177 cards
Traditional potential experience formulas understate experience payoffs, whereas tracking actual 5-day work experience demonstrates an average daily wage increase of $3.83 to $4.19 (an approximate 80% daily wage increase, p < 0.05), which equals nearly 10 years of potential experience in non-experimental Malawian estimates.
Non-notified job seekers (Group C) experienced a sharper, more abrupt increase in the exit rate to work around day 90 compared to advance-notified job seekers (Group B), confirming theoretical predictions of anticipatory behavioral smoothing; however, advance notification yielded no statistically significant difference in unconditional 365-day re-employment probabilities (68.3% for Group B vs 69.7% for Group C).
School enrollment increased by 5.9 percentage points (p < 0.05) for older girls aged 11–17 (where baseline enrollment was only 51%), compared to a statistically insignificant 3.1 percentage point gain (p = 0.23) for younger girls aged 6–10 (where baseline enrollment was already 90%).
Candidates receiving algorithmic writing assistance to fix grammatical and stylistic errors were significantly more likely to receive a job offer from an employer in their first month using the platform.
Intensified active policies produced a net benefit to society of €2,000 per unemployed individual exposed to the intensive regime, driven by a 2.2-week (15%) reduction in average unemployment duration.
Intensification of the policy regime increased unemployment exit rates by 20% to 40% (a 20–22% time-invariant overall increase), reducing average unemployment duration by 2.2 weeks (approximately 15%). The perceived risk of activation accounted for the job-finding increase, whereas actual program participation exhibited zero or negative direct impacts.
Applicants with English first names and Chinese last names received a 12.5% callback rate, which was significantly lower than applicants with full English names (15.7%, p < 0.05) and not statistically different from applicants with full Chinese names (11.3%).
Treatment assignments yielded no statistically significant differences in ALMP participation days (group D mean = 0.056–0.080), meeting invitation frequencies (group D mean = 0.011–0.013), or vacancy referrals (group D mean = 0.031–0.054) across 90, 180, and 365 days (all p-values > 0.10).
The job fair produced minimal direct hiring (1 hire per 12 attending firms), but the temporary reduction in meeting costs corrected underlying misperceptions, driving a 26% increase in job board visits among jobseekers within 6 weeks and a 12 percentage point increase in professional vacancy advertising by firms (p < 0.01).
Treatment increased vacancies receiving jobseeker attribute marketing and evaluation services (valorization/immersion) from 0.003 to 0.007 (a 0.004 increase, p < 0.01), though overall firm baseline adoption of this specific service remained low (0.3% control mean).
Programs providing integrated career readiness training achieved a 10 percentage point increase in Year 2 employment rates (70% vs. 60%, p < 0.01) and an average weekly work increase of 3.4 hours per week (a 13.9% gain, p < 0.01).
Estimated meeting intensity had a strong, statistically significant positive impact on exit rates to employment (coefficients of 0.47 in Storstrøm and 2.30 in Southern Jutland), whereas estimated program participation rates had a negative or non-significant impact (-0.51 in Storstrøm, 1.73 in Southern Jutland), proving meetings are a more effective substitute for long active programs.
Jobseekers assigned to high-quality caseworkers achieved higher baseline human capital investment index scores (9.50 vs -7.33) and received significantly more training proposals in the first quarter (88.3 vs 77.7, p < 0.01) than those assigned to low-quality caseworkers. However, financial transfer treatment effects on actual training completions did not differ significantly between high- and low-quality caseworkers (-1.36 vs +4.85, p > 0.10).
Immigrants arriving prior to the 1970s eliminated their wage gap within 10 to 15 years. In contrast, immigrants arriving in the 1990s maintained a 25% lower median wage compared to native-born workers even after 10 to 15 years of host-country residence, despite holding substantially higher post-secondary degree completion rates.
From 1979 to 2018, net productivity grew by 69.6% while worker hourly compensation grew by only 11.6%. Real earnings for the top 0.1% grew by 340.7% (fifteen times faster than the bottom 90%), while earnings for the bottom 90% grew by 32.9%.
Acquiring short-term work experience increased average daily wages by $3.83 (an 80% wage increase) but resulted in a negative coefficient (-0.386, p > 0.10) on the pay frequency permanence index, indicating that wage gains were concentrated in less permanent, daily-paid positions.
Counselor match survival curves for treatment vacancies diverged from control vacancies on Day 1 (p = 0.016 log-rank test), proving that dedicated counselor pre-screening significantly accelerates successful candidate match velocity.
Graduating from a Top 200 world-ranked foreign university produced no statistically significant increase in callback rates compared to lower-ranked foreign schools. Adding a Canadian master's degree, local references, or government degree accreditation also failed to produce statistically significant callback improvements for applicants with foreign work experience.
Adding Canadian references, official degree accreditation, or permanent resident status to foreign-experienced resumes generated point estimates near zero with no statistically significant increase in callback rates.
Cost-benefit calculations by the Danish Economic Council establish that an intensified active policy regime delivers a net societal value of €2,000 per unemployed individual exposed to the intensive regime.
Professional stigma ads reduced program sign-ups by 26% (-0.032 percentage points, p < 0.001) and social stigma ads reduced sign-ups by 39% (-0.047 percentage points, p < 0.001) relative to control (0.121% sign-up rate), proving low-touch ad formats amplify stigma salience.
Financial support enabled high training completion; YearUp provided full-time weekly stipends yielding 3-year earnings increases of $7,011 (a 40% increase, p < 0.01), while Project QUEST paid 100% of tuition, fees, and emergency assistance, resulting in sustained 11-year earnings increases of $4,616 (a 15% increase, p < 0.01).
Direct monetary transfers equivalent to one or two months of median business profit significantly increased firm formalization rates, whereas informational talks produced zero impact.
Daily wage gains reflect a uniform rightward shift in the overall earnings distribution for individuals receiving job opportunities, proving that the $3.83 to $4.19 daily wage increase (p < 0.05) represents broad-based income growth rather than outlier-driven mean distortions.
Mandatory participation eliminates non-market leisure time value ($h = 0.013$, equivalent to ~2% of UI benefit levels or >1% of worker productivity), disciplining worker outside options and lowering equilibrium wages by 0.03%, which encourages firms to expand overall vacancy postings by 2.4%.
Out of all measured human capital investment components, driver's license course enrollment was the only specific outcome significantly increased by cash transfers, rising by 3.0 percentage points (p < 0.05) from a control mean of 41.9% at 12 months, with automobile usage rising to 55.4% by 24 months (p < 0.05).
Providing promotional advertising alone doubled applicants per vacancy from 25 to 51 [7], but produced no measurable impact on firm hiring [8]. Candidate identity verification alone yielded zero statistically measurable increase in hiring relative to the comparison group [8].
While higher search effort per worker resolves the first friction (reducing empty vacancies), it exacerbates the second friction (increasing double-job offers and candidate loss), causing aggregate matching efficiency to flatten and diminish beyond optimal application thresholds.
Treatment effects on permanent contract vacancies (+0.031 in below-median vs +0.038 in above-median, p = 0.550) and permanent contract jobseeker hires (+0.056 in below-median vs +0.038 in above-median, p = 0.711) showed no statistically significant difference between slack and tight markets.
75% of firm managers overestimate the cognitive ability of tertiary-educated applicants (perceiving a tertiary Raven score premium $>4\times$ the actual difference of 5.3 vs 5.0 questions); 55% of high-school jobseekers expect permanent employment within 1 year compared to 5.8% actual observed employment.
Formal apprentices were 47.5 percentage points less likely to have no education (p < 0.0001), 34.8 percentage points more likely to have completed primary education (p < 0.0001), 2.10 years older on average (p < 0.0001), and 23.0 percentage points more likely to aspire to formal wage employment (p < 0.0001) compared to traditional apprentices in control firms.
High-school educated jobseekers doubled their permanent employment rate from 5.8% to 11.6% (ITT = +5.76 percentage points, p < 0.05) and increased formal employment by 45% (+4.82 percentage points, p < 0.10), driven by an 8.7% reduction in reservation wages (p < 0.10); post-secondary jobseekers showed no statistically significant change in permanent employment (-2.6 percentage points, p > 0.10) or reservation wages (-3.5%, p > 0.10).
For job seekers with low employability (predicted median duration exceeding 6 months), assignment to an early integration agreement (month 1 or 3) increased the 1-year re-employment probability from 45% (month 6 reference group D) to 53% (group A) and 54% (group C) (p < 0.01), representing an 8 percentage point or 18% relative increase; for high-employability job seekers, early integration agreements had no statistically significant effect (82% vs 82–83%).
Mature providers with deep sector employer connections generated far higher outcomes; Per Scholas achieved a 41 percentage point gain in targeted IT employment (61% vs. 20%, p < 0.01) and a 19.6% long-term earnings increase in Year 6 (p < 0.01), whereas newer providers lacking established employer relationships generated smaller or non-statistically significant initial earnings gains.
Providing prescreening hiring services directly to firms reduced employer evaluation costs and accelerated vacancy fill rates across participating businesses.
Log vacancy stocks in experiment counties increased significantly, peaking at an 18.2% increase (p < 0.01) in May/June 2006 (3-4 months post-assignment), resulting in a projected 2.4% net increase in overall economy-wide vacancy supply under full program rollout.
Resumes with English female names received a 4.9 percentage point higher callback rate than English male names (p < 0.01). However, no female callback premium existed for Indian, Pakistani, Chinese, or Greek female applicants compared to their male ethnic counterparts.
Startup simplification increased the total number of registered businesses by encouraging former formal salaried workers to launch new enterprises, but had zero effect on the registration likelihood of pre-existing informal firms.
At baseline French market tightness ($\theta = 0.42$), displacement externalities reduce net aggregate job creation to 68%–73% of the measured firm-level treatment impact ($ED(\sigma)/\sigma AB \approx 0.68\text{--}0.73$), demonstrating that equilibrium spillover losses offset roughly 27%–32% of firm-level hiring gains.
Exit rates began diverging as early as week 2 upon letter receipt, with exit rate increases of 20–40% occurring weeks before mandatory 18-week full-time activation was scheduled to begin.
Endogenizing wage bargaining via Bertrand competition confirms that mandatory job search assistance lowers reservation wages ($w_l$), increases firm vacancy creation, and leads to a monotonic welfare decline as program participation intensity increases across the labor market.
34.38% of firm managers cited "insufficient work experience" as their primary reason for not hiring candidates met at job fairs (the single most cited reason), while 59% of firm professional vacancies required 2 or more years of formal experience compared to jobseekers where 50% believed no experience was necessary for high-skilled roles.
Arc elasticity of labor participation (extensive margin) with respect to the piece rate is 0.58, whereas the elasticity of daily output quantity (intensive margin) with respect to the piece rate is only 0.06 (controlling for reservation rates).
Female labor force participation in India was 27 percent in 2018 (roughly half the global average) and continues to decline, with migration spells remaining shorter and less frequent compared to males.
Female labor force participation in India stood at 27 percent in 2018 (half the global average); baseline loneliness among young migrant workers equals levels found in older adults aged 32–47 and chronically ill individuals.
In long-term firm employment, workers face non-monetary relational incentives, such as reputational damage or loss of future referral privileges, when recommending bad job candidates. Short-term experimental research must substitute explicit financial performance bonuses to simulate these ongoing organizational reputational stakes.
Treated firms exhibited a 24% increase in permanent contract vacancy postings (0.047 additional vacancies over a 0.199 control mean, p < 0.01) and a 10.2% increase in permanent contract hires of registered jobseekers (0.046 additional hires over 0.450 control mean, p < 0.05), generating 17.5 additional theoretical workdays per firm (p < 0.05).
Callback rates for resumes listing foreign experience with multinational companies remained depressed at 6.8% to 7.1%, showing no statistically significant advantage over resumes listing local foreign company names (4.0% to 9.9%), and failing to close the gap with domestic resumes (15.7%).
Large firms (more than ten employees) increased candidate engagement under both advertising alone and joint treatments [8]. In contrast, small firms (less than ten employees) increased engagement exclusively when assigned to the joint combination group [8].
Among firms with high professional recruitment needs, market exposure significantly reduced the proportion of professional hires holding university degrees by 7 percentage points (p < 0.10) and reduced expected 12-month future hiring targets by 30% (from 45.6 to 32.3 workers, p < 0.10), aligning recruitment criteria with candidate availability.
Training was strictly firm- and project-specific (questionnaire terminology for a single health survey), and scores were never disclosed to outside employers. Yet participants achieved an 80% daily wage increase ($3.83/day), ruling out general human capital or skill signaling as the primary driver.
Increasing fixed referral fees from Rs. 60 to Rs. 80 or Rs. 110 produced zero statistically significant impact on referral quality, choice of candidate, or task performance.
Participants who upgraded from office to flexible home work produced 19 percent less output per minute and experienced work sessions that were on average 10 minutes shorter due to home interruptions [10].
Women were willing to forgo higher wages in exchange for job schedule flexibility, and providing remote/flexible formal work arrangements led to significant increases in female labor force participation.
Resumes with foreign qualifications and foreign experience received a 6.0% callback rate compared to 15.7% for native Canadian resumes. Adding one domestic Canadian job increased callback rates from 6.0% to 8.5%, while replacing all foreign experience with Canadian experience raised callback rates to 11.0% (p < 0.01).
Self-reported survey data indicated that 15% of candidates had no prior work experience, whereas submitted résumé data indicated that 35% of candidates had no prior work experience, revealing a 20 percentage point underreporting gap on formal application documents.
Women-led enterprises formalized at significantly lower rates than male-led firms following formalization incentive programs due to domestic care burdens, heavy family financial demands, and higher business closure rates.
Participants who attended business training with a friend experienced a 4 percentage point (40 percent relative) reduction in reporting their primary occupation as housewife, accompanied by significantly increased labor hours worked and higher total sales compared to control subjects.
In-person labor inspections increased worker formalization in Brazil and Mexico, but increased worker dismissal rates in Mexico and reduced overall job creation due to higher informal employment costs [11, 12].
The program increased business formalization between 2009 and 2015 without causing formal salaried workers to shift into micro-entrepreneurship, but generated average monthly tax default rates between 49% and 51% from 2018 to 2020 [4, 5].
Treated jobseekers received a net income increase of only €40 per month (p < 0.05), while 27.7% faced bill payment difficulties, 24.4% forwent medical care, and 19.4% skipped meals. Rather than increasing consumption of temptation goods (€0 impact), treated individuals were 5 percentage points more likely to save money (45.4% control mean, p < 0.05) and saved €36.80 more in the preceding quarter (an 18% increase, p < 0.05).
Firms in the combination advertising and verification group increased online portal hiring by 8 percentage points (68 percent increase) while maintaining baseline recruitment volume through traditional informal networks, resulting in net total vacancy expansion.
Urban educated men worked only 23.4 hours per week conditional on employment; acquiring 5 days of work experience increased labor supply on the intensive margin by 1 additional day worked per week on average (p < 0.10).
Net permanent hiring gains were entirely concentrated among firms with prior agency contact (+0.124 permanent hires of registered jobseekers, p = 0.021; +0.116 quality hires staying off unemployment for >=12 months, p = 0.045), whereas firms without baseline contact experienced no significant hiring impact (-0.005 hires, p = 0.435).
Cash transfers increased total months in the program by 7.9 months (ITT, p < 0.01) from a control mean of 11.4 months, and total counselor meetings by 5.3 (ITT, p < 0.01) from a control mean of 7.3. Information exposure increased by 48% (p < 0.01), yet the estimated treatment effect on the standardized human capital investment index and final employment outcomes was precisely zero.
The interaction between experimentally assigned work experience and baseline international employer experience on inverse hyperbolic sine daily wages is positive and statistically significant (interaction coefficient = 1.731, p < 0.05; Bonferroni-adjusted p = 0.041).
Short-term work experience increases subsequent employment in administrative, managerial, and clerical roles while reducing engagement in agricultural and related primary occupations across the 8-month post-intervention follow-up.
Daily output increases from 6.02 units on Day 1 to 8.12 units on Day 4 (+2.10 units sorted per day, p < 0.01), while error rates drop from 2.20 to 1.69 errors per unit (-0.51 errors, p < 0.01), robust to worker fixed effects.
Severely disconnected jobseekers in the control group exhibited negative baseline human capital investment index scores (-11.84), but cash transfers generated a statistically significant positive effect of +14.30 points (p < 0.05; p < 0.01 for equality test across groups), driven primarily by gains in self-assessed employment prospects. Conversely, less disconnected jobseekers showed a slight negative response (-5.61 points).
At eight months, untreated job seekers in treated communities were 2.1 percentage points less likely to obtain long-term employment than job seekers in communities where no counseling was offered, and in highly competitive job markets, untreated peers experienced a 7.7 percentage point drop in employment relative to pure comparison areas [6, 7, 12].
In communities with extremely competitive job markets, untreated job seekers in treated areas experienced a 7.7 percentage point reduction in long-term employment compared to pure comparison areas where no counseling was offered.
Business process outsourcing employment increased by 4.6 percentage points (p < 0.01) with no statistically significant decline in agricultural labor (-0.008, p > 0.10) or non-BPO paid work (-0.012, p > 0.10), proving total female labor supply expanded from previously inactive populations.
Mandatory active labor market policy regimes act as a direct time tax, shortening average unemployment duration by 2.2 weeks (~15%) and generating a net social value of €2,000 per treated individual without requiring positive post-program skill gains.
Desires to work for pay increased by 13 percentage points pre-marriage (p < 0.05; control mean 30%), 10 percentage points post-marriage before children (p < 0.05; control mean 19%), and 11 percentage points after children leave school (p < 0.05; control mean 23%).
In areas with high youth unemployment (>25%), cash transfers increased counselor meetings by 6.58 (p < 0.01) versus 4.16 in low-unemployment areas (p < 0.01; equality test p < 0.01). Total personal monthly income increased significantly by €57.23 (p < 0.01) in high-unemployment regions, compared to a non-significant €19.05 increase in low-unemployment regions.
Vacancies created by treated firms showed no difference in offered wages (log wage diff = -0.017, p > 0.10) or required experience, but were 12.1 percentage points more likely to be low-skilled/unqualified positions (p < 0.01), confirming that vacancy cost reductions selectively unlock lower-productivity job creation.
Men in the full treatment group were significantly more likely to work in the public sector, compensating for household income that had shifted under the direct control of female household members.
Low-tech manufacturing facilities in developing nations experience worker turnover rates of 75 percent or more annually.
Prioritizing small and medium-sized firms (5 to 250 employees) in low-tightness micro-markets concentrated treatment among high-need sectors, achieving a 24% increase in permanent vacancies (p < 0.01) while keeping the proportion of treated firms in any single micro-market below 2% median.
Employers maximize profit by offering the lowest piece rate (5 MWK) combined with quality monitoring, exploiting low worker bargaining power. For male workers, monitoring at low piece rates increases employer profit by nearly 50% compared to unmonitored contracts.
Resumes listing foreign job experience at multinational firms received a callback rate of 6.8%, compared to 7.1% for resumes listing local foreign company names, showing no statistically significant benefit from brand recognition (p > 0.10).
Participation in activation reduced non-participants' exit rate from unemployment by ~5% (p < 0.10 in proportional hazard; -2.7 percentage points probability of employment at 3 months, p < 0.05). Comparing participants directly to non-participants within the same region overstates program efficacy by ignoring this 25%+ crowding-out spillover.
Relying heavily on informal social networks for job recruitment creates long-term structural barriers that perpetuate socioeconomic inequality across demographic groups. Exclusively circulating employment opportunities within closed social circles prevents non-connected individuals from accessing economic mobility regardless of their individual skill level.
While employment increases occurred across both RA and non-RA roles, the overall post-intervention daily wage gains were primarily driven by employment in non-RA positions rather than continued research assistant work.
Post-intervention daily wage returns were driven primarily by non-RA positions (IV coefficient = $3.803, p < 0.10 for non-RA wages vs. $3.288 for RA wages; inverse hyperbolic sine log daily wage coefficient = 0.541, p < 0.10 for non-RA vs. 0.044 for RA), demonstrating that wage gains spilled over into general market positions rather than remaining confined to research roles.
Resumes with Greek names experienced a statistically significant 4.0 percentage point reduction in interview callbacks (11.8% callback rate vs. 15.8% for English-named resumes), generating a penalty virtually identical to the 4.4 percentage point penalty experienced by Chinese, Indian, and Pakistani applicants.
Trainees who received objective feedback about their soft skills updated their expectations in line with reality regarding eligible jobs, though they became less willing to accept lower-quality employment.
Jobseekers' stated wage distributions closely matched prevailing firm salary data across occupational tiers, whereas jobseeker expectations of permanent employment (55% expected permanent job within 1 year vs 5.8% actual) and firm estimates of tertiary applicant cognitive test scores (overestimating tertiary ability premium by more than 4 times) exhibited massive structural biases.
The callback ratio between English-named and foreign-named resumes remained virtually constant at 1.3 to 1.9 across all occupational deciles, showing a 3.4 percentage point callback penalty in low language-requirement jobs like computer programmer/maintenance technician (decile 1, callback ratio 1.4) compared to a 6.7 percentage point penalty in high requirement jobs like executive recruiter/community counselor (decile 10, callback ratio 1.7).
The callback ratio favoring English-named applicants over foreign-named applicants remained virtually unchanged across skill deciles, ranging from 1.4 to 1.9 for low-language jobs like Computer Programmer (12.7% vs 9.3% callback) up to high-language jobs like Community Counselor (16.2% vs 9.5% callback).
Children whose mothers were assigned any job were 9 percentage points more likely to report that their fathers helped with cooking, cleaning, and childcare, representing a 16 percent relative increase over the 58 percent control group mean.
Experimentally induced short-term work experience resulted in a negative point estimate on payment frequency (-0.318 ITT, -0.386 IV, p > 0.10), indicating that the observed wage gains occurred through less permanent, daily-remunerated positions rather than permanent monthly roles.
Employers frequently conflate two distinct benefits of employee referral networks: pre-hire screening (reducing asymmetric information about applicant skill) and post-hire peer monitoring (mitigating moral hazard during task execution). Failing to isolate pre-hire screening capability prevents firms from designing targeted recruitment incentive structures.
Introducing worker and manager incentive systems directly increased worker efficiency and attendance, contributing to a 9.4% overall increase in plant output and a 16.6% productivity increase in the first year.
High performance pay increased the likelihood of referring coworkers by 8 percentage points and reduced relative referrals by 7 percentage points, while varying fixed fee amounts had zero effect on referral quality.
Individualized in-person assistance significantly increased formalization and tax compliance in Rio de Janeiro and Bogotá, whereas virtual messaging and group workshops produced zero statistical impact on formalization rates [8, 10].
Vocational trainees received significantly more job offers and maintained higher long-run employment rates compared to firm-provided apprentices because trainees could better certify their skills to new employers.
Personalized assistance initially increased business formalization, but the effect dissipated completely over time as newly formalized firms chose not to renew registration after experiencing zero gain in sales or profits.
Retention and case management supports enabled sectoral program trainees to work 1.3 months longer (a 9.6% increase, p < 0.01) over two years and achieve a 10 percentage point increase in jobs offering benefits (57% vs 47%, p < 0.01) relative to control group members.
Firms receiving combined advertising and verification services were 11 percentage points (73 percent) more likely to still employ their online portal hire 6 months post-posting compared to comparison firms.
Empirical findings are forthcoming as the evaluation is ongoing, with follow-up surveys planned via phone 3 months post-intervention [8, 10].
Nonexperimental wage returns relying on potential experience proxies require approximately 10 years of estimated potential experience in Malawi to match the wage gain produced by 5 days of actual measured work experience, demonstrating that potential experience drastically underestimates true per-day returns while overstating accumulated work history.
The historical model successfully stratified the experimental sample into 1,688 high-employability ($\le 6$ months) and 2,475 low-employability ($> 6$ months) individuals, correctly identifying 87% of long-duration cases in baseline data without endogenous stratification bias.
Treated firms experienced a 122% increase in vacancies assigned the prerequisites pre-selection service (0.028 increase over a 0.034 control mean, p < 0.01), where counselors co-defined 1 to 5 unobservable screening criteria with employers.
Vacancies in treated firms experienced a 58% increase in pre-screening and filtering service usage (3.4 percentage points over 5.8% control mean, p < 0.01) and a 57% increase in private listings, leading to a 39% reduction in direct jobseeker application volume (2.67 fewer applications, p < 0.01) and 0.84 fewer firm-initiated candidate searches (p < 0.01).
Treatment increased the proportion of vacancies designated as private (hidden from public internet access) by 57% (0.030 increase over 0.053 control mean, p < 0.01), which mechanically reduced direct jobseeker applications by 39% (-2.67 applications, p < 0.01).
Women aged 18–24 in treatment villages were 2.8 percentage points more likely to enroll in vocational training (p < 0.01) relative to a control mean of 0.5%, whereas men showed no significant change (0.3 percentage points, p > 0.10).
Combining PES pre-screening with statutory trial periods resulted in a 14% increase in hires who remained off unemployment registers for at least 12 months (0.046 quality hires, p < 0.01), showing that 100% of the net permanent contract hiring increase successfully survived probation.
Specific active labor market programs exhibited zero positive direct impacts during participation, showing statistically significant negative lock-in effects on unemployment exit rates during program weeks and negative post-program effects across training, subsidized job, and education categories.
Emphasizing that training was subsidized for financial hardship produced no statistically significant change in application rates (+0.003, SE = 0.022, control mean = 0.426) and generated no significant heterogeneous treatment effects across machine learning groups (p = 0.315).
Annual earnings for participants were $5,080 higher (a 22% increase, p < 0.01) in Year 6 and remained $4,616 higher (a 15% increase, p < 0.01) in Year 11 [10, 25, 26]. In Year 11, participant employment reached 83.8% compared to 73.7% in the control group (a 10.1 percentage point increase, p < 0.01), representing the longest-sustained earnings gains in US workforce evaluation history [10, 22, 26].
Mature provider Per Scholas increased Year 2 targeted sector employment by 41 percentage points (61% vs. 20%, p < 0.01) and Year 6 earnings by 19.6% (p < 0.01). Newer providers with less sectoral experience, such as Towards Employment, yielded no statistically significant long-term earnings increases.
Free early childhood education significantly reduced the post-motherhood spike in female informality, helping address the 76.4% informal contract rate among Brazilian domestic workers in late 2023.
Job seekers who received public branded certificates were 5.2 percentage points more likely to be employed, increased total hours worked by 20%, and earned US$130 over 3 months (generating an earnings gain 5.6 times greater than the per-participant program cost).
YearUp generated a total net benefit to society of $15,349 per participant ($38,484 in total benefits compared to $23,135 in total costs), returning $1.66 to society for every $1.00 spent on program operations (a 66% net positive return, p < 0.01).
Public sector ALMP placements show minimal wage gains in developing economies, whereas 5 days of private/international firm work experience in urban Malawi increased average daily wages by $3.83 to $4.19 (~80% wage increase, p < 0.05).
Displaying detailed public information about workers' previous performance significantly increased job matches and overall earnings for strong performers on the platform.
Quality monitoring decreases output quantity by 0.440 to 0.689 units per day (p < 0.01) while significantly improving quality by reducing sorting errors by 0.561 to 0.744 errors per unit (p < 0.01). Explicit quality monitoring does not increase worker reservation rates (WTA effect = -0.431 MWK, p > 0.10).
From 2000 to 2019, Black workers' median wages were 75.6% of white workers' median wages; sectoral interventions generated earnings gains for workers of color driven by a 33 percentage point increase in IT sector placement (37% vs 4%, p < 0.01) in YearUp and an 8 percentage point increase in technical health roles (29.4% vs 21.5%, p < 0.01) in Project QUEST.
Randomized sorting ensured baseline equivalence between groups, establishing that sectoral programs causally increased 2-year earnings by $4,500 in SEIS (18% increase, p < 0.01), 5-year earnings by $2,716 in WorkAdvance (11.5% increase, p < 0.01), and 11-year earnings by $4,616 in Project QUEST (15% increase, p < 0.01).
Acquiring experimentally induced work experience reduced the probability of submitting a reference letter by 7.9 percentage points (p > 0.10) and decreased the frequency of reference letter submissions by 0.381 times, demonstrating that formal signaling via reference letters did not drive the observed wage gains.
Under standard unincentivized conditions, male recommenders do not systematically refer the highest-quality candidates from their social networks; offering explicit financial performance incentives significantly increases the ability and test scores of referred job seekers.
High-ability original participants successfully selected high-skilled referrals when offered performance-based pay, whereas low-ability original participants failed to recruit high-skilled candidates under any incentive structure.
Performance pay incentives decreased relative referrals by 7 percentage points and increased coworker referrals by 8 percentage points.
Tested job seekers receiving hard-to-observe skill certificates were significantly more likely to receive a formal and permanent job offer while performing the same amount of job search as comparison candidates.
Short-term work experience produced a statistically non-significant effect on monthly reservation wages (IV estimate = +$124.40, p > 0.10) and no significant increase in active job search duration (IV estimate = +0.089, p > 0.10), ruling out reservation wage shifts as the mechanism driving daily wage returns.
Receiving short-term work experience resulted in a $124.40 increase in monthly reservation wages that was statistically insignificant (p > 0.10), and produced no statistically significant increase in the proportion of time spent actively searching for work (coefficient = 0.089, p > 0.10).
Resumes with English names received a 15.7% callback rate compared to 11.3% for resumes with Indian, Pakistani, or Chinese names (a 4.4 percentage point reduction, p < 0.01). Resumes with Greek names experienced a similar 4.0 percentage point reduction (11.8% callback rate, p < 0.05).
While only 15% of respondents reported no prior work experience in self-reported survey modules, 35% lacked work experience on their submitted résumés, revealing a 20 percentage point discrepancy caused by infrequent résumé updates and differing worker definitions of suitable CV entries.
Four years after baseline, formal dual apprentices increased non-routine analytical task involvement by 0.243 standard deviations (p < 0.0001) and decreased overall routine task intensity by 0.213 standard deviations (p = 0.008).
Abstract logic and pattern-matching tasks successfully stratified participant cognitive capacity in low-literacy informal labor settings without reliance on formal schooling credentials.
Increased average schooling levels served as the primary mechanism explaining the 12 percentage point drop in Brazilian labor informality from 2003 to 2012, counteracting the informalization pressure from rising minimum wages.
Treated job seekers had a statistically significant 2.5 percentage point employment advantage over untreated peers at 8 months, but by 12 and 20 months, employment rates between treated and comparison groups were identical.
Treatment vacancies saw a 76% reduction in firm-initiated candidate CV database searches (0.84 fewer matches from a control mean of 1.102, p < 0.01) and a 39% reduction in direct jobseeker applications (2.67 fewer applications from a control mean of 6.862, p < 0.01), proving firms actively delegate sourcing labor to public employment counselors.
Over two years, participants earned $4,500 more (an 18% increase, p < 0.01) than control group members, with Year 2 earnings $4,000 higher (a 29% increase, p < 0.01). Year 2 employment reached 70% compared to 60% for the control group (a 10 percentage point increase, p < 0.01). Participants were 14 percentage points more likely to work jobs paying at least $11/hour (59% vs. 45%, p < 0.01) and 10 percentage points more likely to secure jobs offering benefits (57% vs. 47%, p < 0.01).
Women hold approximately 45 percent of all BPO jobs in India—substantially higher than in most other industries—with starting salaries ranging from 5,000 to 10,000 INR (US$110-220) per month, which was twice the average pay of non-BPO jobs for similar education levels.
Subsidized dual apprentices achieved a 15.1 percentage point higher rate of formal skill certification (p < 0.0001) relative to a 9.5% control mean, contributing to a 15% increase in total earnings (9,394 FCFA/month, p = 0.0169) 2 years post-program completion.
64% of youth searched specifically for permanent jobs (versus 11% seeking temporary work), work stability was the second most cited desirable job characteristic (20.4% versus 6.7% for working hours), and only 5.4% cited starting a business as a reason for non-search.
Market exposure reduced worker reservation wages by 6.6% (p < 0.10) and wage mismatch by 4.2% (p < 0.10), while increasing job board visits by 3.01 visits (p < 0.05); treated firms increased professional vacancy advertising by 12 percentage points (p < 0.01) and job board recruitment by 9.6 percentage points (p < 0.05).
Monthly cash transfers restricted to non-formal workers created a statistically significant disincentive for program beneficiaries to accept formal employment contracts.
Experimentally induced short-term work experience increases the likelihood of hearing about a job opportunity through contacts met during the intervention by 22.2 percentage points (p < 0.05) [19, 21]. It increases the standardized social network index by 0.306 standard deviations (p < 0.10) [19, 21].
Baseline survey results indicate that Yuva Parivartan graduates experience initial placement difficulties and disproportionately depend on personal social connections to find employment.
Treated firms received nearly 1 additional spontaneous candidate profile push per firm from public counselors compared to control firms which received virtually zero (a 488% jump in total counselor interactions including spontaneous profile pitches, p < 0.01), effectively uncovering latent job demand.
Evaluating worker productivity in experimental field settings requires objective, non-credentialed skill measures that avoid cultural or educational bias while reflecting real-world cognitive problem-solving ability. Unvalidated task metrics risk measuring test familiarity rather than actual job-relevant cognitive capability.
Youth receiving automated recommendation letters earned US$1,349 more cumulatively across four years and stayed in jobs 0.36 months longer. In South Africa, women using reference templates saw higher application review rates and increased 3-month employment rates, closing gender hiring gaps.
Failure to fulfill agreed search obligations or submit application verification results in a mandatory 2-week total unemployment benefit withdrawal ("Sperrzeit"), with repeated infractions triggering complete benefit disqualification for extended durations.
Short-term employment gains at eight months were driven entirely by job seekers finding fixed-term contracts with zero increase in permanent employment contracts, and all employment advantages vanished completely after twelve months.
Addressing social stigma shifted the applicant demographic index by 0.174 standard deviations toward higher socioeconomic status (p < 0.05). Professional stigma messaging increased the proportion of wealthy applicants by 62% relative to control (p < 0.01) and currently employed applicants by 28% (p < 0.05).
Stigma salience information significantly increased the belief that society looks down on entry-level jobs (+0.10 points, p < 0.01) and decreased expected family pride (-0.15 points, p < 0.10), while producing zero statistically significant change in expected starting salary (+57 EGP/month, p = 0.27) or 5-year expected salary (-92 EGP/month, p = 0.46).
Merely mentioning social stigma reduced job fair attendance by 1.4 percentage points overall and by 12.0 percentage points for the bottom predicted response group (p = 0.022). Adding testimonials attenuated this negative effect to -0.044 percentage points (statistically insignificant), demonstrating that salience forces can offset mitigation efforts.
Stigma treatments shifted the applicant baseline characteristics index by +0.174 standard deviations in Exp 1 (p < 0.05) and +0.914 in Exp 2 (p < 0.05), increasing the applicant share of wealthy jobseekers by 62% (p < 0.01) and currently employed jobseekers by 28% (p < 0.05).
Vocational job training programs significantly increased participants' probability of securing formal employment and improved overall employment quality compared to non-participants.
Evaluating employee productivity strictly by task completion or completion speed overlooks the supervisory time and firm resources consumed when workers make repeated mistakes. Comprehensive performance metrics must penalize error rates and managerial intervention alongside completion speed to reflect true net productivity.
Excluding month 1 eliminates survey recall error where participants erroneously report recruiter-assigned experimental work as post-intervention employment, ensuring that estimated daily wage increases ($3.83 to $4.19 per day, p < 0.05) reflect independent labor market returns.
The average positive formalization effect of incentive programs in Brazil was driven 100% by high-income entrepreneurs, as small low-productivity firms derived negligible benefit from formalizing.
Attaching formal third-party accreditation to foreign educational credentials produced virtually zero effect on callback rates (point estimates ranging from -0.014 to +0.007, statistically insignificant), failing to reduce the callback gap between foreign-educated and native-educated applicants.
Job exit rates spiked significantly in the weeks leading up to mandatory participation (a "threat effect"), accounting for a substantial share of total job exits before individuals ever received actual coaching or assistance.
Empirical estimations revealed non-positive or negative exit hazard coefficients following program completion across multiple program types (e.g. vocational training post-completion coefficient -0.49 for control group members in Storstrøm).
Official written notifications warning of legal penalties significantly increased the proportion of targeted informal firms that completed formal registration.
Sectoral programs substantially increased industry-recognized credential attainment; earnings gains were driven by placement in higher-paying credentialed jobs rather than total hours worked, yielding hourly wage increases of nearly $4/hour (a 31.7% increase, p < 0.01) in YearUp and a 14 percentage point increase in jobs paying over $11/hour (59% vs. 45%, p < 0.01) in SEIS.
While permanent contract hiring gains remained persistent through the entire 16.5-month period, treatment effects on vacancy postings with the Public Employment Service peaked during the sanctuary period (+24%, p < 0.01) and progressively decayed starting May 2015 as proactive counselor contact ceased.
Professional stigma messages decreased application rates by 2.9 percentage points on average (-0.125 percentage points for the bottom Machine Learning group, p = 0.015), while personal stigma messages showed virtually zero effect (+0.006 average, +0.047 bottom group, p = 0.756), proving these stigma dimensions operate through distinct psychological mechanisms.
While individuals offered counseling were 2.5 percentage points more likely to find long-term employment after eight months than untreated peers in the same communities, untreated peers in treated communities were 2.1 percentage points less likely to find jobs than job seekers in pure comparison communities where 0% were treated [6, 7].
Reinforced pre-screening led to a 14% increase in quality permanent hires who remained off unemployment rosters for at least 12 months (0.046 additional quality hires over a 0.330 control mean, p < 0.01), confirming counselor vetting effectively counteracts negative productivity signaling.
The interaction between accountability treatments and baseline reciprocity was consistently negative (e.g., Monetary Penalty x Baseline Reciprocity coefficient = -0.269, p = 0.022), demonstrating that accountability sanctions selectively increase prosocial performance among unmotivated workers who have higher baseline absenteeism.
Proves mathematically that in the intermediate "unsafe regime" where minimum qualification $z$ binds, a downward belief shock ($\mu$) increases search expenditure ($s$) to avoid remaining unmatched under vacancy penalty ($\kappa$).
Mandatory upfront screening ensures candidate readiness for technical training; programs utilizing strict screening achieved up to a 41 percentage point increase in targeted sector employment (61% vs. 20%, p < 0.01) and 19.6% long-term earnings increases (p < 0.01).
The urn-ball functional form micro-founds crowding-out externalities, yielding structural elasticity estimates where increased search intensity of treated workers directly depresses the contact rate per unit of search effort for non-treated workers.
Targeting useful vacancies led to a 24% increase in permanent contract vacancy listings (0.047 vacancies, p < 0.01) and a 10.2% increase in permanent contract hires (0.046 hires, p < 0.05) yielding 17.5 additional theoretical workdays per firm (p < 0.05), while temporary/fixed-term hires saw no significant net increase.
Facilitating direct recruitment channels led to a over 10-fold increase in BPO employment among young women in treatment villages (control mean 0.4% vs treatment 5.0%, p < 0.01) despite job locations requiring relocation or long commutes.
Short-term labor earnings in the treatment group fell by 10,494 FCFA/month (-25%, p < 0.001) due to reductions in wage work (-6,414 FCFA) and self-employment (-6,381 FCFA). The monthly program stipend of 10,213 FCFA fully offset these forgone earnings, resulting in zero net change in total short-term income.
While a naive microeconometric evaluation calculates a false positive net return of 4,094 DKK (~550 EUR) per participant, structural simulation proves social welfare is maximized at only 20% treatment intensity and drops by 0.13% under full (100%) rollout, while government UI program expenditures are minimized at 30% intensity.
Starting BPO salaries for entry-level candidates with basic English and computer skills ranged from 5,000 to 10,000 INR ($110-220 USD) per month in 2003, representing 200 percent of the average pay for non-BPO workers with similar educational attainment.
25.9% of formal apprentices substituted out of traditional informal apprenticeships ($\omega = 0.259$, p < 0.001); overall youth entry into any apprenticeship increased by 52.8 percentage points despite this 18.5 percentage point windfall drop.
Non-routine analytical tasks increased significantly within sectors (+0.139 standard deviations, p < 0.01), driving a significant within-sector reduction in overall routine task intensity (-0.175 standard deviations, p < 0.05).
Pooled earnings across all four providers increased by $1,945 (a 14% increase, p < 0.01) in Year 2 and $2,716 (a 11.5% increase, p < 0.01) in Year 5 [10, 11]. Mature providers yielded higher impacts; Per Scholas participants achieved a 41 percentage point increase in IT sector employment in Year 2 (61% vs. 20%, p < 0.01) and a 19.6% long-term earnings gain in Year 6 (p < 0.01) [10, 12, 13].
By Year 3, participant annual earnings grew by $7,011 (a 40% increase, p < 0.01), driven by a 33 percentage point gain in IT sector employment (37% vs. 4%, p < 0.01) and a 9 percentage point gain in financial services (23% vs. 14%, p < 0.01) [10, 21]. By Year 5, quarterly earnings remained $1,857 higher (a 33.5% increase, p < 0.01), generating $1.66 in public benefit per $1 spent ($15,349 net benefit per participant) [10, 20].