding meant vast improvements both in output and welfare. Coyle goes on to conclude that the GDP became the economic indicator par excellence because during the period that consumer-driven modern economies were formulated, it was just the best indicator of growth and welfare.

Karabell approaches the proliferation of economic indicators such as GDP slightly differently, suggesting that GDP was standardized and institutionalized largely out of international economic anxiety — policymakers and citizens wanted some assurance that the economy was not going to collapse immediately. He also argues that GDP’s importance was in part ideological. Though it was the intent of the earliest architects of national statistics in the 1930s to “replace endless ideological jousting with solid data” during the Great Depression, in effect, those same statistics became the instruments that would define the ideological jousting of the postwar era (Karabell 75). With World War II ending and domestic production ramping up, fears of inequality were supplanted by a newfound “need” to out-produce the Soviet Union. GDP became a distilled, simple figure to entrench the belief that United States had the better mousetrap. According to Karabell, in order to make this argument hold, the United States had to convince the world that GDP was in fact the best measure of an economy (Coyle touches on this point as well). Essentially, the United States had to establish the rules of the game. And because it was in the position to do so, the United States began doing what it did best in those postwar years: standardize and export.

The United States, with the support of Britain, began standardizing national income accounts. They did this not just through international economic agreements such as Bretton Woods, but through organizations such as the United Nations as well, which under the aegis of promoting peace also began promoting liberal capitalism. So that it might achieve its astonishingly utopian goal of “banishing…destructive economic forces,” the United Nations began standardizing national income accounting and proliferating global economic data (Karabell 89). As Karabell notes, both the World Bank and the International Monetary Fund also played a significant role in collecting and disseminating these statistics. It would be a vast understatement to say that there are far-reaching implications to the fact that supranational organizations, backed by the United States, effectively dictated what a sustainable economy ought to look like.

It is, however, interesting to note the degree to which Karabell’s work implies (even as it is not explicitly mentioned) that the standardization of economic indicators such as GDP has played a crucial role in securing U.S. economic hegemony. Essentially, exporting standard methods of national accounting allowed the United States via the UN, World Bank, and the International Monetary Fund to create a world in its own image by systematizing and defining what economies ought to look like. Rationalizing economies into bundled data sets made it more efficient for those organizations, with the blessing of the United States, to adjudicate the creditworthiness of less-developed economies. The “structural adjustments” that have historically been attached to the loans for those countries deemed creditworthy have created massive new markets for both goods and financial capital. And increased access to financial capital markets has been absolutely central to securing U.S. economic hegemony, which hinged on securing open access to markets for goods and capital. Of course, economic indicators, which U.S. and British economists invented, have often been the justification for these policies. And just as open capital markets have buttressed the U.S. economy, they have toppled others. The U.S. role in the Latin American debt crisis would be just one such example. More generally speaking, one might point to the vast amount of capital currently flowing back into the United States from less-developed economies and the structural adjustments imposed upon debtor nations by their lenders.2

To be sure, global capital markets cast a volatile pall over the entire history traced by both authors. If the first part of this narrative has been a history of the ways in which the rationalization and institutionalization of GDP has obscured questions of economic welfare while simultaneously securing U.S. capitalist hegemony, the second, related part is best understood in terms of the limitations and outright failures of these indicators to grapple with the increasingly prominent role played by debt financing and capital markets since the 1970s. The nagging problem of value elided by data-driven representations of capital in the early twentieth century, in other words, has never really gone away. While the economic turmoil of the 1970s needs no rehearsal here, it is worth noting that because the makeup of Western economies shifted dramatically, finance became increasingly important in GDP calculations. Understanding the role of finance in GDP is far from a straightforward task, perhaps because it requires abstracting and quantifying “fictitious capital.” Yet, for thirty years, many praised the “contribution of financial services to the economy,” essentially arguing that it alone rescued the American economy in the 1970s and has been fueling growth ever since (Coyle 98). Without a doubt finance has played a key role in fueling capitalist expansion and securing U.S. economic hegemony over the past thirty years. “Nonetheless, given the devastation [financial services] have caused,” Coyle argues that “we have to ask how that devastation can be reconciled with the importance of finance in the GDP figures.” More importantly still, the crisis encourages a “reevaluation of…the economic theory on which policy has been based for the past generation.” Questioning just how much value the financial industry has added to the economy, Coyle wonders, “have we been aiming at the wrong target all these years” (98)? For readers here, the answer to this question will certainly be, “of course.” But the question Coyle is asking is not really concerned with whether or not capitalism is the most economically just system (for her the answer to this question is clearly, “yes”).

Her question is really more complicated than that. And it perhaps suggests a far more interesting problem than Coyle perhaps realizes. What she is questioning when she questions the “target,” at which policymakers have been aiming, is whether or not output statistics — or statistics more generally — are capable of measuring productive labor at all. To put a finer point on it, Coyle is really asking two related questions: What constitutes productivity? And, which activities generate value? And she begins her answer by assessing the value of the financial services industry in terms of “how it is counted in GDP” (99). To summarize: it is complicated, and not just because it is impossibly complex or opaque. Rather, because, according to the OECD, if financial services were calculated like most other services, financial output would be “very small, if not negative” (100). In order to turn finance into a sizeable percentage of GDP, according to Coyle, the UN introduced a measure which entails “that increased risk-taking is recorded as increased real growth in financial services.” In other words, risk is turned into a productive activity. Taking risks generates output even though banks could do so by leveraging equity rather than generating any “real” output. That is, debt essentially becomes and generates productivity. Coyle notes that the profits were imaginary, creating a “statistical mirage [that] affects all countries’ GDP.” Citing a study from 2011, she points out that the current method of calculating GDP “overestimates the service output of the commercial banking industry by 21 percent” (101). And this has wide-reaching political and economic consequences. Not least of which, it vastly overstates the importance of commercial banking to the economic health of a particular country, turning an industry that may well actually have a negative effect on economic growth into the engine of the economy by virtue of statistical abstraction.

The inflation of the “output” of finance capital tells only one part of the current drive to quantify the increasingly abstract quality of value in service-based economies. In 2013, the BEA announced that it “shifted the way it measured national output. The result was a $400 billion adjustment” (Karabell 3). “Shift” is putting it mildly. The BEA dramatically revised GDP to include “creative work undertaken on a systematic basis to increase the stock of knowledge, and use of this stock of knowledge for the purpose of discovering or developing new products” (4). Research and development at Apple or Pfizer and “work” undertaken by Richard Florida’s “creative class” is now included in GDP. Productivity, in other words, seems increasingly to be a feature of what would more commonly be understood as costs — research and development or debts and risk. Coyle defines what counts as productive and what does not as the “productive boundary” (38). And this, Coyle argues, is “arbitrary” because in reality the division is “simply a matter of convenience” — some things are just easier to quantify (105). It is a surprisingly cavalier way of thinking about the problem of what constitutes productivity — even if it is intended to be merely a description of the practice of defining GDP and not a value judgment. It stands out particularly because in the previous section on the “great crash” of 2008, Coyle explains the arbitrariness of what gets included in GDP and how this effectively dictates the political contours of economic policy.

Despite or perhaps because of this increasingly capacious definition of productivity, GDP is proving somewhat suspect. Both Karabell and Coyle suggest as much, even if neither author is willing to give up on it completely. Current revisions to GDP notwithstanding, recent actions by the BEA suggest that confidence in GDP is waning. On April 25, 2014, the BEA began releasing its “gross output” number quarterly rather than annually (and often a year late). The hope is to provide a more accurate “statistical tool” to understand what society actually makes. According to the Wall Street Journal this differs from GDP because rather than measuring the use economy, or the amount of final goods sold, gross output measures the “make” economy — gross output is intended to measure actual production: “[v]alued at more than $30 trillion at the end of 2013,” gross output is “almost twice the size of gross domestic product, and far more volatile.”3 The point here is that depending on the lens through which one tries to order the economy, one gets a very different understanding, not of how it works, but of what comprises it in the first place. Some would hold that this is just what happens when you slice the same thing in different directions. But such claims assume that there is something to slice — a quantifiable entity that exists out in the world.

This brings us back to the point that opened the essay: these indicators do not measure “the economy” but shape it by determining which economic activities are included in productivity figures and thus adjudicating what activities add value. We have already seen just how arbitrary this is. For example, creative labor counts but domestic labor does not, no matter how creatively undertaken. By noting the various ways economic indicators are still grappling with the problems of productivity and value in Adam Smith’s work, each of these works points to the myriad ways economic indicators struggle to represent the production of value. In the process, each book inadvertently highlights the limitations and difficulties that inhere when attempting to quantify something as abstract as value. Unfortunately, both works take value for granted. They assume that goods and services enter the market with an inherent value that is simply measured and aggregated at the moment of exchange. The problem with economic indicators, in both Coyle’s and Karabell’s arguments, is the extent to which value is abstracted and aggregated. In reality, the abstraction begins long before the bureaucratic attempts to measure productivity. As Marx argued in Capital, value itself is abstract, defined only after exchange takes place. Value is not measured through exchange, but invented. The difference between Marx’s view of value and the one assumed by Coyle and Karabell is crucially irreducible. If value is always already a representation that obscures the social relations embodied in exchange, revisions to the way data is aggregated and analyzed can scarcely address the problems of value and productivity as Smith and Marx each conceived them.

Rather than tarry with questions of value and the limitations of statistical modes of representation, Karabell and Coyle focus their criticisms of economic indicators on the extent to which these indicators largely ignore questions of social welfare. Coyle argues that although GDP has become a kind of abstract ideological tool, it was never intended as such. Rather, Simon Kuznets, who was instrumental in formulating the data for early attempts at national income accounting, “saw his task as working out how to measure national economic welfare rather than just output” (Coyle 13). Output, argued Kuznets, often measures “costs implicit in our economic civilization” and counts them as a net benefit (Coyle 14). Karabell cites Robert Kennedy who echoes this sentiment, arguing that output statistics measure “everything…except that which makes life worthwhile” (49). The efforts here to rethink the contours of the economy are admittedly closer to the original intent of economic indicators than the questions of output that have buttressed the economy since the middle of the twentieth century.

Yet, rather than really grapple with the ways in which qualitative questions about welfare and inequality might escape the grasp of big data, both of these books retreat further into the world of statistics, much in the same way each work treats value. And this is the real limitation of each of these histories. Coyle’s work points to limitations of GDP to define market production, arguing it is losing its relevance in a global economy that is now, more than ever, primarily “an intangible entity” (131). And though she notes the fact that the “financial crisis has given extra urgency to the need to rethink the concept of economic value,” she suggests the path to “radically” rethinking “the economy” is to entrench our commitment to the statistical modes of thinking. Which is to say, her solution to the current “statistical fog” is to do the thing that produced the problem in the first place, but better.

Karabell’s solution too retreats to the safety net of big data. His proposed solution is to form more refined and customizable statistical representations based on expanded amounts of data. He rightly notes, “the myth of national numbers that reflect national realities distorts how we attempt to solve [problems like unemployment] collectively and inhibits how we meet our challenges individually” (245). What we need, he claims, is a “dashboard” of “bespoke” indicators, “tailored” to the “particular questions” of governments and individuals. The rationale for a turn to bespoke and localized indicators is an attempt to move beyond the “blunt” instruments of macro-level indicators (249). And though his solution would not be quite as blunt, it is nonetheless a call for more widespread data, and in this regard it is essentially an attempt to more deeply (and specifically) quantify more aspects of daily life. It is essentially finding ways to aggregate the data the supporters of human capital have been arguing we should have had all along. Like Coyle, the future that Karabell wants is actually much closer to the history he has been critiquing than perhaps he would want to admit.

Both authors, in other words, are doubling down on big data. And it seems somewhat striking that those two histories, which begin with a healthy dose of skepticism toward statistical abstraction, should conclude by claiming we need more of it. My point here is not to criticize such histories, exactly. Both of these provide perfectly lucid overviews of their subject matter and raise plenty of interesting questions. Rather, I hope to point out that the modes of thinking championed in these books goes very little distance in solving the problem of value that preoccupied both Smith and Marx — a problem that remains central to the antagonisms at the heart of capitalist economies. Granted, economics as a discipline has become almost solely data-driven, but that does not mean it should be.4 It seems quantifying questions of output and value are essentially deepening the commitment to reifying and rationalizing social relations, which will in turn only further distort them. If data can, as Karabell argues, “alter our perception of reality,” it seems at least plausible that the commitment to statistical representation is producing a vision of the world that veils, or at best elides, the contradictions and antagonisms that prompted these histories in the first place. (5). Which is to say, perhaps more than anything else, what these economic histories reveal is the limit to bourgeois economics as a framework for its own critique.

Notes

  1. “What Is Industry Gross Output?” Bureau of Economic Affairs, Frequently Asked Questions, 10 March 2006
  2. Leo Panitch and Sam Gindin, The Making of Global Capitalism: The Political Economy of American Empire (London: Verso, 2013).
  3. Mark Skousen, “At Last, A Better Economic Measure,” The Wall Street Journal (22 April 2014).
  4. Deirdre N. McCloskey, The Rhetoric of Economics (Madison: U Wisconsin P, 1985).