How economies work · 5 of 9
GDP, inflation, unemployment
Say what each of the three measures and what each one leaves out.
Between October 2009 and October 2015 the American unemployment rate fell from 10.0% to 5.0%. Over the same six years the share of adults in the labor force fell from about 65% to about 62.5%. Both numbers improved the story politicians told and only one of them was about people finding work. Learning to read the second number is most of what this lesson is for.
The three numbers and what each one is
GDP is the market value of everything produced in a country in a period. It counts transactions. is the change in the price of a fixed basket of goods. It counts prices. Unemployment is the share of people who want work and cannot find it. It counts a specific and narrow definition of wanting.
Each is a genuine attempt to compress an economy into one figure, and each fails in a specific way that is written into its definition rather than hidden in the methodology. The failures are the useful part, because they tell you which second number to go and find.
Unemployment: watch the denominator
To be counted as unemployed you must have no job and have looked for one in the last four weeks. Stop looking and you leave the labor force, which is the denominator, and you also leave the numerator. The rate falls.
Below is a population of 252 million adults in five boxes. Try to improve the headline rate.
Two of the six moves improve the headline without anybody getting a job, and one of them makes it worse while nothing bad happens at all. That is not a defect in the statistic. It is what the statistic means, and it is why every serious labor market discussion cites the participation rate alongside the rate itself.
The 2009 and 2015 figures are Bureau of Labor Statistics series LNS14000000 and LNS11300000, both free on the BLS site and on FRED. U-6 is published monthly in BLS Table A-15, where the definitions above are also set out. At the April 2010 peak, U-3 was 9.9% and U-6 was 17.1%.
Inflation: the basket is not your basket
A price index prices a fixed basket, weighted by what an average household spends. Shelter is roughly a third of the American consumer price index, and about three quarters of that shelter weight is owners' equivalent rent, which is an estimate of what homeowners would pay to rent their own houses. It is a real cost and nobody writes a check for it.
This matters because the average household is a construct. A 23-year-old renting in a city and a retired couple who own their home outright face the same index and very different prices. When your personal experience disagrees with the figure, the usual explanation is that your basket is not the published one.
ConventionCore inflation strips out food and energy, which sounds like removing the things that matter. The reason is that those two prices are volatile and mean-reverting, so including them makes it harder to see the trend that monetary policy can actually act on. A central bank watching a headline number driven by one bad harvest would respond to weather. Read core for the direction and headline for what you are paying.
GDP: it counts spending, so it counts the wrong things too
GDP counts market transactions. Anything valuable that happens without money changing hands is invisible to it, and anything wasteful that involves money is visible.
Unpaid work is the biggest hole. A parent caring for their own child produces nothing measurable; the same parent taking a job and paying somebody else to do the same care produces two increases in GDP. Environmental damage is the other direction: the output shows up and the depletion does not. And rebuilding after a hurricane adds to GDP while the hurricane subtracted nothing, which is why the figure sometimes rises after a disaster.
None of this makes GDP useless. It makes it a measure of activity rather than a measure of wellbeing, and the trouble starts when people use the second word while quoting the first number.
What to do with this
Pair every headline with its companion. Unemployment goes with participation or the share of adults working. Inflation goes with the components, because knowing that shelter drove it tells you more than the total does. GDP goes with per head, and with median income if the question is about living standards.
And when a number moves a lot in one month, check whether the definition moved with it. In April 2020 the unemployment rate hit 14.7%, and the BLS said in the same release that misclassification in the survey meant the true figure was several points higher. The agencies publishing these numbers are usually the first to tell you where they break. It is the people quoting them who leave that part out.
Test yourself
01Unemployment falls from 6% to 5% and the share of adults with a job also falls. What happened?
People left the labor force. The rate improved because the denominator shrank faster than the numerator, and the share of adults actually working is the figure that shows it.
Whether that is bad news depends entirely on who left. Retirements from an aging population produce the same two numbers as a wave of people giving up on finding work. The rate cannot distinguish them, so you have to go and look at the age breakdown, which is published.
02Your rent went up 8% and the inflation figure came in at 3%. Is the figure lying?
No, it is answering a different question from the one you are asking. The index prices a fixed basket for an average household, and shelter is about a third of it. If you are a renter who just signed a new lease, shelter is closer to half of your spending and you are paying today's price rather than the average of everyone's existing leases.
Your personal inflation rate depends on your basket. Somebody who owns their home outright, drives little, and eats at home experienced a different number from yours in the same month, and both numbers are real.
03Name two things that raise GDP which nobody would describe as an improvement.
Rebuilding after a disaster, because the destruction is not subtracted and the reconstruction is added. Also: a longer commute burning more fuel, higher spending on prisons or on treating an illness, and any unpaid work that becomes paid, such as a parent going back to work and hiring childcare. The last one is the clearest case. The same care happens, and GDP goes up because money changed hands.
A tutor that knows this lesson. It asks before it explains, and it will not tell you what to do with your own money.
Educational material, not investment or policy advice. Figures are cited where they come from a filing or a statistical series, and labelled as illustrative where they do not.
