We, including many economists, sometimes forget what is economic growth in a normative sense, that is, what we should count as “good” economic growth. Economic growth does not consist in producing more of this or that good (or service). It does not even necessarily consist in producing a larger quantity of all goods. Nor does it consist in producing the largest value of goods calculated by weighing the quantities produced with any set of prices. To have any normative (moral) significance, to be evaluated as good or not good, economic growth requires more than that.
It is not surprising that, in order to say something about the goodness of economic growth, we need some moral criteria. The evaluation of any policy or social situation ultimately requires the same. This is what nearly a century of “new welfare economics” and several decades of “social choice theory” should have taught economists. (I reviewed some aspects of this idea in a 2006 Independent Review article, “Social Welfare, State Intervention, and Value Judgments.”)
Define economic growth as increasing consumption and wealth; and wealth as future (discounted) consumption, which will flow from capital such as machines, factories, office buildings, cars, or houses. Nothing original in that. But is consumption anything that some political authority wants people to consume? At least if our moral criterion is individualist, consumption is instead what individuals themselves want to consume. And there is no other way to measure what individuals want, given the scarcity and cost of things, than by observing the demand they express on free markets.
This leads an economist to refine his definition of economic growth as the increase in the total consumption of goods and services by all individuals, calculated by weighting the quantities by the prices determined on free markets. (Adding up apples and oranges is impossible without weighing factors.) Together with supply and cost, market demand determines the relative prices of different things given the preferences of all individuals. When they are determined by supply and demand on free markets, prices lend normative significance to the value of the total consumption. This does not measure the elusive concept of total welfare but it is as close an indicator as we can get.
The implications of this idea are deeper than it appears at first sight (which is one reason why economic theory is useful), and a short post cannot do justice to all the related problems. But it is difficult to develop or discuss any non-arbitrary concept of economic growth without considering this starting point. “Non-arbitrary” means grounded in the preferences and consent of all individuals as when we adopt the individualist moral criterion of classical liberalism and libertarianism.
To see how this makes sense, consider the polar opposite: economic growth as the increase of the production and consumption of what is deemed to be worth more by some dictator or by some group of experts or by some elected assembly or by some mob. In such a case, economic growth has no more normative foundation than, say, the rate of growth of the wealth of Louis XIV’s family (and court). Here is a numerical example.
Assume an economy with only two goods: apples and oranges. Further assume that every individual in that society prefers apples to oranges, except those in Louis XIV’s household and court. Louis XIV (or his faithful Finance Minister, Jean-Baptiste Colbert) determines that an apple is worth one denier (a unit of currency at the time); and an orange, two deniers. He orders that, in year 1, 1000 apples be produced for his subjects and 500 oranges for his household and court, for a total “value” of 2,000 deniers ([1000 X 1] + [500 X 2]. In year 2, the king succeeds in increasing the volume of oranges produced by 10% while the production of apples remains the same, for a total production “value” of 2100 deniers ([1000 X 1] + [550 X 2]). The rate of economic growth has thus been 5%. In our individualist perspective, this number has no normative value at all.
In an individualist and classical-liberal perspective, as opposed to an absolute-monarchist perspective or an absolute-majoritarian perspective or any other dictatorial or collectivist system, the morally desirable rate of economic growth corresponds to the growth of what individuals want to consume weighed by the prices that they themselves contribute to determine on free markets according to their own preferences. Because all consumers adjust their behavior to these prices, the price of a good represents the marginal value that each consumer attaches to it: this is what gives normative significance to the growth of production.
READER COMMENTS
Jon Murphy
Jul 14 2021 at 11:02am
Good discussion. The one thing I would add is even modern statistics of economic growth are limited. Simon Kuznets, in his discussion of GDP (which he invented) argues that interpreting GDP as a measurement of economic growth only makes sense in a market economy. As I recall, that reason is why he initially wanted to keep government spending out of GDP as the market price system for government expenditures didn’t give quite the same signal as for private transactions, and thus one couldn’t really add G to private consumption as they are not comparable.
So, when we discuss things like the economic growth of non-market economies (eg USSR, China, Cuba, etc), comparing their GDP or GDP per capita to market economies doesn’t quite work. GDP doesn’t give a full story. The USSR always looked impressive on paper, but once things started opening up, the truth came out: their “growth” was all imaginary. From my admittedly-limited and anecdotal experience, China is much the same way. They are not nearly as wealthy as their GDP, or even their GDP per capita, would suggest.
By the way, unfortunately I cannot provide the exact location for Kuznets discussion of GDP. My library is in storage.
Pierre Lemieux
Jul 14 2021 at 1:12pm
Jon: We are probably saying fundamentally the same thing, that is, a necessary condition for GDP to have normative significance is that prices be determined on free markets.
It is a related but different matter to (correctly, like Kuznets) argue that the government’s contribution to GDP cannot be measured by its expenditures, that is, by the value of the inputs it purchases to produce its services. The reason is that these inputs would have rendered equivalent services, or at the very least services of not-zero value, elsewhere in the economy. The contribution of, say, Ford to GDP is not the value of its inputs but its profits (its “value added”); so should it be for the government. This, I think, is the argument.
You must not forget that GDP is, by definition, the value of production; and that its alternative calculation from the expenditure side (Y=C+I+G) only provides an equivalent measure of GDP because everything that is produced is consumed (an accounting identity). The G there is not the government’s contribution to GDP but what it consumes out of GDP. (I quickly mentioned this issue of government expenditures as government production in my Regulation review of a book by Diane Coyle.)
The caveats in your second paragraph are well taken.
Jon Murphy
Jul 14 2021 at 3:18pm
True true. Good point
Scott Sumner
Jul 14 2021 at 5:53pm
Jon, China is nothing like the old USSR. Visit any Chinese shopping mall and you’ll immediately see the difference. My impression (based on frequent visits to China) is that China’s GDP/person relative to the US is probably higher than the official figures suggest.
Pierre, I had trouble following your post. Are you measuring real or nominal GDP? If real, how do you overcome the index number problem? What has happened to the price of telephones since 1990?
Jon Murphy
Jul 14 2021 at 6:08pm
I’ll trust you on China. That’s why I qualified my comment. I’ve not been in a while and I’d not be surprised to hear things have changed
Pierre Lemieux
Jul 14 2021 at 6:37pm
Scott: I am abstracting from inflation and discussing only relative prices. I am not trying to overcome the index-number problem, which is what I tried to convey with my caveat:
What we do know is that the farther prices are from free-market prices, the farther the indicator is from any normative value. Does that make my argument clearer?
(For our readers, the classic article on evaluating GDP is Paul A. Samuelson, “Evaluation of Real National Income,” Oxford Economic Papers, January 1950.)
Scott Sumner
Jul 16 2021 at 2:09pm
I’d put it this way. Other things equal, estimates of an economy using market prices are better than estimates using non-market prices. However, there are economies where market prices are almost meaningless indicators of economic growth due to hyperinflation. Thus the official GDP indicator in a country like the Soviet Union might have measured growth more accurately (despite all their problems) than the market prices in a country with hyperinflation.
So again, I agree that other things equal, market prices are much more meaningful than non-market prices, but growth estimates based on market prices aren’t very meaningful in an absolute sense where there’s lots of inflation.
David Henderson
Jul 14 2021 at 4:25pm
Good post, and especially good timing, given that it’s July 14.
Pierre Lemieux
Jul 14 2021 at 6:38pm
David: I am so removed from national days that I did not notice that, at least consciously.
David Seltzer
Jul 14 2021 at 5:53pm
Pierre: you wrote, “the price of a good represents the marginal value that each consumer attaches to it: this is what gives normative significance to the growth of production.”
As a former market-maker and hedge fund manager competing in various capital markets, the accepted moral “ought” was; your word is your bond. In my experience, one data point, traders adhered to that tenet out of self-interest. The implied contract was mutual trust. When an instrument traded, both parties were confident profits would be paid and losses absorbed. In that arena, I often heard various traders being referred to as fair people.
john hare
Jul 14 2021 at 6:29pm
There is a scenario that interests me, and arouses my curiosity as to how an economist would respond. This is a contrived scenario that I consider impossible, but it stands in for a number of others that might come about in the fullness of time.
A wave of overpowering morality sweeps the country such that people quit committing crimes. The prisons slowly empty and a couple of million go from a drain on the economy to productive citizens. Police force and private security employment drops over a similar period of time and these people start producing something tangible instead of preventing harm. A high percentage of lawyers, locksmiths, and judges join the well educated workforce.
So in this fantasy scenario, several million people start producing value, though at the same time most of their wages drop as they enter unfamiliar professions. Lawyers taking massive pay cuts in particular. The vast majority of the population is far better off as they are no longer suffering from crime, or supporting those committing, adjudicating and punishing it. The question being, has the economy grown or contracted from an economists’ viewpoint??
That is clearly a fantasy scenario. The real question is, “What happens in fields where an economic deadweight becomes an asset, but slows throughput of cash?”
Phil Murray
Jul 15 2021 at 9:54am
John: I think the economy has grown. If we behave ourselves, we can devote fewer resources to producing and consuming locks, security services, and the like. We can devote more resources to producing and consuming other goods and services we deem to be more valuable. Don’t worry about lawyers’ incomes or cash. People who used to spend $300 an hour on lawyers that now spend $50 an hour will be happy to buy other goods with the $250. Lawyers who don’t like that can acquire new skills. Your scenario is similar to how much better off we are in peace than perpetual war.
John hare
Jul 15 2021 at 11:56am
I agree that the true economy has grown. The problem that I see is that some measures seem to conclude that the $250.00 an hour haircut on the lawyers represents a contraction. I’m asking about accounting, not true productivity.
Pierre Lemieux
Jul 15 2021 at 10:50am
John: Your scenario is not implausible: in fact, it happens all the time. In the early 20th century, the resources used to manufacture horse buggies and most of the labor used to produce phone calls were rapidly diverted to other consumer demands. The number of persons occupied in agriculture declined from 90% in the late 18th century to 1% today; agricultural labor moved to other industries. The container eliminated most of the dockers’ jobs. Same for people who made slide rules and, a bit later, pocket calculator. Consumers, who loved beaver hats in the 18th century, changed their minds, eliminating the industries of beaver hunting and trade. And so forth. The consumers benefited from that; otherwise, they would have continued demanding manually-switched phone calls and horse buggies. To answer your question succinctly, we know (if we value individual choices and human flourishing) that “the economy has grown” because it is producing more of the things consumers want and less of what they don’t want or don’t want anymore.
Note two further points (again very succinctly): In the economic perspective, we produce to consume, not the other way around. And even lawyers like cell phones, cars, low shipping costs, etc.
Thomas Lee Hutcheson
Jul 14 2021 at 10:33pm
I’m with you 95% of the way, but the relative prices that obtain in the market depends, given the distribution of preferences, on the distribution on of income.
If Louie REALLY likes apples compared to oranges and receives a huge share of French national income, the free market price of apples relative to oranges will be high, regardless of how much everyone else may like oranges.
Pierre Lemieux
Jul 15 2021 at 10:22am
Thomas: That’s a very good point. I now regret to have erased a caveat where I explicitly recognized the issue. Here is what I would have added. The normative judgments that are are at the basis of policy evaluations are essentially, for most people, about the distribution of income. If Joe agrees that most of the income should go to Louis XIV (through the taxes, implicit and explicit, that expropriate the production of ordinary people), then Joe would argue that my numerical example represents good economic growth (as you point out). Not so for somebody who believes in the formal equality of classical liberalism as a means to general human flourishing. Now, it is a result of both welfare economics and social choice theory that these divergent value judgments cannot be aggregated and that a “social welfare function” must be either nonsense or imposed by some dictator. Consequently, there is really only one solution: laissez faire!
Gene Laber
Jul 15 2021 at 8:23am
Pierre and Jon,
“The G there is not the government’s contribution to GDP but what it consumes out of GDP.”
At the federal level G includes such things as defense spending, data collection and analysis by BEA, BLS, etc. At the state and local level it includes fire and police services, road maintenance, etc. Aren’t those outputs? Why are they not production?
Pierre Lemieux
Jul 15 2021 at 10:31am
Gene: That’s the problem. Given that it does not (or even cannot) sell its output at freely determined market prices, what tells us that the government produces what individuals really want? Suppose that the government starts manufacturing four-finger gloves for the invisible minority of citizens who have lost one finger. It hires employees of car manufacturers to do that. What tells ua that consumers did not prefer the (few) cars that are no more manufactured to the many four-finger gloves that are now produced instead? In the national accounts, the whole expenditures of government for manufacturing four-finger gloves would be added to GDP and only the profits lost by car companies would be deducted.
Gene Laber
Jul 16 2021 at 10:44am
Pierre,
I understand the point you are making when you write,
“Given that it does not (or even cannot) sell its output at freely determined market prices, what tells us that the government produces what individuals really want?”
But that reflects problems associated with collective decision making. Suppose a HOA decides to provide garbage collection services for homeowners who were previously contracting individually for such service. Do we know that the decision of the HOA board will result in the same level of output as the sum of the individual choices? But both sets of expenditures would show up in the C component of the GDP identity because there is no government involvement. If, however, a nearby city annexed the development and substitutes its garbage collection services for the individual choices, I interpret your previous comments to mean that since the expenditures now appear as part of G, rather than part of C, you do not want to count them as “production.” Do I misinterpret you?
Craig
Jul 16 2021 at 9:57pm
Could an economy be BOTH growing AND shrinking? Let’s posit a situation where the GDP of China is 1tn RMB and that the exchange rate of RMB to USD is 1:1 so that from a dollar pov it is 1tn USD. But now let’s suppose the FX rate changes so that 1 dollar can buy 11 RMB. We will assume that in RMB terms, the economy of China grew by 10% to 1.1tn RMB but in dollar terms the GDP of China fell to 100bn USD. So from a Chinese pov the economy grew by 10%, from the US point of view, the Chinese economy decreased by 90%
Jon Murphy
Jul 17 2021 at 11:21am
The relevant question is not whether exchange rates have changed, but why. The price of money is a price nonetheless.
To answer your question, we need to know why the exchange rate changed.
Comments are closed.