About the Research & Source
This summary is a simplified, educational interpretation of the original academic paper. All ideas and findings are drawn directly from the researchers’ work and restated here for wider accessibility.
“Duration Dependence and Job Search over the Spell: Evidence from Job Seeker Activity Reports”
by Jonas Cederlöf, Sara Roman
Note: This educational summary is not affiliated with the authors or arXiv. It is intended purely for academic and illustrative purposes. For the original paper, theoretical framework, and formal proofs, visit the official arXiv publication.
TL;DR — Plain-English takeaways
The paper tackles a simple question with big policy stakes: as unemployment lasts, do people actually search less, or do results worsen mainly because of who remains unemployed? It also examines what changes at unemployment-insurance (UI) exhaustion and how much extra effort still converts into interviews.
- Search effort ≈ flat. A small dip appears right before job start (people already secured a job) and a ~10% fall occurs at UI exhaustion as programs crowd out search.
- Callbacks fall with duration, but mostly selection. Only about 10–14% of the decline is “true” duration dependence; the rest reflects dynamic selection.
- More applications help, with tapering gains. Doubling applications over three months ≈ +30% chance of getting at least one interview.
- Heterogeneity matters. Non-natives get fewer callbacks and steeper within-spell declines; tighter markets amplify duration dependence; age patterns differ.
What’s the core question?
The study asks: Why do job-finding and callback rates drop the longer someone stays unemployed? Is it because unemployment itself hurts people’s chances, or because the kind of people who remain unemployed changes over time?
The authors test two competing explanations:
- 1. Dynamic selection People with stronger job prospects leave unemployment sooner. Over time, those left are typically the ones facing tougher odds—so averages decline even if individual effort doesn’t.
- 2. “True” duration dependence Being unemployed longer directly reduces one’s chances—because employers may view long spells as a negative signal or because skills erode with time.
Why does this matter?
Understanding why job-finding rates decline over time isn’t just academic — it shapes how governments design unemployment policies and how employers view job seekers.
Depending on which mechanism dominates, the right policy looks very different:
- If dynamic selection dominates: Prioritize early and targeted support for harder-to-place workers. Focus on personalized matching, training, and outreach before discouragement sets in.
- If true duration dependence dominates: Tackle stigma and signaling problems. Encourage employers to consider candidates with longer spells, promote “fresh skill” certificates, and maintain candidate visibility over time.
Data & Method
The study relies on a uniquely rich dataset that tracks what job seekers do each month, how their behavior changes, and how their chances of finding work evolve.
- 2.4 million monthly activity reports from Sweden’s Public Employment Service (2014–2019), detailing applications, interviews, and participation in training or programs.
- Focuses on UI-eligible job seekers, linked to administrative registers with demographics, past earnings, and benefit data.
- Key idea: Compare each person to themselves over time (“spell fixed effects”) to isolate real behavioral change from who exits unemployment.
- Timing adjustment: The last ~2 months of each unemployment spell are excluded (often a “waiting-to-start” period) to capture active job search only.
How the data and analysis fit together
Key Findings
The data show two simple truths: most of the fall in interview chances comes from who stays unemployed, not from people giving up, and job search effort stays nearly constant throughout unemployment.
Why interview chances fall over time
Most of the decline in callbacks is explained by dynamic selection—people with better prospects find work early—while only a small part is due to true duration dependence (being unemployed longer actually hurting chances).
Based on fixed-effects estimates of callback probabilities.
Job-search effort over the unemployment spell
Job seekers send roughly the same number of applications month after month. There’s only a small dip at UI exhaustion and right before they start a new job.
Average search effort remains stable until the final months of unemployment.
Do more applications help?
In short: yes — but with limits. Sending more applications clearly increases your chance of getting an interview, although each extra effort yields a smaller additional payoff — a case of diminishing returns.
- A 50% increase in monthly applications (for example, from 4 to 6) raises your next-month callback rate by about 8%, even when effort in nearby months is similar.
- Doubling the number of applications over three months increases the likelihood of getting at least one interview by around 30%.
Diminishing returns to job search effort
The more applications you send, the higher your chance of getting interviews — but returns flatten as you go.
What happens when unemployment benefits run out?
When people reach the end of their unemployment insurance (UI) benefits — the payments that support them while job hunting — their number of applications drops by about 10%.
This doesn’t mean job seekers stop looking. In most cases, they shift their time toward government-backed programs such as training courses, job-practice placements, or counseling sessions. These activities are valuable, but they also reduce the time available for sending new applications.
The ~10% decrease appears across both short and long benefit durations.
Who is most affected?
The impact of unemployment duration isn’t the same for everyone. It varies by age, background, and local labor-market conditions.
- Age: Younger job seekers’ search effort declines more noticeably over time, while older workers tend to maintain — or even slightly increase — their job-search activity.
- Nativity: Non-native workers apply to more jobs but still receive about 29% fewer callbacks and face roughly 17% higher sensitivity to unemployment duration.
- Labor-market tightness: In markets with more competition or fewer openings, the decline in callbacks over time is about 50% stronger.
Differences across groups
Duration effects grow stronger for non-natives and in tight job markets.
Policy & practice takeaways
- Address employer-side biases and stigma toward long-term unemployment — not just “motivation” among job seekers.
- Keep candidates visible: promote fresh skill certificates, structured outreach, and updated references to counter negative duration signals.
- Design public programs to complement active job search — especially around benefit milestones — so they enhance employability without displacing application activity.
Conclusion
This study challenges a common belief about unemployment: people don’t lose motivation over time — they keep searching just as hard. What really changes is how the labor market responds to them.
The decline in job-finding rates is driven mostly by who remains unemployed — not by falling effort. Those still looking after several months tend to face tougher odds from the start. Meanwhile, a smaller share of the slowdown comes from “true” duration effects, such as employer hesitation toward candidates unemployed for longer periods.
Policy should therefore look beyond job seekers’ behavior and focus on the environment they face: tackling stigma, supporting skill renewal, and keeping candidates visible to employers. Activation programs and training can help — but only if they complement active search instead of replacing it.
In short: long-term unemployment is less about people giving up, and more about how the market treats them once time passes.