Historically, Gross Domestic Product (GDP) has been accepted as a proxy measurement of well-being in society, though its faults and flaws when used for this purpose are legion and increasingly apparent. GDP measures the volume of monetary exchange within the economy—the commotion of money and nothing more. That makes it an infinite planet statistic, unsuited to the challenge of creating a sustainable foundation for the human economy. Thus a movement to go "Beyond GDP" and measure sustainable delivered well-being has been gaining momentum around the world.
Amid the profusion of alternative indicators that attempt to replace, revise, or supplement GDP, there's a clear division in methodology: While all of them rely on economic metrics, some add objective data from physical, earth, and sometimes social sciences. The Genuine Progress Indicator (GPI) falls in this category. Other alternative indicators use survey research methods to incorporate subjective reports of well-being, as does the Bhutanese measure: Gross National Happiness (GNH).
Meanwhile, governments and nongovernmental organizations are under increasing pressure to justify their budgets and ensure maximum return on every dollar spent by showing how their programs and policies produce real and palpable benefits in the world. To accomplish this they must evaluate their success in achieving benchmarks—outcomes chosen as goals.A movement to go beyond GDP has been gaining momentum around the world.
A synthesis of these two movements and these two methodologies holds great promise. It would establish an indicator set capable of assessing sustainable well-being—a macro-scale policy goal—while also helping to clarify least-cost, maximum-benefit options at the program or microeconomic level. The combination of the two movements—to get "beyond GDP" and to institute outcomes-based budgeting—is likely to be more successful more quickly than either movement alone.
Thus, a strong contender for an indicator set to measure the efficient delivery of sustainable well-being could be created by blending elements of the Genuine Progress Indicator, the Gross National Happiness Index, and Outcomes-Based Budgeting (OBB).
The Genuine Progress Indicator
The Genuine Progress Indicator (GPI) grew out of the Index of Sustainable Economic Welfare developed by Herman Daly and John Cobb in 1989. Beginning with a basic measure of total personal consumption, GPI adjusts that figure for changes in income inequality, and then it adds the value of unpriced, nonmarket goods and services, and subtracts the value of unpriced costs. While local implementation varies, GPI generally consists of a dashboard of about two dozen sub-indicators that fall into three groups: economic, environmental, and social.
In 2009, Maryland became the first state to compile the GPI officially (as a "complement" to Gross State Product). The Maryland GPI uses 26 metrics and is reported on a website that is exemplary in its clarity. In 2012, Vermont became the second state to endorse the official compilation of a GPI, entrusting the work to the Gund Institute. The Maryland methodology is being followed to achieve a degree of interstate comparability.
One problem with GPI is that its usefulness as a guide to state experience (and policy) is compromised by the lack of state-specific data. Some GPI calculations rely on "stepping down" national-level data to the state level. For example, Vermont represents less than 1 percent of all U.S. households. In the absence of good data about the value of domestic production in Vermont, the state GPI is derived by taking a fraction of the national figure. To be a useful guide to policy design and implementation at the state level, GPI needs greater state-level accuracy.
Another problem: While GPI is a decent approximation of sustainable economic well-being, it doesn't measure total or overall well-being, because it doesn't assess the benefits we derive from healthy social capital. The GPI does measure "lost leisure time," and because maintaining social, community, and family connections takes time, it's reasonable to expect that lost leisure time would be inversely related to strength of social capital. But that hypothesis could be demonstrated directly, not through a proxy for social capital.
And another problem: Ultimately, the elements of GPI rest on assumptions about what will and won't, what does and doesn't contribute to human well-being. Those assumptions may be wrong.
To recount a personal anecdote: As a graduate student in southern California I was disturbed one night by the sound of a helicopter hovering low over the neighborhood, shining a strong searchlight into the streets and alleys nearby. "Oh, that," said a neighbor. "That's the police, chasing some car thief, most likely." If the goal is to reduce crime through increasing the rate at which perpetrators are apprehended, helicopters with searchlights may make sense. If the ultimate goal of controlling crime is to give citizens a sense of safety and security in their homes, the tactic may be counterproductive.Sustainable delivered well-being is the ultimate end for which other indicators are instrumental.
To generalize: Objective circumstances like crime rates or lost leisure time are relatively easy to measure, but they can lead policymakers to commit the fallacy of misplaced concreteness—mistaking the map for the terrain, the instrument reading for the thing it measures. Only by incorporating some element of subjectively reported well-being can an indicator guard against the commission of this fallacy. Sustainable delivered well-being is the ultimate end for which other indicators are instrumental ends; measuring sustainable delivered well-being directly through survey research allows for the assumptions behind the indicator set to be calibrated against reality.
Although it's a definite improvement over GDP as a measure of sustainable well-being, GPI thus falls short in its measure of the nonmarket costs and all the nonmarket benefits that humans experience in society, and has conceptual problems as a measure of total sustainable well-being. It may in fact be a good measure of total sustainable well-being—but there's no way to know that, not without some additional and separate measure of sustainable delivered well-being against which to test it.
Assessing Gross National Happiness
Gross National Happiness, developed in Bhutan, uses survey research to determine the quality of self-reported well-being—happiness—experienced within a society.
The concept traces to an announcement made in 1972 by Bhutan's then king, Jigme Singye Wangchuck, who had recently succeeded his father. Bhutan, a Buddhist nation, had long been isolated from modern commercial society, but change was clearly in the offing. On assuming office, the young king assured his people that he would be more concerned about Gross National Happiness than Gross National Product. The Centre for Bhutan Studies, under the leadership of Karma Ura, undertook to give practical effect to the king's commitment by developing a survey instrument to measure the country's GNH.
The Canadian health epidemiologist Michael Pennock had a major role in the design of the instrument; many of the topics covered by the GNH survey are well-established metrics by which Health and Human Services administrators have long assessed the status, and the outstanding needs, of the populations they serve. A pilot survey was completed in 2006 and the first full survey was done in 2007, with a sample size of 950. It included 750 variables. The length of the survey, and the challenge of reaching a rural population in a mountainous country with few roads, made it difficult for enumerators to visit more than one respondent per day.
The survey instrument collects data about life conditions and subjectively reported well-being in nine broad domains:
- economic well-being;
- psychological well-being;
- time use;
- community vitality;
- cultural participation;
- environmental quality; and
- good government.
Ura and Pennock have also collaborated on the development of policy screening tools that can be used to examine the potential impacts of projects or programs on GNH—resulting in what might be called a "GNH Impact Statement," similar to an Environmental Impact Statement. Use of one such tool played a significant role in Bhutan's rejection of membership in the World Trade Organization. After completing an exercise based on the screening tool, it was decided that participation in the WTO would either lower GNH or have an impact that was unknowable. (Wisely, Bhutan has implemented a conservative decision-making process: In the face of unknown consequences, the default is "no action.")
Confident that the indicator set represents the aspects of life that Bhutanese citizens value most, its Gross National Happiness ministry has undertaken a nationwide program of "Education for GNH," designed to adapt curricula in the country to be productive of the outcomes that the GNH instrument assesses.
Many of the questions in the Bhutanese survey are specific to the Bhutanese experience of happiness and well-being, but the principles behind them are transcultural. For instance, the Bhutanese survey asks about participation in festival days, which are important in village community life. When translated to another culture, the question, at its most generic, is about instances of positive, collective, and celebratory social engagement with neighbors and fellow citizens.
Even questions that appear completely specific to Bhutanese culture—How often do you think of your karma?—can have transcultural application and be scored on a standard Likert scale from zero (not at all) to five (all the time). "Karma," Ura has said, "is, in this context, moral responsibility. This is like asking: How often do you think of the moral consequences of your actions?" Michael Pennock uses (what he calls) a "de-Bhutanized" version of the survey in his work in Victoria, British Columbia.
When tailored to a local culture, GNH offers an accurate and comprehensive assessment of self-reported well-being in that society. The nine domains capture elements of well-being not included in the GPI assessment, and are particularly strong in assessing psychological well-being and the strength of social capital—community vitality, cultural participation, good governance, and civic engagement.
But GNH has other limitations. While self-reports can capture many aspects of well-being, not all self-reported well-being is actual well-being. Survey research is not an appropriate methodology for assessing some crucial metrics, notably environmental harms: Respondents could be misinformed or uninformed about the quality of their environment. The ability of that environment to provide ecosystem services now and into the future is a matter for objective assessment, not survey research. That GNH lacks this objective measurement of environmental health prevents it from being a measure of ecologically sustainable well-being.
The Need for a GPI+
Some form of GPI+ is needed to marry GPI and GNH. By combining the strengths of both, GPI+ would address the shortcomings of each, and in so doing become an accurate measure of ecological, economic, and social sustainability. To get to that point, GPI+ must first address a few flaws.
Since personal consumption, a GDP-derived figure, is the base from which GPI makes its cost-benefit adjustments, GPI continues to encode GDP's assumption that increased personal consumption is always and forever a good thing. This is not a realistic assumption. It contradicts an emergent truth that is supported by standard economic theory and actual research: Beyond certain income levels, additional income (and personal consumption expenditure) has a diminishing marginal return. Adjusting for income inequality doesn't fully correct this problem.
The particular algorithm that incorporates this diminishing marginal utility of additional income will in all likelihood vary from culture to culture (or from subculture to subculture within a culture), and would be most accurate if it were constructed from survey research data from the population being assessed by the macro indicator. (Meaning, inferences from one country to another may not be justified and should probably be avoided.) This would allow the accuracy of the adjustment to be tested against reality.
GNH, on the other hand, does not effectively assess ecologically sustainable well-being. As noted, survey research has limitations as a means of assessing environmental quality and the consequent condition of natural capital (though it can provide useful information about citizen perception of environmental quality and natural capital services).
Bhutan's population is rural and largely agricultural, and as such it retains a higher degree of competence to assess environmental quality than the urban populations of many other nations, which have lost much of their indigenous and traditional knowledge of how natural systems operate and how human actions affect them. Thus, to the extent that GNH can operate as an indicator of ecologically sustainable well-being, it relies on elements of social capital—shared systems of valuation and publicly held knowledge—that industrial-urban nations tend to lack.
Policymakers always face constrained choices—no government budget has ever been infinite. But for the past century humans have benefited from turning past solar income, stored as fossil fuel, into present wealth. As civilization passes Hubbert's Peak and enters an era in which the current subsidy to the economy from past solar income must necessarily decline, additional pressure will be placed on governments to provide their services in the most cost-effective manner. That pressure is being met, in part, by outcomes-based budgeting: the effort to identify the expected costs and benefits for each policy or program that a government chooses to pursue, so that rational, efficient, value-maximizing decisions can be made.
Most approaches to outcomes-based budgeting assess the achievement of "benchmarks," which are proximate rather than ultimate ends. To be made fully rational and efficient, outcomes-based budgeting must also assess the contribution of such benchmarks to ultimate ends and goals, and to the achievement of the overall function of the social system: the enjoyment of sustainable well-being among the populace. The use of survey research is necessary to this purpose.
A comprehensive indicator set, such as GPI+, would allow the extension of outcomes-based budgeting to this macro level by incorporating assessments of the degree to which ultimate ends are achieved. Such an indicator set would enable policymakers to find rational, data-based answers to questions like:
- Will a dollar spent on repairing roadways bring more or less benefit to citizens than a dollar spent on education?
- Which will have a larger payoff in well-being: addressing food insecurity or environmental quality?
- Would the well-being of our people be increased more by an investment of time and effort in cultural programming or in civic engagement?
Combining elements of GNH and GPI into a single indicator set that makes use of subjective reports of well-being thus has several advantages. It represents a more accurate measurement of society's sustainable delivered well-being; it is more readily adaptable to the experience of varied nations and communities because it is capable of assessing benefit flows from built, natural, and social capital; it offers an effective means to further the implementation of outcomes-based policy administration; and it allows the resulting indicator set to be modified to ensure that its assumptions correlate with the lived experience of the citizenry.
The GPI+ Model: A Work in Progress
"GPI+" is at this point a concept, not an accomplished indicator set. The exact mix of indicators that will comprise GPI+ remains open to discussion and consensus-building among stakeholders—including policymakers, legislators, academics, nonprofit administrators, and other practitioners. It's likely that GPI will continue to be compiled in a fairly standardized set of metrics, and that the "plus" part of the measure will be what's at issue.
In the decisions about which indicators will be included, cost of compilation will necessarily be a concern; to the extent practicable, the indicator set should draw on existing databases to avoid incurring additional data-collection expenses. Part of the necessary preliminary work, then, is to inventory existing datasets and to identify appropriate indicators that could draw on them.
One of the strengths of GPI is that it translates indicators into monetary costs and benefits, giving a single number than can readily be compared to the monetary figure for GDP. But some of the non-economic aspects of well-being that GPI+ proposes to include are not readily reducible to monetary costs and benefits. (For example: What is the cost to society of homelessness and substandard and expensive housing?)
Should the indicator set nevertheless proceed to generate monetary values for these metrics, through assumptions and proxies similar to those used for other indicators that are not obviously monetary? (The cost of giving all citizens access to affordable housing could be computed from average rental costs, average construction costs, and the like, and that cost could stand as the social cost of substandard housing.)
Or should GPI+ consist of a standard GPI figure supplemented by a dashboard of other, non-monetary metrics? This remains an outstanding philosophical and methodological question that needs to be answered by the community of practice. One possible synthesis: The non-monetized indicators collected into a dashboard could be assigned scores—more points for better values, fewer points for less desirable values—so that this portion of the indicator could also be summed in a single number.
Several other comprehensive systems of assessment proceed in this way, notably the Failed State Index, which is compiled annually by the Fund for Peace and Foreign Policy magazine. The assignation of point values is admittedly discretionary, but not totally arbitrary or subjective; and continuity of method between years allows for meaningful comparison.
Next Steps: Vermont Forges Ahead
In the spring of 2012, Vermont became the second state to endorse officially the compilation of a state GPI, and the first to do so through legislative authorization. The legislation directed the Gund Institute for Ecological Economics to prepare a GPI estimate and to report it to the legislature and the Administration in the spring of 2013.
Separately, at the behest of Gov. Peter Shumlin, Secretary of Administration Jeb Spaulding oversaw a process by which state agencies began implementing an outcomes-based strategic planning process, identifying benchmarks they propose to achieve. The baseline from which progress will be measured is described by a variety of existing datasets that are called, collectively, "The Governor's Dashboard," which was unveiled in the spring of 2012.
The elements of the Dashboard measure some of the things that a GPI+ for the state would measure, and they supplement Gross State Product as an indicator of both economic and total well-being. (Indicators include a set of metrics that surveys the health of the agricultural economy, a set of measures that assess educational attainment, and such other measures as rates of infant mortality, obesity, and coverage by health insurance. There are no subjectively reported elements.)
Clearly measures like "poverty rate by county," "percentage of Vermonters who smoke," "infant mortality," and "high school graduation rate" offer the opportunity for fairly clear linkages between policy goals and monetary expense.
The relationship between Vermont GPI and the Governor's Dashboard remains to be worked out. Together they represent a step toward the GPI+ that's described here but fall short of its full realization. Whether the two sets of indicators will be integrated as a policy tool depends very much on the manner of use both will have within the legislature and state agencies; strong support for integrated use of GPI and OBB has come from both.
Additional strong support has come from the office of newly elected state auditor Douglas Hoffer, who foresees using the cost-benefit structure of GPI and its valuation categories to inform the criteria against which state programs and efforts will be audited. This represents a new way of thinking about governmental efficiency, and a new reading on the role of the office of auditor.
If these efforts lead to practical integration of GPI, OBB, and subjectively measured well-being in the state of Vermont, it is likely that the state's experience will offer useful guidance to those making similar efforts in other states, and perhaps even those making similar efforts in other nations.