← all systems · see it in The System
Unemployment
Unemployment is the state of people of working age who are without paid work, are looking for it and are available to take it; in aggregate it is the pool of such people and its measure, the unemployment rate (the unemployed as a share of the labour force: employed plus unemployed). Statisticians count it by labour force surveys on the ILO definition (no work in the reference week, active search in the last four weeks, available to start), with broader measures (U1-U6, LU1-LU4) for the discouraged, the marginally attached and the underemployed. Economists sort it by cause into frictional (search between jobs), structural (mismatch of skills, places and jobs, including technological), cyclical or Keynesian (too little aggregate demand) and classical (wages above the clearing level), and read it through its flows (job loss in, job finding out) and its relations with output (Okun's law), inflation (the Phillips curve, the natural rate or NAIRU) and vacancies (the Beveridge curve). Sociologists and psychologists study what it does to people and places: lost income and the latent functions of work (time structure, contact, purpose, status, activity), worse mental health, scarred earnings, and political effects. Societies answer it with relief, insurance, demand management and public works. Walked as the state and its measure, both booked, and the economy it reads.
Properties published scienceour proposal
Sourced fields, and our own readings.
| born at | contested (C-2861): Tudor England's vagrancy and poor laws (1530s-1601) lumped and then separated the able-bodied jobless; recognised as a category with industrialisation; counted systematically in the 20th century |
| born precision | n/a |
| boundary | Inside: people of working age who have no paid work in the reference period, are seeking it and are available (the unemployed), the pool they form with its flows in (job loss, entry, re-entry) and out (job finding, leaving the labour force), the measure of that pool (the rate and its wider variants), and the mechanisms that set it (search, mismatch, deficient demand, wages above clearing). On the boundary: the potential labour force (discouraged and marginally attached people) and the time-related underemployed, counted by the wider measures but not by the headline rate. Outside: the employed (one hour of paid work in the week makes a person employed), people outside the labour force (students, retirees, homemakers not seeking work), and the economy, firms and state that act on the pool. |
| nested level | Built from people (job seekers) in the labour force; a state of the labour market (its unused labour) and a measure the economy reads itself by; runs on people's search, firms' hiring and firing, and the statistics agencies' counts. |
| scale | social |
Measures (23) published science
| world unemployment, 2018 | 172 million (5% of the reported global workforce) people |
| monthly job finding probability (US, 1948-2007) | 45 (post-war average) percent per month |
| monthly employment exit probability (US, 1948-2007) | 3.4 percentage points per month |
| share of rate fluctuations from job finding (US since 1948) | three-quarters (exit probability one-quarter) share |
| steady-state rate vs next month's actual rate | 0.99 correlation (quarterly-averaged) |
| Okun coefficient | -0.15 (Japan), -0.45 (United States), -0.85 (Spain) change in unemployment rate per 1% change in output |
| callback rate by unemployment duration | about 7% at one month to 4% at eight months (about 45% lower), flat after percent of applications |
| present value earnings loss after displacement | 1.4 years (unemployment below 6%) to 2.8 years (above 8%) of pre-displacement earnings years of earnings (men, US, 5% discount over 20 years) |
| graduating in a recession: initial earnings loss and fade | about 9 percent at first, halved within 5 years, gone by 10 (rise of unemployment by 5 points) percent of earnings; years |
| mental health: unemployed vs employed | SMD 0.19 (95% CI 0.08-0.30); re-employment SMD -0.27 (between-group) standardised mean difference |
| latent function deficit, employed vs unemployed (status) | d = 0.94 (status); 0.59 time structure; 0.51 activity; 0.45 collective purpose; 0.39 social contact Cohen's d |
| suicide risk associated with unemployment (63 countries, 2000-11) | relative risk raised about 20-30%; 41,148 suicides associated in 2007, 46,131 in 2009 relative risk; deaths per year |
| UI benefit elasticity of duration (US) | a 10% increase in benefits raises durations 4-8%; 60% of the effect is liquidity percent |
| robots and employment (US commuting zones, 1990-2007) | -0.18 to -0.34 points of employment-to-population per robot per thousand workers; wages -0.25 to -0.5% percentage points |
| US unemployment rate, 1933 | 25% percent of labour force |
| Germany and US unemployment rate, 1932 | about 25% percent |
| UK average unemployment, 1950s-1960s | 1.6% percent |
| UK unemployment, 1945-1970 peak and 1986 | peak 3.3 percent (1945-70); over 12 percent (1986) |
| US unemployment rate and U-6, April 2010 | 9.9% (U-3); 17.1% (U-6) percent |
| US household survey sample | 60,000 households households per month |
| ILO search reference period | last four weeks or one month |
| US unemployment insurance replacement | about one third of previous income; under 40% of covered apply and receive share of income |
| inactive workers in ant colonies (cross-domain) | approximately 50% of workers on average share of workers |
Open contentions (10) open contention
Where sources or readings disagree. Both stay on the map.
- C-1072 Do economies return to full employment on their own?
booked: The map doesn't rule. It records the Keynesian reading that they don't, alongside the Austrian, monetarist and classical views.
the other reading: demand-side remedies work (Keynesian); only supply-side and the natural rate (monetarist, supply-side) - C-2028 Does raising the minimum wage cost jobs?
booked: The map books the reanalysis finding of a limited effect and keeps the textbook reading beside it.
the other reading: minimum wages cause unemployment; small or no employment effects - C-2608 Does lower unemployment mean higher inflation in the long run?
booked: The map keeps the competing readings; none is ruled.
the other reading: stable exploitable trade-off; short-run only; natural rate in the long run - C-2619 What causes unemployment?
booked: The map keeps the competing readings; none is ruled.
the other reading: the market clears; unemployment frictional and voluntary; involuntary unemployment from deficient demand or rigid wages - C-2823 Why do long-unemployed people find work less easily?
booked: The map keeps the competing readings; none is ruled.
the other reading: employer screening; employer ranking; human capital depreciation (skills rust) - C-2856 Do unemployment benefits lengthen job searches through moral hazard or through cash?
booked: The map keeps the competing readings; none is ruled.
the other reading: moral hazard (reduced incentive to search); liquidity (constrained households can afford to search longer) - C-2858 Does unemployment cause ill health, or the other way round?
booked: The map keeps the competing readings; none is ruled.
the other reading: unemployment causes worse health; poor health causes unemployment (selection); recessions improve some health measures - C-2859 Does high unemployment raise unemployment for good (hysteresis)?
booked: The map keeps the competing readings; none is ruled.
the other reading: hysteresis: temporary shocks have permanent effects; natural-rate view: the rate returns to a supply-determined natural level; partial persistence: scars that fade - C-2860 Which measure of unemployment should be the headline?
booked: The map keeps the competing readings; none is ruled.
the other reading: the ILO headline rate; broader underutilization measures; employment-to-population ratios - C-2861 When did unemployment begin?
booked: The map keeps the competing readings; none is ruled.
the other reading: a modern category, born with industrial wage labour; joblessness among people able to work is older (Tudor vagrancy laws, 1535-1601 poor relief); only the name and the count are modern
Cross-scale patterns seen here (18) our proposal
Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Unemployment. Each links to every other system that shows it.
- nested ladder (structural)
The US alternative measures form a ladder, each a wider form of the last: U3 ⊂ U4 (+ discouraged) ⊂ U5 (+ marginally attached) ⊂ U6 (+ involuntary part-time); the ILO's LU1 ⊂ LU2/LU3 ⊂ LU4 splits at a half-rung (underemployment or potential labour force first). - same role, different builds (structural)
One job, carrying people of working age through a gap in paid work, is met by different builds: punishment and moving-on (Tudor 'sturdy beggars'), the workhouse (until 1948), state insurance (UK 1911, US 1935), union-run funds (the Ghent system), public works (WPA, CCC), a proposed job guarantee. Ac - templating (structural)
The ILO definition is a module swapped into national statistical systems and still working: the US CPS and Eurostat apply it, so counts interoperate; the 2013 and 2023 resolutions add the LU1-LU4 modules on the same base. - the whole sets the local reference (structural)
The market as a whole sets how a local reading is read: employers read an applicant's spell length as a signal, and read it less in weak markets; and national rates do not close into one frame (France's rate 60% higher than the US while more French prime-age men worked). - trade one property for another (structural)
Several trades recur: economic efficiency bought with frictional unemployment (searchers who wait for a fitting job); insurance (liquidity) bought with longer spells; low unemployment bought with inflation (Phillips); and, across domains, an ant colony's long-term persistence bought with short-term - who opens the door (structural)
Out of the pool the door is opened by the receiver: the employer posts the vacancy and decides the callback, reading the spell length; into insurance the state installs the registration door. A port, installed ahead of the crossing. - borrowed-then-held (partial)
Support for the jobless was first supplied from around the labour market (the Church helped the poor until the monasteries closed in the 1530s; families) and later held by its own institutions tied to the unemployed status (insurance from 1911, registration at employment agencies). The job moves bet - conservation law plus closure (partial)
The stock-flow balance (in steady state inflows equal outflows; LF = E + U) leaves the job finding and exit rates open; a second law, a model of matching or the measured hazard rates, closes it, after which the steady-state rate tracks the actual rate (correlation 0.99). - lightning (partial)
A single closure that reshapes a whole place: in Marienthal one textile factory shut and a village was changed (isolation, lost time structure, slower walking). A stored difference released at once is not shown; the discharge is the closure. - namespace completeness (the map predicts its gaps) (partial)
The ILO's three statuses are 'mutually exclusive and exhaustive' with a priority order; the scheme itself names the in-between cases it leaves (the potential labour force, time-related underemployment), which the later LU2-LU4 and U4-U6 measures fill. - runaway from an unstable balance (partial)
The demand spiral: job losses cut consumption, which cuts demand and causes further job losses; cyclical unemployment 'can boomerang back and exacerbate' its macro causes. The balance point is not shown in the sources read. - same shape, opposite sign (partial)
One transition run both ways: job loss withdraws the manifest and latent functions of work and raises symptoms (SMD 0.19); re-employment restores them and lowers symptoms (SMD -0.27). The process is the same change of employment status, its sign reversed. - smooth to sharp in finite time (partial)
In the Shapiro-Stiglitz efficiency-wage model the no-shirking wage rises smoothly as unemployment falls and goes to infinity at full employment, a finite point: a model singularity, not an observation. - symmetry breaking (name the cut) (partial)
Workers alike before a shock are cut into insiders (employed at the time) and outsiders; the cut sets later wages and so later unemployment ('a fundamental assymetry in the wage setting process'). The symmetric before is implied, not shown. - The carry (partial)
The wait face: unemployment is the settling time between jobs ('the time period between jobs'); a pool drained about 45 percent a month settles to its steady state month by month; the callback penalty settles in about eight months; graduates' losses take 8-10 years to fade. Overflow face: people who - The Payment to Entropy (partial)
At the hand-off between jobs part of a worker's value arrives unusable: displaced men lose 1.4-2.8 years of earnings in present value; skills rust in long spells; the economy's output lost (Okun). Form: human and match capital lost per hand-off, larger when the hand-off happens in a slump. - Umwelt: each system carves its own slice of one shared medium (partial)
The same jobless population is carved differently by each counter: sample surveys (the ILO's preferred), official estimates, social insurance claims, employment-office tallies; and within one survey the U1-U6 slices. Employers carve it by spell length. - causal ceiling (surface)
No observer can read the natural rate directly ('no one knows exactly what the NAIRU is'); the limit read in the source is practical (a wide margin of error), so this is near the dark twin.
Valence: what it trades, and with whom (12) published science
Each exchange with a neighbour, read from both sides: who gains, who pays, and what carries it.
- Human (Homo sapiens) via sys:labour-market · Unemployment gain
- Government via sys:government · Unemployment mixed
- Family (the human family as a system) via sys:human · Unemployment mixed
- Labour market via sys:labour-market · Unemployment mixed
- Firm (the economic unit) via sys:labour-market · Unemployment mixed
- Mental health via sys:human · Unemployment loss
- Trade unions, friendly societies, cooperatives via sys:labour-market · Unemployment loss
- Economy via sys:economy · Unemployment loss
- Labour force surveys and statistics via sys:labour-force-survey · Unemployment gain
- Inflation (economics) via sys:human-society · Unemployment mixed
- Industrial robots and automation via sys:technology · Unemployment loss
- Gross domestic product via sys:labour-market · Unemployment loss
Partners and parts published science
Built from / builds (3)
Sits inside / holds (2)
expressed by (10)
- Templating / Standardization
- Search/Foraging
- Optimal Stopping & Patch-Switching
- Social Learning / Public Information Use
- Resource Denial/Attrition
- Self-Monitoring
- Triage & Prioritized Allocation
- Deterrence
- Bridging Supply Gaps (Store & Dormancy)
- Scheduled Build-and-Dismantle (On-Demand Structure)
floor filled by (10)
- Labour market
- Human (Homo sapiens)
- Government
- Family (the human family as a system)
- Firm (the economic unit)
- Technology
- Mental health
- Trade unions, friendly societies, cooperatives
- Economy
- Labour force surveys and statistics
Kinds (11)
- Frictional unemployment
- Labour underutilization
- Structural unemployment
- Cyclical (Keynesian) unemployment
- Seasonal unemployment
- Classical (real-wage) unemployment
- Technological unemployment
- Hidden (covered) unemployment
- Long-term unemployment
- Youth unemployment
- Voluntary and involuntary unemployment
Notes published science
- Ant colonies keep about half their workers idle. In simulations and in real Myrmica colonies the idle ones step in only when the busy ones are worn out, and colonies with idle workers survive longer, trading short-term output for staying alive (src:hasegawa-2016-lazy-workers). Economists call unemployment's one benefit that people are 'available for hire' (src:wp:unemployment).
- In Marienthal, the Austrian village whose only factory shut in the Depression, the jobless did not riot: they withdrew, lost their sense of time, and even walked more slowly (src:franke-2024-unemployment-disorders).
- Send the same résumé with a longer gap and the phone rings less: callbacks fall about 45 percent over the first eight months out of work, then stop falling, and the penalty is stronger when jobs are plentiful, because employers read a long gap as a bad sign only when others are getting hired (src:kroft-2012-duration-dependence).
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
the boardroom (shared floors)
- L5: The pool and its rate: the stock of unemployed set by inflows and outflows (in steady state equal), the unemployment rate and its wider measures, the natural rate (NAIRU), and the relations through which the economy reads it: Okun's law (output), the Phillips curve (inflation), the Beveridge curve (vacancies); persistence and hysteresis
- L6: What it does to people and places: income lost and the latent functions of work (time structure, social contact, collective purpose, status, activity) missing; symptoms of depression, anxiety and distress up, down again on re-employment; earnings losses of 1.4-2.8 years after displacement; scarred cohorts; whole places (Marienthal); suicide risk; distrust of democracy, xenophobia, unrest
- L7: What is done about it: the unemployed search, retrain, move or give up; societies answer with relief (the Poor Relief Act 1601, the workhouse), insurance (UK 1911; US 1935), demand management (fiscal and monetary policy), public works (WPA, CCC), training and job-integration agreements; the full-employment goal and its limit at the natural rate
unemployment as economists and sociologists describe it: people of working age and the work on offer, through the count, the flows and their causes, to the rate, what it does to people and places, and the remedies (reader's view)
- L1: The working-age population and the work on offer: people who are employed, unemployed or outside the labour force, and the jobs firms and governments offer; only work for pay counts, so self-employed farmers 'cannot be unemployed'
- L2: The definition: not in employment (not one hour of paid work in the reference week), seeking in the last four weeks or month, available to start; three exhaustive statuses with employment first
- L3: Flows and spells: separations and entries in, job finding and exits from the labour force out; a monthly job finding probability around 45 percent and an employment exit probability around 3.4 points (US, 1948-2007); duration and its dependence
- L4: Kinds and causes: frictional (search), structural (mismatch, technological), cyclical (deficient demand), classical (wages above clearing), seasonal; voluntary and involuntary
- L5: The rate and how the economy reads it: the unemployment rate and its wider measures; the natural rate or NAIRU; Okun's law to output; the Phillips curve to inflation; the Beveridge curve to vacancies; persistence and hysteresis
- L6: What it does to people and places: lost income and the five latent functions of work (time structure, social contact, collective purpose, status, activity); higher symptoms of depression, anxiety and distress; scarred earnings; places like Marienthal; distrust, unrest and political upheaval
- L7: What people and societies do: the unemployed search, take other jobs or leave the labour force; societies give relief and insurance, manage demand (fiscal and monetary policy), run public works and training
way in: Demand for goods: recessions and the business cycle
- L1: Spending on goods and services across the economy; recessions, and shocks such as war, civil disorder and natural disasters
- L2: Firms need fewer workers when demand falls (labour demand is derived from the demand for goods); wages are sticky
- L3: Job finding falls and layoffs rise: since 1948 the job finding probability accounts for three-quarters of the rate's fluctuations, the exit probability for one-quarter
- L4: Cyclical (deficient-demand, Keynesian) unemployment: more unemployed than vacancies even if all were filled
way in: New entrants and re-entrants: people joining the labour force
- L1: People reaching working age, graduates, and re-entrants such as former homemakers
- L2: The choice to join the labour force: in a boom people give up studying and start a job hunt
- L3: Counted as unemployed until hired; those who enter in a recession start lower: graduates lose about 9 percent of earnings at first, fading over 8 to 10 years
- L4: Frictional unemployment of entrants; youth unemployment higher than general unemployment in many countries
way in: Search: job seekers and vacancies finding each other
- L1: Jobs and workers that differ in skills, pay, hours, place and taste
- L2: Active search: contacting employers, interviews, applications, agencies, answering advertisements, within the last four weeks or month
- L3: Time to a match: searchers take time for a better match; employers screen by spell length, callbacks falling about 45 percent over the first eight months
- L4: Frictional (search) unemployment, mostly voluntary in the simple distinction
way in: Technology and trade: tasks taken over and jobs moved
- L1: New technologies and inventions, competition from globalisation and international trade
- L2: Machines take over tasks: robots compete against human labour in the production of tasks
- L3: Displacement in exposed places: one more robot per thousand workers lowers the employment-to-population ratio by 0.18-0.34 points across US commuting zones
- L4: Structural (and technological) unemployment: skills that do not fit the jobs on offer; lasts longer than frictional
way in: The count: labour force surveys and registers
- L1: Households sampled (the US Current Population Survey: 60,000 households a month) or people registered at employment offices
- L2: The questions: any paid work in the reference week (one hour is enough), search in the last four weeks, availability
- L3: Classification: three exhaustive statuses with employment first, then unemployment; weighting to the population; durations tallied
- L4: Published measures: the headline rate (U3, LU1) and the wider ones (U1-U6; LU2-LU4 adding underemployment and the potential labour force), seasonally adjusted, monthly
way in: The seasons: seasonal work
- L1: The year's seasons and the work that follows them
- L2: Jobs that exist only in season: construction, migratory farm work; insurance in highly seasonal industries pays workers in the off-season
- L4: Seasonal unemployment, removed from the headline figures by seasonal adjustment
way in: Unemployment insurance and relief
- L1: The benefit system: payments by governmental bodies to the unemployed, from bare subsistence to a share of the lost wage
- L2: Eligibility: registered, unemployed through no fault of one's own, often required to seek work; in the US about a third of previous income, fewer than 40% of the covered apply and receive
- L3: Benefits lengthen search: a 10% rise lengthens US durations 4-8%; 60 percent of that is a liquidity effect, not moral hazard
way in: Wage setting: wages held above clearing
- L1: The wage and who sets it: minimum wage laws, insiders and unions in bargaining, employers paying efficiency wages
- L2: Insiders bargain for themselves: the employed set wages to secure their own jobs; outsiders have no voice
- L3: More seekers than vacancies at the going wage; under efficiency wages the threat of unemployment keeps effort up, so wages never fall to full employment
- L4: Classical (real-wage) unemployment; with insiders, shocks shift equilibrium unemployment itself (hysteresis)
way out: Benefits paid to the unemployed
- L1: Household income and consumption held up; time to search
- L3: Payments: in the US based on previous income, usually about a third of it, for a limited time
- L7: The decision to insure and relieve: who gets what, for how long
way out: Demand policy: fiscal and monetary stimulus
- L1: Demand for goods and so for labour
- L3: More money, lower interest rates, more spending
- L7: The decision to raise demand: deficit spending or expansionary monetary policy
way out: Lost income, less spending
- L1: Demand for goods and services
- L3: Consumption cut back
- L6: Income lost: the unemployed cannot meet their obligations
way out: Output not produced
- L1: Gross domestic product
- L3: Output below potential: Okun's law, coefficient about -0.4 to -0.5 for the US
- L5: Unemployment above its natural level
way out: Public works and government employment
- L1: People out of the pool and into work
- L3: Jobs on public projects: the WPA (1935-43), over three million young men in more than 2600 Civilian Conservation Corps camps
- L7: The decision to employ the unemployed directly: the workhouse, make-work programmes, a job guarantee
way out: The reserve's pull on wages and prices
- L1: Wage growth and inflation (the Phillips curve; a small slope)
- L3: Wages held in check: the reserve army's downward pressure; the threat that keeps effort up
- L5: The pool's size: people available for hire
way out: The toll on the person
- L1: The person's body and life: symptoms, life expectancy, relationships
- L3: Health and earnings: lower earnings for years after displacement; skills lost in long spells
- L6: Distress and the loss of latent functions
way out: Votes, protest and upheaval
- L1: Governments and regimes
- L3: Votes against governments, demonstrations, unrest
- L6: Lost trust and grievance