Skip to main content

Kennedy School Review

Topic / International Relations and Security

Counting What Counts: GDP Redefined


What did the BP oil spill in 2010 mean for the U.S. economy? Progress. At least that’s the conclusion of the economy’s de facto benchmark—gross domestic product (GDP). As the massive oil slick seeped into the Gulf Shore, J.P. Morgan representatives noted that economic activity generated by cleanup efforts would likely outweigh losses to tourism or fishing (Di Leo 2010). And what of the hundreds of miles of property damage and ecosystem deterioration? Not counted. The bankers correctly concluded that our standard barometer of economic welfare would likely register the largest oil spill in history as a net gain.

Restoring GDP’s Purpose

Critics of GDP, who range ideologically from Robert F. Kennedy to Reagan adviser William Bennett, have been exposing such perversities since national income accounts were first federally instituted in 1934. In fact, the first such critic may have been the creator of GDP himself. Simon Kuznets, the young economist tasked by Congress with measuring the output of a Depression-era economy, warned after generating the GDP framework that “the welfare of a nation can scarcely be inferred from a measurement of national income” (U.S. Senate 1934, 7).

The succeeding eight decades have proven Kuznets right. GDP has achieved its narrow original purpose of measuring aggregate economic activity while remaining a woefully inadequate gauge of national welfare. Agnostic to inequality, GDP has portrayed per capita income as rising over the last decade despite falling median earnings (DeNavas-Walt et al. 2011, 8). By ignoring future growth potential, GDP has improved with the depreciation of machinery and the extraction of finite coal deposits while overlooking the extent to which we educate our youth or cultivate entrepreneurship. By omitting nonmarket values, GDP growth has tended to accelerate with crime rates, smog levels, and commuting time while slowing with vacation days and family-cooked dinners.

Such mismeasurement is profoundly problematic insofar as GDP serves as the headline metric for policy-making success. Examples are common. In arguing for passage of the controversial U.S.-Colombia free trade agreement last year, the Obama administration sidestepped the thorny issues of distributional impact and environmental ramifications by reducing the cost-benefit analysis to a single rubric: a projected $2.5 billion addition to GDP (White House 2011). Continuing to conflate GDP with national progress means subordinating interests of social cohesion and sustainability to the single-minded pursuit of ever-increasing raw output.

Progress Toward Measuring Progress

Growing concerns about GDP’s adequacy as a comprehensive benchmarking tool have spurred economists, think tanks, and multilateral institutions—aided by advances in data collection and statistical analysis—to generate a few dozen supplemental indicators. From the 1973 release of the Sustainable Measure of Economic Welfare to last year’s proposed Quality of Development Index, the list of alternatives continues to grow. In 2008, French President Nicolas Sarkozy commissioned a team of top economists, led by Nobel laureates Joseph Stiglitz and Amartya Sen, to study the inadequacies of GDP and assess the merits of the proliferating alternatives. The commission concluded, “the time is ripe for our measurement system to shift emphasis from measuring economic production to measuring people’s well-being” (Stiglitz et al. 2009, 12).

Numerous governments have already started making the shift. Following a European Union–sponsored conference in 2007 entitled “Beyond GDP,” the federal statistical agencies of Germany, France, and the United Kingdom each began constructing broader measures of sustainability and welfare. Last year, Bhutan unveiled a second generation of the Gross National Happiness measure that drives its distributional and regulatory policies. Within the United States, Maryland recently became the first state to shift to twenty-first century benchmarks by adopting the Genuine Progress Indicator.

Is there one particular indicator type that merits a position alongside GDP for U.S. federal policy making? And how could political and institutional obstacles be overcome to win federal adoption of such new measures? To answer both questions, we use three overarching objectives that guide our recommendations: accuracy, feasibility, and impact. We seek to propose a set of new indicators and a means of implementation that—considering the technical and political constraints—can offer a comprehensive and rigorous measurement of national progress so as to better inform policy makers and broaden GDP-driven public discourse.

Finding a New Framework

In assessing the merits of more than two dozen competing alternative indicators, we presuppose the importance of aggregating multiple variables into a singular benchmark, rather than offering a dashboard of many separate variables. Simplicity is a key prerequisite for any effective counterweight to GDP. If needed, aggregate indicators can always be disaggregated into component variables to reveal causal linkages for policy intervention.

Singular indicators tend to fall into one of three categories: subjective metrics, composite indexes, and adjusted-GDP measures. Subjective indicators, such as Gross National Happiness, employ the logic that the best judges of a people’s welfare are the people themselves. Relying on surveys that ask respondents to rate their quality of life, these measures intend to capture a nation’s average well-being. But the aim of such subjective measures differs from our own. While we seek to assess the policy-relevant components of welfare, self-reported welfare can vary with weather, sports outcomes, and other factors outside the realm of policy.

By contrast, composite indexes like the U.N.’s Human Development Index grade countries on policy-relevant dimensions of welfare based on objective performance (e.g., literacy rate for education, carbon emissions for environment) and then take the average grade as overall welfare. But this approach raises other problems, including the question of how to weigh distinct variables in creating a composite number. To avoid making controversial and arbitrary decisions about the relative importance of health versus income versus education, most indexes simply assign equal weights across all variables. However, the choice of equal weights, while more palatable, is no less arbitrary.

Instead of assigning weights, adjusted-GDP measures, such as the Genuine Progress Indicator, impute prices for a wide range of welfare-relevant values that are excluded from GDP, from the annual worth of household labor to the cost of a year’s coal extraction. Adding the benefits and subtracting the costs from GDP yields a more holistic, dollar-denominated expression of sustainable economic progress. Of course, imputing values for nonmarket goods and services raises significant accuracy challenges. Debates over appropriate pricing methodologies, many of which remain unresolved, have fed reams of scholarly journals.

While uncertainty complicates the aggregation of variables in both adjusted and composite indicators, adjusted-GDP measures still hold greater potential for contending with GDP because the unit of measurement remains in dollars. Consider that opponents of drilling for oil in the Arctic National Wildlife Refuge want to contest an assertion that doing so will boost GDP by $350 million. Which is the better counterargument: “the ensuing loss of natural capital will actually cost a net $147 million” or “the project will lower our welfare index from a 0.7 to a 0.63”? Given the narrative-swaying potency of the former, we recommend adjusted-GDP as the framework best equipped to parallel GDP.

GDP’s Modern Companions

What variables should be included within this framework? That depends on the particular variety of welfare we seek to measure. Do we wish to simply gauge current welfare or to incorporate changes in assets that affect future welfare (e.g., depreciated machinery, extracted coal, increased educational attainment)? Do we wish to measure purely economic well-being or to incorporate social and environmental components of policy-relevant welfare (e.g., public parks, leisure time). The answer depends on the particular policy questions at hand. If the city of Baltimore, for example, wanted to assess this year’s quality-of-life impact of a sprawl-reducing urban renewal project, the best indicator would include current social and environmental factors (e.g., commuting time, smog) but probably not changes in assets (e.g., debt or depreciation incurred for the project). By contrast, if the city wanted to assess the contribution of the project to sustainable economic growth, the appropriate gauge would subtract asset depletion (e.g., additional debt) and add asset creation (e.g., increased social capital) but not account for noneconomic variables like commuting time.

Given such distinct indicator needs, we recommend the creation of four new national indicators—G2, G3, G4, and G5—to assess four definitions of progress: current economic welfare (G2), sustainable economic welfare (G3), current general welfare (G4), and sustainable general welfare (G5). Figure 1 details the particular objectives and adjustments of each new indicator. This series of welfare benchmarks would stand alongside GDP, which would become G1, in the same way that the federal government uses U1 through U6 as complementary measures of unemployment.

Figure 1 — Four twenty-first century indicators.

Overcoming the Obstacles

The politics of such statistical reform are treacherous. Previous attempts to enshrine alternative indicators have had to contend with three formidable obstacles: methodological uncertainty, interest group opposition, and institutional inertia. Methodological concerns, as outlined above, partially explain why the Key National Indicator System (KNIS), a federally  funded initiative featuring a 200-indicator dashboard, chose not to report an aggregated headline metric. The second obstacle can occur when commercial or constituent groups fear negative impacts from a change in indicators. In a representative case, when the U.S. Bureau of Economic Analysis decided to compute environmental satellite accounts in 1992, the coal industry’s hostile response prompted two West Virginia representatives to kill the initiative in Congress. The third obstacle is the pervasive tendency of bureaucratic organizations to resist change. Despite bipartisan consensus on the need to update the antiquated federal poverty measure with the more comprehensive Supplemental Poverty Measure, it took the Census Bureau more than fifteen years to release the new indicator.

In spite of such obstacles, there are viable pathways toward adoption of new comprehensive indicators. To circumvent interest group opposition, the first step is to build a winning coalition in Congress. Social conservatives (seeking to formally account for work in the home), market-oriented moderates (seeking to better capture the benefits of entrepreneurship), economic progressives (seeking to adjust income for inequality), and environmentalists (seeking to internalize the cost of carbon emissions) could form the backbone of a surprising alliance for the development of new, twenty-first century indicators.

To overcome the genuine methodological challenges, the work of formulating new indicators could be delegated away from Congress to a bipartisan commission of experts, along the lines of the KNIS Commission, which was appointed by leaders of both parties in 2010. The commission could decide which variables enjoy sufficient methodological certainty to be initially included while advancing research on more ambiguous variables for their eventual incorporation.

Acting on such a commission’s recommendations, Congress could distribute the work of computing, aggregating, and reporting the new indicators to multiple qualified agencies and provide a clear mandate and necessary funding so as to minimize risk of bureaucratic inertia. The nation’s more than twenty designated statistical agencies could be tasked with collecting and computing data, while the Bureau of Economic Analysis, which currently produces the national income and product accounts, could spearhead the aggregation of the new indicators. The President’s Council of Economic Advisers could then play a lead role in reporting the indicators, particularly through its annual “Economic Report of the President.”

We Get What We Measure

Once instituted, this set of GDP supplements would have profound impacts on both the public narrative and the policy-making process. Imagine the symbolic power of headlines like “GDP Grows, Welfare Falls” and blog-disseminated graphs revealing a growing gap between GDP (G1) and sustainable economic welfare (G3). Such press could catalyze a shift to a broadened definition of progress. In time, policy makers themselves could use such tools to more adequately weigh critical policy decisions and to set more comprehensive policy goals. The Government Performance and Results Act (GPRA) Modernization Act of 2010 (GPRAMA) already requires that the Office of Management and Budget establish “long-term, outcome-oriented goals for . . . crosscutting policy areas” every four years (Dodaro 2011, 3). New G2 through G5 indicators could inform such holistic benchmarking.

The conventional wisdom that “we get what we measure” is as true for policy making now as it was in the era of Kuznets. Most people today would agree with his eighty-year-old warning that GDP does not capture all that policy making should seek to “get.” The time has come to move beyond one-dimensional growth and to pursue growth that is shared, sustained, and translated into a higher quality of life. The time has come for indicators of real progress.



DeNavas-Walt, Carmen, Bernadette D. Proctor, and Jessica C. Smith. 2011. Income, poverty, and health insurance coverage in the United States: 2010. Current population reports. Washington, DC: United States Census Bureau.

Di Leo, Luca. 2010. Oil spill may end up lifting GDP slightly. Wall Street Journal, 15 June.

Dodaro, Gene L. 2011. Managing for results: GPRA Modernization Act implementation provides important opportunities to address government challenges. Washington, DC: Government Accountability Office, 10 May.

Stiglitz, Joseph E., Amartya Sen, and Jean-Paul Fitoussi. 2009. Report by the Commission on the Measurement of Economic Performance and Social Progress.

White House. 2011. Fact sheets: U.S.-Colombia trade agreement and action plan. Washington, DC: Office of the Press Secretary, 6 April.

U.S. Senate. 1934. National income 1929-1932. 73rd Congress, 2nd Session, Document No. 124. Washington, DC: Government Printing Office. 


We would like to thank the forty individuals that we personally interviewed—representing institutions including the U.S. Department of Commerce, the World Resources Institute, the Heritage Foundation, the Institute for Policy Studies, and others—for lending insights that inform this article’s analysis


This article was originally published in the 2012 edition of the Kennedy School Review.

Ben Beachy is a 2012 Master in Public Policy candidate and Public Service Fellow at the John F. Kennedy School of Government at Harvard University. He spent the previous six years working as an economic policy analyst in Nicaragua and as a national organizer in Washington, DC, with the advocacy organization Witness for Peace.

Justin Zorn is a 2012 Master in Public Policy Candidate and Public Service Fellow at the John F. Kennedy School of Government at Harvard University. He recently completed a Fulbright Fellowship in Singapore and a master of science at Oxford University, both focused on long-term planning in governments and nonprofits.


Photo source here.