Unemployment

GDP is a versatile measure, representing spending, income and production, but it is only one of our headline economic numbers. Another dimension is captured in thinking about aggregate labor markets. How many people are working or not working could be measures, but these will often rise over time purely due to population growth.

What we would like to build up to is something closer to a real measure, the percentage of the available population that is working or not working. As a convention, we focus on the percentage of the available population that is not working.

To end up with this we need to develop specific definitions with clearly outlined adjustments to total population. These are critical for us to calculate our measures and interpret them correctly. Data for the United States is collected from two different surveys, which will be briefly described at the end of the subchapter.

Who “counts” for labor statistics?

As a starting point we begin with the total population. Not everyone in the population is able to work. The simplest cut is due to age. Because of labor laws, we disinclude everyone below the age of 16. We also have to take away people who practically cannot work. That includes people in active-duty military and people in institutions. These include prisons and nursing homes. Everyone who remains is the civilian noninstitutional population.

We need to divide the civilian non-institutional population into some key groups that will allow us to achieve our headline numbers.

Definitions

Labor force

The pool of available and willing workers. Sum of the employed and unemployed.

Employed

Everyone who currently has a job, as long as they worked at least one hour during the survey week. This is regardless of the type of job, or if it is full time or part time.

Unemployed

Those currently without a job who have looked for a job at some point in the last four weeks.

Out of the labor force

Those currently without a job who are not actively looking for one for any reason. Examples include students, people without jobs for family reasons and discouraged workers.

In one sense this classification is quite rigid. Any qualifying activity overrules any non-qualifying activity. For example, part time employment as a student places that individual in the labor force. And if you have a job but are looking for a better job or already have a second job, you simply count as employed.

However, that also means that who is included as employed is quite broad. Working one hour part-time counts the same as working a full 40 hour week. Being employed says very little about personal circumstances and success.

The connections between these definitions can be summarized by the following equations.

$$\begin{array}{rl}\text{Civilian non-}& \text{institutional population}\\ & =\text{Labor force + Out of the labor force}\\ \text{Labor force}& =\text{Employed + Unemployed}\end{array}$$

Headline numbers

Now we present our two headline numbers regarding aggregate labor markets.

$$\begin{array}{rl}\text{Labor force particip}& \text{ation rate}\\ & =\frac{\text{Labor force}}{\text{Civilian non-institutional population}}\\ \text{Unemployment rate }(U_3)& =\frac{\text{Unemployed}}{\text{Labor force}}\end{array}$$

For example, take a country with a civilian non-institutional population of 200 million, where 114 million people are employed and 6 million are unemployed. The labor force is $114+6=120$ million, so the labor force participation rate is $\frac{120}{200}=0.60$ or $60\%$, and the unemployment rate is $\frac{6}{120}=0.05$ or $5\%$.

The labor force participation rate gives us the percentage of the available population that is willing to work at present. Due to structural changes the rate rose to a high of just over 67 percent in 2000. Since the COVID-19 recession in 2020 the labor force participation rate for the entire United States population has been mostly between 62 and 63 percent.

Our second value is the unemployment rate, the fraction of available and willing workers who do not currently have a job. It is important to remember that it is not the fraction of the population who do not have a job. The standard measure is referred to as $U_3$. It has been collected for many years and thus has something valuable to say about labor market performance and the aggregate employment situation.

The graph below shows the unemployment rate for the United States this century, ending in January of 2025.

However, many people are “missing” from this definition and because of that other measures are available.

Definitions

Marginally attached workers

Workers out of the labor force, but who have looked for a job in the past year.

Discouraged workers

Those currently without a job who have given up looking for a job only because they believe there are no suitable opportunities available. Subset of those out of the labor force. A subset of marginally attached workers.

Some additional unemployment measures add these types of workers to the labor force and unemployed. There are also adjustments for those with part-time jobs who wish they had full time jobs. You are not responsible for knowing any details about these measures, it is simply good to be aware that deeper information about the employment situation is available beyond the standard unemployment rate.

Types of unemployment

Returning to the standard unemployment measure, next we separate unemployed workers depending on the reason they do not have a job.

Definitions

Frictional unemployment

Unemployment due to temporary factors, such as moving, switching jobs or graduating school.

Structural unemployment

Unemployment resulting from fundamental shifts in the economy, such as skills mismatch or not being able to move where jobs are located.

Cyclical unemployment

Unemployment resulting from changes in demand for workers due to the business cycle

Frictional unemployment can be summarized as not currently having a job, but expecting to find one soon. Frictional unemployment would only be zero in a command economy with guaranteed employment. In market-based economies, it is always positive. Anything that speeds up the job search process can make it go down. A past example of this is the introduction of internet job boards at the turn of the century and a current one could be the development of AI job matching tools.

Similarly, structural unemployment is not having a job and having skills or being in a field where it is difficult to find a job where you currently are. So those who are structurally unemployed you may need additional training, education or to move to find a job again. This can rise and fall over decades, due to changes in technology, education, and job training.

Cyclical unemployment stems from business cycles. During recessions labor demand is weaker and unemployment rates rise. During expansions, labor demand is often stronger and unemployment rates fall. A notable exception to these patterns is the recovery following the recession of 2000, when GDP growth returned but unemployment continued to rise for a few years after.

These divisions of unemployment enable us to provide one more definition that will have more importance later on.

$$\begin{array}{r}\text{Natural rate of unemployment}=\frac{\text{Structural + Frictional}}{\text{Labor Force}}\end{array}$$

Put another way, the natural rate of unemployment is the non-cyclical portion of unemployment. For the United States in recent years has often been thought to be 4-5%. When the unemployment rate is equal to the natural rate of unemployment, economists sometimes say that the economy is at full employment.

Measurement

This is not required material for the AP Exam, simply some real life context to how the data is collected.

In the United States, the values we have just defined are measured using two distinct surveys. Each is conducted on a monthly basis by the Bureau of Labor Statistics, part of the Department of Labor.

The Current Employment Statistics survey is a survey of businesses. This is used to track changes in total nonfarm payroll employment.

The Current Population Survey is a monthly survey of households, this is what is used to calculate the unemployment rate. As a part of this demographic data is also collected, allowing for measurement of unemployment rates across various groups.