Metrics of economic performance

This article is basically a rant against popular ways of measuring economic performance, and when they go wrong.

Nothing here should be new to someone who studies economics professionally, but far too many non-experts make these mistakes even when they are otherwise smart people.

This is important because measures tend to get used as metrics when enough people believe in them, and flawed metrics lead to bad decisions.

Failure modes of common measures

The following are two measures that people frequently use as metrics, despite their many flaws.

• Unemployment Rate

  Using the unemployment rate as a metric of economic performance means that you have to oppose many forms of technological progress.

  Almost every major invention that improves the economy in the long term also causes short-term disruptions, which frequently lead to people losing their jobs. These are necessary temporary setbacks that lead to permanent improvements to overall productivity later.

  The industrial revolution would have been a bad thing if you use the unemployment rate as a metric. Think of all the weavers that lost their jobs to the automated weaving loom.

  Paying people to do pointless things like lowers the unemployment rate. This is why it is so disastrous that this is used as a metric. It incentivizes politicians to create pointless jobs, just so that they can claim they lowered the unemployment rate. Some are even arguing for "guaranteed jobs" as if that was a good thing. This is explained far better than I ever could here.

• Gross Domestic Product

  The GDP completely ignores the distribution of the produced goods and services, as well as the purposes to which they are put.

  Adding pointless bureaucratic overhead increases the GDP rather than decreasing it.

  Fixing a mistake (e.g. pollute a river so it needs cleaning, then clean it) increases GDP more than not making the mistake in the first place.

What actually matters about the economy

People disagree wildly on how economic performance should be measured. It's unreasonable to try to come up with a single unifying measure that satisfies everyone. Nevertheless there are two aspects I would like to emphasize.

In practice these two aspects are tightly related to each other, but it's useful to think of them as separate concerns because their problems are of two very different kinds and many measures people use in practice only ever look at one of them:

• Total production of goods and services

  This is the aggregate of all goods produced and all services rendered in a society. It does not matter by whom or to whom, but it does only take into account goods and services that improve the lives of people directly.

  The problem here is difficult but largely objective: We need to figure out how to quantify and compare the different types of goods and services.

  This problem can be tackled using a market basket approach: What are the monetary costs of each of a set of goods and services? The question is which goods and services are worth keeping track of, and how to combine them into a single unifying measure.

  The market basked can probably not be reduced to a single number without losing a lot of information in the process: How do you quantify something like the availability of public museums and compare it to the cost of a loaf of bread?

• Fairness of distribution

  This measure how fairly all the goods and services produced in a society are distributed to people.

  The problem here is highly subjective: People disagree wildly on what constitutes a 'fair' distribution and this disagreement is based on differences in ethics.

  For example, the widely used GINI index is optimal only when every person receives the exact same amount of money. This completely ignores what each person does, or how much each person actually needs.

Of the common measure listed above, the unemployment rate cares only about fairness of distribution and ignores total production, while the GDP does the opposite.