Why Unemployment Changes: Job Flows, Frictions, and Cyclical vs Structural Forces

1) The mechanics: job-finding and job-separation rates

Unemployment changes because people are constantly moving between three states: employed, unemployed, and (sometimes) out of the labor force. In this chapter, focus on the two flows that directly move the unemployment rate: separations (people losing or leaving jobs and entering unemployment) and job finding (unemployed people getting jobs and exiting unemployment).

Think of the unemployment rate as the water level in a bathtub. Separations are the faucet filling the tub; job finding is the drain. The unemployment rate rises when inflows exceed outflows, and falls when outflows exceed inflows.

A simple flow equation (intuition, not heavy math)

Let U be the number of unemployed people and L the labor force. Each month:

• Some employed workers separate into unemployment at rate s (job-separation rate).
• Some unemployed workers find jobs at rate f (job-finding rate).

A compact way to express the change in unemployment is:

ΔU ≈ (separations into unemployment) − (hires from unemployment)

In a very simplified two-state world (employed vs unemployed), the unemployment rate tends to settle near a “steady” level where inflows equal outflows. That steady unemployment rate is approximately:

Continue in our app.

• Listen to the audio with the screen off.
• Earn a certificate upon completion.
• Over 5000 courses for you to explore!

Or continue reading below...

Download the app

u* ≈ s / (s + f)

This formula is powerful because it shows two different ways unemployment can rise:

• Separations increase (higher s): more people are pushed into unemployment.
• Job finding falls (lower f): people stay unemployed longer.

Practical step-by-step: diagnosing a rise in unemployment

1. Ask: are layoffs/quits changing? If layoffs spike, s likely rose. If quits fall sharply, it can signal workers feel less confident (and may also affect matching dynamics).
2. Ask: are hires slowing? If firms post fewer openings or take longer to fill roles, f likely fell.
3. Look for duration clues. If unemployment duration rises (people unemployed longer), that often points to a lower f.
4. Connect to the story. A sudden shock (e.g., demand drop) often hits f quickly; a sectoral shift may keep f low for specific groups for longer.

Mini example with numbers

Suppose initially s = 0.02 (2% of employed separate into unemployment per month) and f = 0.40 (40% of unemployed find jobs per month). Then:

u* ≈ 0.02 / (0.02 + 0.40) ≈ 4.8%

If a downturn reduces hiring so f falls to 0.25, then:

u* ≈ 0.02 / (0.02 + 0.25) ≈ 7.4%

Unemployment rises even if separations do not change, simply because the drain clogged.

2) Frictional unemployment: search and matching

Frictional unemployment is unemployment that comes from the time it takes to match workers and jobs, even in a healthy economy. People search for better fits; firms search for suitable candidates. Because information is imperfect and the process takes time, some unemployment is normal.

Why matching takes time

• Information frictions: workers do not instantly know every job opening; firms do not instantly know every qualified applicant.
• Heterogeneity: jobs differ (hours, location, tasks); workers differ (skills, preferences).
• Screening and hiring processes: interviews, background checks, onboarding.
• Timing: vacancies open at different times than workers become available.

Matching function intuition

Economists often summarize the hiring process with a “matching” relationship: hires depend on both the number of unemployed workers (U) and the number of vacancies (V). When there are many vacancies and few unemployed, job finding is easier; when there are many unemployed and few vacancies, job finding is harder.

Two useful rates:

• Job-finding rate (f): roughly, the probability an unemployed worker finds a job in a period. It tends to rise when V is high relative to U.
• Vacancy-filling rate: roughly, the probability a vacancy gets filled. It tends to rise when U is high relative to V.

Practical step-by-step: reducing frictional unemployment (micro-to-macro)

1. Improve information: better job matching platforms, clearer job descriptions, transparent pay ranges.
2. Speed up processes: shorter hiring cycles, standardized assessments, faster credential verification.
3. Lower mobility barriers: childcare availability, transportation access, predictable scheduling.
4. Support search: effective unemployment insurance design can help people search without forcing immediate poor matches, but overly long or poorly designed support can reduce search intensity; the details matter.

Frictional unemployment is not “bad” in itself: some search leads to better matches, higher productivity, and higher wages over time. The key is whether frictions are unusually high or whether the economy is simply reallocating workers efficiently.

3) Structural unemployment: skills, geography, and sectoral shifts

Structural unemployment happens when there is a persistent mismatch between what employers need and what workers can offer, or where jobs are located versus where workers live. Unlike frictional unemployment (short-term search), structural unemployment can last longer because it requires adjustment: retraining, relocation, or sectoral transformation.

Common sources of structural mismatch

• Skills mismatch: job growth in fields requiring skills that unemployed workers do not have (e.g., demand for data analysts while many job seekers have experience in declining occupations).
• Geographic mismatch: vacancies concentrated in certain regions while unemployed workers are elsewhere, with barriers like housing costs, family ties, or licensing rules.
• Sectoral shifts: long-lasting changes in demand across industries (e.g., a permanent decline in one sector and expansion in another).
• Institutional barriers: occupational licensing, non-compete clauses, credential requirements that limit mobility.

How structural forces show up in the flow mechanics

Structural unemployment often appears as a persistently low job-finding rate for certain groups or regions, even when the overall economy is doing reasonably well. It can also show up as vacancies staying high while unemployment remains elevated, suggesting that jobs exist but are not being filled by available workers.

Practical step-by-step: telling structural from frictional in real life

1. Check concentration: Is unemployment high in specific industries, occupations, or regions?
2. Compare vacancies by sector: Are there many openings in growing sectors while unemployment is concentrated in shrinking ones?
3. Look at wage signals: If wages rise strongly in high-vacancy sectors, that suggests scarcity of suitable workers (a mismatch).
4. Assess persistence: If the problem lasts through the recovery phase, structural factors are more likely.

Policy tools often discussed for structural unemployment

• Training and reskilling: targeted programs linked to employer needs, apprenticeships, short credentials.
• Mobility support: relocation assistance, housing supply policies, portability of benefits.
• Credential and licensing reform: reduce unnecessary barriers while maintaining safety and quality.
• Place-based policies: incentives for firms to locate in high-unemployment areas (with tradeoffs and mixed evidence depending on design).

4) Cyclical unemployment: aggregate demand and the business cycle

Cyclical unemployment rises in downturns and falls in expansions because overall spending in the economy changes. When demand for goods and services weakens, firms sell less, so they need fewer workers. When demand strengthens, firms expand and hire.

How cyclical forces move the flow rates

Cyclical unemployment can operate through both s and f, but often the most important channel is a drop in hiring:

• Job-finding rate falls (f ↓): firms post fewer vacancies, slow hiring, or become more selective.
• Separations may rise (s ↑): layoffs increase, temporary contracts are not renewed.

In many downturns, unemployment rises sharply because the job-finding rate collapses: fewer opportunities are available, so unemployment duration increases.

Practical step-by-step: mapping a demand shock to unemployment

1. Spending drops: households and firms cut purchases and investment.
2. Revenue falls: businesses see lower sales.
3. Vacancies fall first: firms pause expansion plans and reduce job postings.
4. Hiring slows: job-finding rate declines; unemployed workers take longer to find jobs.
5. Layoffs may follow: if weakness persists, separations rise.
6. Unemployment rises: both higher inflows and lower outflows raise the unemployment stock.

This sequence helps explain why recoveries can feel “jobless” at first: firms may stop layoffs and stabilize, but hiring can remain weak for a while, keeping f low and unemployment elevated.

5) Wages, vacancies, and bargaining power: the vacancy–unemployment relationship

Labor markets do not adjust only through quantities (employment and unemployment). They also adjust through prices (wages) and through vacancies (how many jobs firms are trying to fill). A key empirical pattern is a relationship between vacancies and unemployment: when vacancies are high, unemployment tends to be low; when vacancies are low, unemployment tends to be high.

The vacancy–unemployment relationship (intuition)

Imagine plotting:

• Unemployment rate on the horizontal axis
• Vacancy rate (open jobs as a share of the labor force or employment) on the vertical axis

Typically, the plot slopes downward: tight labor markets have many vacancies and few unemployed; slack labor markets have few vacancies and many unemployed. Movements along this relationship often reflect cyclical changes (demand-driven). Shifts of the whole relationship can reflect changes in matching efficiency or structural mismatch.

How bargaining power fits in

• When unemployment is low and vacancies are high, workers usually have more outside options. Bargaining power shifts toward workers, and wage growth tends to strengthen.
• When unemployment is high and vacancies are low, firms have more applicants per opening. Bargaining power shifts toward firms, and wage growth tends to weaken.

Wages, however, do not always adjust instantly. Contracts, norms, minimum wages, and concerns about morale can make wages “sticky,” so the economy may adjust more through quantities (employment/unemployment) than through wages in the short run.

Reading the signals: a practical checklist

What this suggests for policy debates

Because unemployment can rise for different reasons, policy arguments often hinge on diagnosis:

• If unemployment is cyclical: policies that support overall spending (monetary or fiscal tools) can raise vacancies and hiring, increasing f and reducing unemployment.
• If unemployment is structural: boosting demand alone may raise wages in scarce-skill areas without quickly reducing unemployment for mismatched workers; targeted training, mobility, and barrier reduction may be more effective.
• If unemployment is frictional: improving matching efficiency (information, placement services, reducing hiring delays) can raise hires for a given level of vacancies.

In practice, economies often face a mix: a downturn can be mostly cyclical at first, then leave behind structural issues (skills erosion, sectoral change) that keep job finding lower for some workers. The flow approach (separations and job finding) plus vacancy and wage signals helps separate these forces.