New Metrics

A Truer Measure of Economic Well-Being? Replacing GDP with GPI, Part One

The chorus decrying the shortcomings of Gross Domestic Product (GDP) as a measure of economic well-being has been rising, crescendoing recently with the release of the report by Nobel Laureates Joseph Stiglitz and Amartya Sen commissioned by then French President Nicholas Sarkozy. Here in the US, states such as Maryland and Vermont are beginning to adopt an alternative to GDP: the Genuine Progress Indicator (GPI).

By Bill Baue | Published 11 years ago | About a 7 minute read

To find out more about this trend, #NewMetrics channel co-curator Bill Baue recently caught up with Eric Zencey, a Fellow of the Gund Institute for Ecological Economics at the University of Vermont who also teaches Political Economy at UVM and Washington University in St. Louis. Zencey played an instrumental role in bringing GPI to Vermont.

Bill Baue: What are the problems with GDP that GPI was created to solve (and how does it solve them)?

Eric Zencey: GDP is a deeply foolish indicator of how we're doing economically. It measures the gross volume of monetary transactions in the economy — but it doesn't care what the money got spent on, and it doesn't care about the elements of our economic well-being that come to us without a price tag. To take up the first category: In GDP, defensive and remedial expenditures are treated as positive contributions to our well-being, when they ought properly be counted as costs. Hurricane Katrina did something like $82 billion in damage in New Orleans, and as that damage is repaired, GDP goes up — but the expenditure doesn't produce an $82 billion increase in our economic wellbeing. It just puts things back — begins to put things back — to where they were. The problem is that GDP doesn't have any way to account for the present service value of the houses, appliances, furnishings, automobiles, roads, etc. that we already have. GDP notices when we buy them, but that's it; it's an amnesiatic indicator. If you bought it last year it's like it no longer exists. So when we lose these things, and the services they provide to us, GDP doesn't have a line item that can show the loss. GPI fixes this by counting the ongoing service value of such things. And it counts the money we spend on them as a cost, not a benefit. The benefit is the use we get from them, not the fact that we spent money on them. In this way, GPI basically applies the principles of double-entry bookkeeping to the economy as a whole. If you're in business you have to deduct expenditures from income to get an accurate sense of how you're doing. Any business person knows that; it's about time we started applying that concept to the whole economy.

To take the second category: We get plenty of economic benefit from outside the market system, and GDP doesn't count these contributions to our wellbeing. I'm not talking about that old saw, "the best things in life are free" — I'm talking about concrete economic benefits, like the work done by volunteers, the do-it-yourself stuff we accomplish at home, the services that we provide to ourselves in households without spending money. If I take care of my aging parents in my house, GDP doesn't know about it. But if I park them in a nursing home, suddenly the service value of their care is monetized and counts in GDP (Whether my parents are better off or worse off depends on their needs and the quality comparison between in-home and nursing home services; but whatever the quality comparison, elder care is an economically valuable service.). Same thing with housework and garden chores: Pay someone to do it, GDP goes up, though there isn't any net gain in the economic benefit you're enjoying (Well, okay, to be precise about it: If you pay someone to do that kind of work, you'll have increased leisure time — which GDP ignores, but GPI counts as a benefit.).

GPI attempts to account for the dollar value of these non-market contributions to our economic well-being.

One of the biggest failures of GDP is that it doesn't deduct environmental costs from economic benefits. We all benefit from a category of good called ecosystem services — the sun warms us and dries laundry on the line; forests provide flood control and climate moderation and carbon sequestration; soil fertility feeds us, and so on. If we cut down a forest to get lumber, GDP counts the value of the lumber that's produced as a positive contribution to our well-being but doesn't deduct the lost ecosystem services as a cost. You can see that this is exactly parallel to the lost cars and houses from Hurricane Katrina: GDP counts the money that moves and ignores the benefits that have been lost. That's why I've been saying GDP is an infinite planet statistic — by not valuing ecosystem services, it's as if it assumes we'll always have enough of them. As we're beginning to learn, that's not true.

Baue: Can you give a sense of how we got here — what's the history behind GPI?

Zencey: As an alternative to GDP, GPI traces back to the work of economists William Norhdaus and James Tobin who, in 1972, came up with a Measure of Economic Welfare (MEW) that corrected some aspects of the GDP accounts. Herman Daly and John Cobb adapted the MEW in 1989 to produce the Index of Sustainable Economic Welfare (ISEW). The basic idea was, count costs and benefits that GDP ignores, put dollar values on them, and see where you end up. The GPI took on its present form in 1995 after some tinkering with the ISEW. The basic framework is, you take a basic measure of GDP (Personal Consumption, which is GDP minus government spending and business investment) and make 24 separate adjustments to it for various costs and benefits.

Baue: Are there any shortcomings that GPI faces? How is it seeking to solve those?

Zencey: GPI is a very rough-cut instrument right now. For instance, it measures "net change in forest cover" and puts a dollar value on the ecosystem services we gain or lose as net forest cover goes up or down. But forest doesn't come as generic acreage — it's very particular. Hardwood forest is different from softwood forest is different from rainforest. Forest in one area may serve as habitat for an economically valuable species, or it may have above-average flood-protection value if it's strategically situated in a watershed. For GPI to be a more precise guide to policy, these broad categories — "net forest change," "net wetlands change," "net farmland change" — could be disaggregated. For instance, other things being equal, forest that's under sustained yield forestry management is more valuable than forest that isn't, and the monetary valuation of its gain or loss should reflect that.

Another issue for some people is those monetary valuations themselves. How do you put a price on something that's not sold on a market? Isn't it necessarily subjective? The answer, in a word, is no. Ecosystem service valuation isn't an exact science, but that doesn't mean it's subjective. We put dollar valuations on things that haven't sold in markets all the time. When you list your house with a realtor, the realtor does a market analysis to come up with an asking price; you don't really know the value of your house until it is sold. Similarly, we can look at comparables for the various things GPI measures and puts a monetary value on. Lost leisure time, or time lost to commuting? Let's value it at the average hourly wage rate. Flood control services of forests? Let's look at damage figures from flooding, or at the cost of building levees and dams (The first is called avoided cost valuation, the second is called replacement cost valuation.). Who decides what valuation system to use? The answer: the community of GPI researchers, meeting annually as a body that sets standards and coordinates practices. We've had two of these meetings so far and we've begun articulating a standardized model for the GPI accounts, and are working our way through some of the technical questions about valuation that need to be answered. In this, GPI is no different from GDP. Most people don't realize it, but the number that gets reported as GDP is the product of thousands of decisions about what will and won't count and how things will be counted.

In part two: applying GPI at the state and company level ...