Unemployment

This is Unit Five of Principles of Macroeconomics: unemployment. Unit four asked what makes output grow; this unit asks about the people not producing it. Carry one idea through. Classification decides everything here: the definition decides who is counted, and the cause decides which remedy works. The capstone comes early, at the entry explaining why anyone is involuntarily unemployed at all. By the end you will compute both rates with their different denominators, classify a person into one of three categories, say why wages resist falling, tell cyclical unemployment from frictional and structural, and match a remedy to a diagnosis.

Five entries on who counts as unemployed, and the two rates reported each month.

Why the measure matters.

MAC 0401 Why Unemployment Matters

Unemployment costs the person income, savings and often health, and it costs the economy the output those workers could have produced, which is the opportunity cost of leaving them idle. It is measured with more care than almost any other statistic, because the definition decides who is counted.

How everyone is sorted.

MAC 0402 The Three Categories

Everyone sixteen and over falls into one of three groups: employed, unemployed, or out of the labor force, and the three are exclusive and exhaustive. Employed means working for pay; unemployed means jobless, available, and actively looking. The base population excludes active-duty military and people in institutions.

Which two are added together.

MAC 0403 The Labor Force

The labor force is the employed plus the unemployed. People out of it are in neither group and not in the denominator of the unemployment rate: students, retired people, full-time caregivers. Someone can leave unemployment by finding work or by giving up looking, and only the first is good news.

The headline rate.

MAC 0404 The Unemployment Rate

The unemployment rate is the unemployed as a percentage of the labor force, not of the whole population: divide the unemployed by the employed plus the unemployed. Dividing by the adult population instead is the most common arithmetic error here, and because that population is larger it always understates the rate.

The rate reported beside it.

MAC 0405 The Participation Rate

The participation rate is the labor force as a percentage of the civilian noninstitutional population, and it answers a different question: one asks how many of those looking cannot find work, the other how many take part at all. Both can fall at once, if people stop looking.

Three entries working those definitions, one calculation and one person at a time.

Computing the two rates.

MAC 0406 Computing the Rates

Work in a fixed order. Add the employed and the unemployed to get the labor force. Divide the unemployed by that labor force for the unemployment rate. Divide the labor force by the civilian noninstitutional population for the participation rate. Write each denominator down before dividing. With two hundred and fifty million people sixteen and over, a hundred and fifty million employed and ten million unemployed, the labor force is a hundred and sixty million, the unemployment rate is six point two five percent, and the participation rate is sixty-four percent. Ninety million appear in neither numerator.

Sorting one person.

MAC 0407 Classifying a Person

Ask two questions in order. Did they do any paid work in the reference week, even an hour, or fifteen unpaid hours in a family business, or were they away from a job they hold? If so, employed. If not, did they make a specific active effort to find work in the four weeks ending with that week, and can they start? Only if both hold are they unemployed. Anyone else is out of the labor force. Wanting a job is not enough, and neither is needing one.

The people the categories miss.

MAC 0408 Discouraged and Marginally Attached Workers

Discouraged workers want a job and stopped looking because they believe none is available. They are a subset of the marginally attached, who looked within the past year but not the past four weeks, and neither counts as unemployed. The underemployed are the mirror problem, working part time and counted as employed.

Four entries on where the numbers come from, and what they look like over time.

Six measures, not one.

MAC 0409 The Alternative Measures

Because one definition cannot capture everything, the statistics agency publishes six measures, U-one through U-six, and U-three is the official rate. U-one counts only the long-term unemployed; U-six adds the marginally attached and involuntary part-timers. U-six is always the largest, so a rate quoted without its measure tells you little.

Where they come from.

MAC 0410 Two Surveys

Two surveys produce the monthly figures: a household survey gives the unemployment rate, and a survey of employers gives the change in the number of jobs. The employer survey misses the self-employed, so the two can disagree, and only the household survey yields a rate.

What the series looks like.

MAC 0411 Patterns in the Data

The rate rises in recessions and falls in expansions, returning to roughly four to six percent. It never reaches zero, and despite a far larger labor force there is no long-term trend up or down. Unemployment lags the cycle, so it keeps rising after output turns up, which is why recoveries feel late.

Comparing one country with another.

MAC 0412 Comparing Countries

Comparing unemployment across countries needs care, because definitions, surveys and labor markets differ, so the same word can measure different things. A country where people leave the labor force rather than search will report a lower rate. Comparisons with low-income countries are hardest, because much work is short-term, subsistence or barter.

Two entries on the unemployment that comes with recessions, and why it exists at all.

Unemployment from the cycle.

MAC 0413 Cyclical Unemployment

Cyclical unemployment is the part that moves with the business cycle: in a recession the demand for labor shifts left. If the wage cannot fall, quantity supplied exceeds quantity demanded, and that gap is the unemployment. Most unemployment is involuntary, which is the puzzle, because with flexible wages the market would clear.

Why the wage does not fall.

MAC 0414 Why Wages Are Sticky Downward

Here the unit's governing idea is at work. If wages fell whenever labor demand fell, the market would clear and involuntary unemployment would barely exist. It does exist, so wages resist falling, and economists offer five explanations that compete rather than exclude one another. Implicit contracts: employers hold wages up in bad times and back in good. Efficiency wages: paying above the market raises effort and retention, so cutting can cost more than it saves. Adverse selection: an across-the-board cut drives out the best workers first. Insiders and outsiders: those employed keep the firm running. And relative wage coordination: nobody wants to be the one whose pay falls while others hold. A firm whose sales fall by a fifth can cut every wage or lay off a fifth of its workers; cutting sends its best engineers away, so it lays off. Three cautions. These are competing accounts, not an agreed list. Laws and unions cannot carry it alone: stickiness appears well above the minimum wage and beyond union contracts. And because the wage does not clear the market, a fall in demand shows up as lost jobs rather than lower pay, which is why the short-run aggregate supply curve in unit eight slopes upward.

Three entries on the unemployment that persists when the cycle is neutral.

Unemployment that never goes away.

MAC 0415 Frictional Unemployment

Frictional unemployment is the time it takes workers and employers to find each other, and it exists even when the number of jobs matches the number of workers. It is not a failure: a good match takes time, and rushing everyone into the first offer would waste skills.

Unemployment from a mismatch.

MAC 0416 Structural Unemployment

Structural unemployment comes from a mismatch: the skills employers want are not the skills workers have, or the jobs are not where the workers are. A technology can make one set of skills obsolete while creating demand for another. Retraining is the usual remedy, and it works slowly.

Telling the three apart.

MAC 0417 Which Type of Unemployment?

Three types, one question each. Is the person between jobs in a working economy, searching and likely to be matched soon? Frictional. Do the skills or the location fail to match what employers want, so searching longer would not solve it? Structural. Did the job disappear because the economy contracted, and would it return if demand recovered? Cyclical. The test separating the last two is whether reviving demand would bring the job back. A welder laid off when a shipyard closed permanently is structural; a cashier laid off in a recession whose employer plans to rehire is cyclical.

Five entries on the natural rate, potential output, and what moves them.

What is left when the cycle is neutral.

MAC 0418 The Natural Rate

The natural rate of unemployment is what remains when the economy is neither booming nor in recession: frictional plus structural, with no cyclical unemployment. It is never zero, because search and mismatch never stop. Natural does not mean good or fixed: it arises from current arrangements and institutions, and it changes over time.

What it means for output.

MAC 0419 Full Employment and Potential GDP

At the natural rate the economy is at full employment and real GDP equals potential GDP. Above the natural rate output is below potential; below it, output is above potential. Full employment therefore includes people who are unemployed, and output above potential can only last a while.

Reading one rate against the other.

MAC 0420 Reading the Rate Against the Natural Rate

Comparing actual with natural tells you where output sits relative to potential, and the relation runs in reverse. Actual above natural means cyclical unemployment and real GDP below potential. Actual below natural means the economy is working beyond its sustainable level, and equal means output is at potential. Subtract natural from actual: a positive answer is a recessionary gap. With a natural rate of four and a half, an actual rate of seven leaves two and a half points of cyclical unemployment. The natural rate is an estimate, so half a point sits inside the margin of error.

What makes it move.

MAC 0421 What Moves the Natural Rate

The natural rate moves. It falls when job search gets easier or the workforce ages, and rises when mismatch grows or wages are left too high. Online job listings lowered search costs. Policy moves it too, on both sides: how long benefits last, and how hard it is to hire and dismiss.

The main policy on that side.

MAC 0422 Unemployment Insurance

Unemployment insurance pays a share of prior earnings for a limited period to workers who lost jobs through no fault of their own, cushioning the loss and keeping spending up in a downturn. Receiving benefits is not what makes someone count as unemployed, and research finds longer benefits lengthen search.

Two entries on reading a published figure, and matching a remedy to a diagnosis.

Reading a published figure.

MAC 0423 Reading an Unemployment Statistic

Before accepting an unemployment figure, ask four questions. Which measure is it, since the six differ by several points? What is the denominator, the labor force rather than the population? What happened to participation over the same period, since a falling rate can mean people stopped looking? And what is the reference period, since the figure describes one week of one month and is revised later? A rate falling from eight percent to seven while participation falls two points is not the same news as the same fall with participation rising.

Matching a remedy to a diagnosis.

MAC 0424 Which Remedy Fits?

The remedy follows the diagnosis, and mismatching them wastes the policy. Cyclical unemployment comes from too little total demand, so the jobs come back when spending does. Frictional and structural unemployment persist at full employment, so demand policy cannot touch them; they call for better job information, help with search and relocation, and retraining. A region that loses its mines permanently stays high through a national boom, and a stimulus will not reopen the mines. The hard case is the mixed one, which is most real recessions, and the shares are not observable at the time.

Look at what you can now do. You can compute both rates, dividing the unemployed by the labor force and the labor force by the adult population. You can classify a person by whether they worked, and if not whether they searched within four weeks. You can say why wages resist falling, and why a fall in demand costs jobs rather than pay. You can tell cyclical unemployment from frictional and structural, by asking whether demand would bring the job back. And you can read a rate against the natural rate, and match a remedy to a diagnosis. Unit six turns to the price level.