There’s a lot of interesting stuff around just now on the question what we should and shouldn’t do with measures of aggregate economic performance and welfare. I talked about this in my BrisScience lecture. I make (again) the point the Gross Domestic Product is a bad measure of a nation’s economic welfare because it’s Gross (doesn’t net out depreciation of physical or natural capital), Domestic (doesn’t net out income paid overseas) and a Product (takes no account of labour input)).
But if GDP isn’t a good measure, what is? There are a bunch of alternatives in the air at present such as Gross National Happiness and the Genuine Progress Indicator (the latter has been advocated by Clive Hamilton and the Australia Insitute. These ideas have been getting a fair bit of criticism lately. Andrew Leigh has a go at Gross National Happiness while Nick Gruen writes on the Genuine Progress Indicator for New Matilda. This is subscription only, unfortunately, but when Nick completes his two-part paper, I’ll try to comment more. Andrew Norton at Catallaxy has also written a lot on this.
My general view is close to Nick Gruen’s. We should be trying to get at a Net measure of Full Income (including leisure and taking account of resource stocks) but none of the attempts so far have been really satisfactory. More on this when I get some leisure (As If!).
John Quiggin 
JQ, I have long been and advocate of the Genuine Progress Indicator (GPI) amongst other alternatve economic viewpoints, probably why I am a member of the Post Austitic Economics Society. The following is a paste from the Redifining Progess peoples web site about GPI, it may at least get some discussion going about that long forgotten notional concept of scarcity when we sprout on about the economic view.
“…The GPI starts with the same personal consumption data the GDP is based on, but then makes some crucial distinctions. It adjusts for certain factors (such as income distribution), adds certain others (such as the value of household work and volunteer work), and subtracts yet others (such as the costs of crime and pollution). Because the GDP and the GPI are both measured in monetary terms, they can be compared on the same scale.
I. Crime & Family Breakdown
Social breakdown imposes large economic costs on individuals and society, in the form of legal fees, medical expenses, damage to property, and the like. The GDP treats such expenses as additions to well-being. By contrast, the GPI subtracts the costs arising from crime and divorce.
II. Household & Volunteer Work
Much of the most important work in society is done in household and community settings: childcare, home repairs, volunteer work, and so on. These contributions are ignored in the GDP because no money changes hands. To correct this omission, the GPI includes, among other things, the value of household work figured at the approximate cost of hiring someone to do it.
III. Income Distribution
A rising tide does not necessarily lift all boats — not if the gap between the very rich and everyone else increases. Both economic theory and common sense tell us that the poor benefit more from a given increase in their income than do the rich. Accordingly, the GPI rises when the poor receive a larger percentage of national income, and falls when their share decreases.
IV. Resource Depletion
If today’s economic activity depletes the physical resource base available for tomorrow’s, then it is not really creating well-being; rather, it is just borrowing it from future generations. The GDP counts such borrowing as current income. The GPI, by contrast, counts the depletion or degradation of wetlands, farmland, and nonrenewable minerals (including oil) as a current cost.
V. Pollution
The GDP often counts pollution as a double gain; once when it’s created, and then again when it is cleaned up. By contrast, the GPI subtracts the costs of air and water pollution as measured by actual damage to human health and the environment.
VI. Long-Term Environmental Damage
Climate change and the management of nuclear wastes are two long-term costs arising from the use of fossil fuels and atomic energy. These costs do not show up in ordinary economic accounts. The same is true of the depletion of stratospheric ozone arising from the use of chlorofluorocarbons. For this reason, the GPI treats as costs the consumption of certain forms of energy and of ozone-depleting chemicals.
VII. Changes in Leisure Time
As a nation increases in wealth, people should have increasing latitude to choose between more work and more free time for family or other activities. In recent years, however, the opposite has occurred. The GDP ignores this loss of free time, but the GPI treats leisure as most Americans do — as something of value. When leisure time increases, the GPI goes up; when Americans have less of it, the GPI goes down.
VIII. Defensive Expenditures
The GDP counts as additions to well-being the money people spend just to prevent erosion in their quality of life or to compensate for misfortunes of various kinds. Examples are the medical and repair bills from automobile accidents, commuting costs, and household expenditures on pollution control devices such as water filters. The GPI counts such “defensive” expenditures as most Americans do: as costs rather than as benefits.
IX. Lifespan of Consumer Durables & Public Infrastructure
The GDP confuses the value provided by major consumer purchases (e.g., home appliances) with the amounts Americans spend to buy them. This hides the loss in well-being that results when products are made to wear out quickly. To overcome this, the GPI treats the money spent on capital items as a cost, and the value of the service they provide year after year as a benefit. This applies both to private capital items and to public infrastructure, such as highways.
X. Dependence on Foreign Assets
If a nation allows its capital stock to decline, or if it finances its consumption out of borrowed capital, it is living beyond its means. The GPI counts net additions to the capital stock as contributions to well-being, and treats money borrowed from abroad as reductions. If the borrowed money is used for investment, the negative effects are canceled out. But if the borrowed money is used to finance consumption, the GPI declines.”
Isn’t subtracting medical costs from car accidents too simplistic? Of course fewer car accidents would make us better off, but it sounds as though GPI treats an accident victim who has their leg cheaply amputated as better off than someone who has it expensively returned to usefulness. Isn’t this medical care a (human) capital investment with future returns?
I am new to this question, so excuse me for some (possibly) random thoughts on it.
Mike,
Point by point, then. Several of these I do not disagree with, but have some questions.
I. Crime & Family Breakdown
Like Tom’s comment on health, I am not sure that treating spending on mitigating and trying to solve human fallibilities as a cost is appropriate. There seems an implicit assumption that without this spending we would be better off. This is (IMHO) incorrect.
II. Household & Volunteer Work
Agreed – if it is possible to calculate the value of this. For example, if I spend a few hours in the garden, I am unlikely to achieve as much as a professional gardener would in half the time. Measuring quality becomes important.
III. Income Distribution
Again, measurement would be a problem. How do you measure the relative differences? Is it some function of the Gini curve and how do you justify the differences?
IV. Resource Depletion
No argument.
V. Pollution
No real problem, here – but some of the measurements may be subjective.
VI. Long-Term Environmental Damage
Isn’t this the same as the above?
VII. Changes in Leisure Time
Interesting study recently here, where measurement of actual leisure comes in. Do you define leisure as time away from work, or do you exclude work around the house (see II above)? What is leisure is probably also fairly subjective. If I enjoy my work, is it “better� than a job I hate?
VIII. Defensive Expenditures
See (I) above.
IX. Lifespan of Consumer Durables & Public Infrastructure
The problem here is that, to achieve a “fiddle� of the figures a short-term reduction in capital expenditure would do it.
X. Dependence on Foreign Assets
This treatment makes sense.
Perhaps a solution would be to adopt a corporate style approach – present both a profit & loss statement and a balance sheet. Recognise capital expenditure as an increase to national wealth, for example, and then depreciate over the life of the asset.
This solution would (IMHO) cover many of the issues above. The problems of valuation remain.
It might be useful to take the shift from GDP to GPI in steps.
Going from GDP to Net National Income in the SNA accounts for IX and X
Including stocks of natural resources in the capital stock accounts for IV and VI
Using a ‘full income’ measure or measuring income/hour worked accounts for II, VII and some of VIII
Valuing negative externalities accounts for V and VIII
I don’t think there’s much dispute in principle about any of these.
What’s left then are I and III. On III, I don’t think the approach of making a deduction from an aggregate number is the best way to go. I’d prefer either to look at median income or to give separate measures of mean and dispersion.
On I, the costs of crime can in principle be treated like pollution. But I don’t think there’s much value in trying to put a money cost on divorce and family breakdown. Even to determine whether increasing divorce is bad, you’d need to put cost on unhappy marriages, and I can’t see that economists have much of value to contribute here.
How do you work out the “stocks of natural resources”, given that it changes with changing technology? eg. Uranium wasn’t a valuable resource before nuclear power, if we get thorium reactors then that will be a resource otherwise its probably not much of one? Additionally technology makes resources available that weren’t previously.
I guess also resource stocks can be devalued by changing demand.
On reflection, I think the GPI model of health is correct if you consider the only benefits of treatment to be that people can go back to work — then improved rehabilitation will show up in the GDP figures.
I think that GDP is what it is and should be supplemented with other measures rather than replaced. Anybody who claims to have one number that encasulates happiness is seriously deluded. If anything it is multivectored and dynamic.
Also amoungst the limitations of GDP is that it measures the production of goods and services for trade. If I increase my welfare by renovating the garden, tidying my house or doing some exercise this is not reflected in GDP. I am sure that John Quiggin has gained a lot of welfare from his engagement in martial arts and most of this is the result of personal effort for the self, not tradable production that would show up in the GDP figures.
I generally think of GDP as the taxable part of the economy. Roughly speaking it represents the tax base (although acts such as conscription do bipass it). GDP is pretty central to the welfare of government which is separate from the welfare of the people.
I think Liberty is a more tangible political goal than happiness. Governments, legislation and rules can certainly make us miserable and should ideally be designed in such a way to minimise unintended misery. However governments are not generally the source of our happiness.
Microeconomic theory measures welfare (utility) in two main ways. One is as a function of the goods and services consumed (direct utility function). The other is as a function of the prices and income faced by the person (indirect utility). If prices are defined appropriately these two approaches are theoretically equivalent – but have different measurement implications.
Generally people couch discussion of broadening welfare measures in terms of the first approach. ‘Goods and services’ are broadened to include environmental and social goods and home-produced services.
With respect to valuing leisure and home production at least, I think there is a lot of merit in using the indirect utility approach. In this case, this amounts to valuing people’s welfare according to their non-wage income and their wage rate. This focuses on the opportunities available to people rather than their actual choices – and is easier to measure (eg WRT trends over time). The direct approach tends to get tied in knots about how to treat self-consumed home production.
More speculatively, I wonder whether this approach could be used for social and environmental benefits. For example, if general crime rates are high, the price of living in a low crime neighbourhood might be measured in terms of the cost of moving to a middle class area.
I am on the same wavelength as Terje – GDP itself is innocent of all charges, as it a pretty good measure of the level of economic activity, and so is useful for things like running current macroeconomic policy and analyzing linkages between particular variables and the rest of the economy.
The problem is when GDP as a proxy for welfare, and the availability of GPI and GNH data is welcome as it makes using GDP for welfare measures unnecessary as well as undesirable.
none of the attempts so far have been really satisfactory
And it will ever be thus, because what’s an important measure of “progress” or “happiness” to one person is trivial, or even erroneous, to another. An increase or decrease in such measures tells us more about what the compilers think is a “good society” than whether we are actually achieving a good society.
Not to mention how the extreme ease of manipulation gives plenty of scope for those more interested in playing politics than advancing knowledge. What you put weight on in your index predetermines the policy prescriptions that flow from your results.
It is all well and good to say that GDP isn’t a good measure, but at least it is measurable. The other alternatives proposed above are both subjective and virtually immeasurable.
the critique of GDP (or GNP) has been around for a very long time – it has regularly resurfaced in debate in statistics, economics and public policy circles; for a nearby view (ie NZ) I recall for example Marilyn Warings criticisms of it from a feminist perspective in the 1980s. There are strong arguments for a move to a more useful and comprehensive measure is only someone could solve the problem of measurement. there are techniques for quantifying eg the value people get out of leisure or recreation (hedonic pricing, avoidance cost etc.) but these are hugely expensive to do on a national scale. anyone out there got an answer?
the critique of GDP (or GNP) has been around for a very long time – it has regularly resurfaced in debate in statistics, economics and public policy circles; for a nearby view (ie NZ) I recall for example Marilyn Warings criticisms of it from a feminist perspective in the 1980s. There are strong arguments for a move to a more useful and comprehensive measure is only someone could solve the problem of measurement. there are techniques for quantifying eg the value people get out of leisure or recreation (hedonic pricing, avoidance cost etc.) but these are hugely expensive to do on a national scale. anyone out there got an answer?
Bruce, I was thinking along the same lines. I might try a post on this, if I can get a bit of free time to work it up.
sorry about missing bruce’s comments, which are apposite to the comment I make about the measurement problem; although I still think the measurement issue is much more a practical one – how to estimate with anything like a reasonable error bar for a whole country – rather than a theoretical one. Also apologies for posting twice – both of these failings I attribute to my computer rather than my intentions. At this point my computer will crash to pay me back for blaming it unfairly…
Three points
1) The Greens already include having the ABS measure GPI in their party platform.
.
I think it is a great idea, though I would do it experimentally for 5 years or so so it can be properly debated before making it official.
.
2) There are alot of industries who would be strongly against this. The mining industry for obvious reasons, and anyone else whose negative externalities are not measured at the moment. It might be politically difficult to implement. I guess it could be done privately, at least initially, but I would be surprised if any NGO had the resources to do it. Might be a good thing for the think tanks to get stuck into.
.
3) If GPI were measured across alot of countries which were then ranked according to GPI per person Australia might come out right at, or very close to the top. Alot of our quality of life comes from intangibles that are not measured in GDP and our natural resource base is enormous.
“How do you work out the “stocks of natural resourcesâ€?, given that it changes with changing technology? eg. Uranium wasn’t a valuable resource before nuclear power, if we get thorium reactors then that will be a resource otherwise its probably not much of one?”
I do not see a problem. If new technology makes previously useless natural resources valuable then our natural resource base has increased and should be measured as such.
Let’s not forget that the ABS has been publishing measures of Australia’s progress, based on ideas not altogether unlike the GPI, for a few years now. See the Summary Indicators at the ABS website here.
Perhaps the best known alternative indicator of national well-being is the simple Human Development Index used by the UN in its annual Human Development Reports. Interestingly, since national performance can be ranked both by GDP/capita and by the HDI, it is quite common to see countries’ performance on each indicator compared. When this is done, some countries are seen to perform much better on human development than their fairly poor GDP/capita outcomes would lead one to expect. From memory, Cuba is a country which is in this category. By contrast, there are some “rich” countries (with high GDP/capita) which perform surprisingly poorly.
This use of the HDI may well point the way to the best use of alternative indicators. I doubt whether any single all-inclusive indicator of national progress or well-being is possible. What is the point of asking “How are you doing?” and getting the answer “42”? It is pretty meaningless. But relative values and trends of several indicators may well be valuable both as national time-series and for cross-country comparisons.
Aren’t we wasting time on the ‘pursuit of happiness’? The writers of the US Constitution got the phrase from Samuel Johnson’s novel ‘Rasselas’, where (from memory) the philosopher Imlac points out that the ‘p of h’ is stupid and pointless, because happiness arises out of other courses of action, e.g. trying to follow Jesus’s 2nd commandment? (My example, not Imlac’s, not inserted for theological but rather for practical reasons.) If happiness is pursued of itself it will stay tantalisingly out of reach.
O6,
Your latest sentence reminds me of a story in which there are two cats. The older cat asks the kitten why it is chasing its tail. The kitten replies that “happiness is in the tip of our tails and I want to catch it”. To which the older cat replies “yes happiness is in your tail but if you get on with the days business it will follow you everwhere”.
Regards,
Terje.
[…] People mostly refer to Gross Domestic Product, which is much the same magnitude, but not a useful concept except in measuring shortrun economic activity posted on Sunday, March 2nd, 2008 at 9:00 am […]