Returning to work after years at home: what the hiring audit studies found

What the audit studies actually measured

An audit study sends fictitious applications to real job postings, varies one thing at random, and counts responses. Weisshaar submitted one application to each of 3,407 openings listed between August 2015 and January 2016, sampled from 50 major US metropolitan areas, across five occupations that require a degree but no licence: human resources managers, marketing directors, accountants, financial analysts and software engineers. Every fictitious applicant was a parent, college-educated, with two jobs since graduation and about 9.5 years of experience, placing them at roughly 32 to 34 years old. Only two things varied: gender, and whether the resume showed continuous employment, an 18-month layoff, or an 18-month lapse annotated with the line left to take care of my children. Across the whole sample, 9.45 percent of applications produced an interview request and 8.34 percent produced a formal rejection. The rest went unanswered, which is normal for this method and is the reason the silence you are experiencing is not diagnostic on its own. An older audit study in the same tradition, Correll, Benard and Paik in 2007, found childless women receiving 2.1 times as many callbacks as equally qualified mothers, which is a penalty for motherhood itself rather than for a gap. Weisshaar's resumes drew higher callback rates than that study's, and she suggests parenthood signals may be swamped by the stronger employment-history signal. What this design proves is causal, because nothing except the assigned condition differed between resumes. What it cannot tell you is how a real person with references, a portfolio, a former manager who will vouch for them, and the ability to explain themselves in a room would fare. The experiment measures one pass over a resume by a stranger. It does not measure you.

Leaving to care for children is read differently from being laid off

The headline result is that the two kinds of gap are not read the same way. Relative to continuously employed applicants, laid-off applicants received about two-thirds as many callbacks: 8.8 percent for fathers and 9.7 percent for mothers, against 14.6 percent for employed fathers and 15.3 percent for employed mothers. Applicants who had left to care for children did worse than both: 5.4 percent for fathers and 4.9 percent for mothers. Weisshaar reports that opt out applicants were about half as likely to get an interview request as their unemployed counterparts, and the difference was statistically significant. The gender differences within each condition were not. That last detail matters and is usually left out of the coverage. In the callback rates, fathers who had stayed home were penalised as heavily as mothers who had, and Weisshaar records no evidence of a motherhood penalty in the main effects. Two qualifications belong with that. In the survey experiment, opt out fathers were rated more harshly than opt out mothers. And when Weisshaar split the audit results by how competitive the local labour market was, gender effects did appear: penalties emerged for continuously employed mothers and were larger still for opt out fathers. So the headline penalty attaches to the choice to leave rather than to being a woman, but the paper does not report a clean absence of gender. Weisshaar's companion survey experiment, fielded by YouGov to 1,000 US respondents in January 2014, points at the mechanism: respondents rated stay-at-home applicants lower on capability, reliability and deservingness by about 0.2 of a standard deviation each, and lower on commitment by about 0.46. Against the laid-off applicants specifically, the opt out penalty was significantly worse on commitment, deservingness and reliability. In other words, screeners appear to treat a layoff as something that happened to you and caregiving as something you chose.

Explaining the gap does not repair it

The standard advice is to name the reason for the break on the resume so the reader is not left guessing. That advice has been tested and it did not hold. Kristal, Nicks, Gloor and Hauser sent 9,022 applications to real UK employers between October 2019 and March 2020 and compared four resume formats: no gap, a 2.5-year gap left unexplained, the same gap explained with the sentence left to become a full-time mother and look after my children, and a version that listed each job by its duration in years rather than by calendar dates. The explained gap performed no better than the unexplained gap. The years format outperformed all three other conditions, including the resume with no gap at all. The authors state plainly that they found no empirical support that explaining the gap works better than leaving it unexplained. Two things follow. First, if you have been rewriting your cover letter to justify the break, that effort is not buying you callbacks in the way you were told. Second, the one intervention that did move the number was a formatting change that costs nothing: writing durations instead of dates, so a reader takes in nine years of experience rather than a hole between 2016 and 2025. That finding comes from one field experiment in one country and has not been replicated widely, so treat it as promising rather than settled.

How many people are standing where you are

This is a large population, which matters mainly because it tells you the market has seen your resume before. In the UK, 1,518,000 people aged 16 to 64 were economically inactive in April to June 2026 because they were looking after family or home, of whom 1,306,000 were women, out of 9,111,000 economically inactive people in total. That is the Office for National Statistics series INAC01, seasonally adjusted, published on 18 August 2026. In the US, the Bureau of Labor Statistics put the labour force participation rate for women aged 16 and over at 56.4 percent in July 2026, down from 57.2 percent in January. BLS does not publish a headline count of people who left work specifically to care for children and want to return, so any figure you see quoted for that is an estimate someone has constructed, not an official series. What the published numbers support is modest: the caregiving break is common, it is overwhelmingly taken by women, and returning is a routine transition rather than an exotic one. What they do not support is a claim about how long your search will take. Nobody publishes a reliable time-to-hire figure for returners, and anyone quoting one to you should be asked where it came from.

What a certification changes, precisely

Here is the honest boundary. The penalty in these experiments is employer behaviour triggered by a signal on a resume. A certificate does not remove that signal, and no study cited on this page tested adding a credential to a resume with a caregiving gap. Anyone telling you a certification raises your callback rate is extrapolating past the evidence. What a credential does change is three specific things. It puts a dated item inside the blank years, so the most recent line on your resume is 2026 rather than 2016. It converts a claim into something a screener can check, because the issuing body maintains a public verification record. And it changes the keyword you are matched on, from a job title you last held a decade ago to a tool the employer named in the advert. That third one is the largest effect in practice, because it moves you out of the pile where the reader is comparing gap lengths and into the pile where the reader is filtering on a product name. None of this repairs the commitment judgement Weisshaar measured. It gives the reader something else to look at first. That is a smaller promise than the one training providers make, and it is the one the evidence supports.

Which credentials carry currency, and what they cost

Pick by what local job adverts name, not by prestige. If the postings you are answering say Power BI, Microsoft's PL-300 is the exam behind the Power BI Data Analyst Associate certification. Microsoft gives you 100 minutes for it and prices it by the country or region where it is proctored rather than publishing one global figure, so check your own country's price before you budget. It also expires: the certification carries a 12-month renewal frequency, with renewal done by a free online assessment on Microsoft Learn. If the postings say project coordinator, PMI's CAPM has the unusual property of requiring no work experience. The eligibility bar is a secondary degree, meaning a high school diploma, GED or global equivalent, plus at least 23 hours of project management education completed before you sit. The exam is 150 questions in 180 minutes, 15 of them unscored pretest items, and you may attempt it up to three times in one year. Holders need 15 professional development units every three years to keep it. On the free end, Google and HubSpot both issue certifications at no cost. I could not load a Google-owned page stating the current validity window for the Google Analytics certification while writing this, so if a course seller quotes you an expiry date for it, ask them for the Google URL. A credential with no fee still costs you the hours, and hours are the resource you are shortest of.

Returnships are the one route with a published conversion number

A returnship is a paid, fixed-term placement aimed at people coming back from a career break, usually structured so the employer has already accepted the gap before you apply. Path Forward, the nonprofit that runs these programmes with employers in the US, states that more than 80 percent of returners who complete a returnship convert to full-time positions, with the remainder generally finding employment within six months, and puts the typical programme length at three to four months with a range from eight weeks to six months. Read that number with care. Path Forward runs the programmes and does not publish the underlying counts, the denominator, or how many applicants are screened out before a placement starts. It is a conversion rate among people who were selected and who finished, not an odds estimate for an applicant. Even discounted, it points at something real: the returnship route removes the specific screen that the audit experiments identify, because the gap is the entry criterion rather than the disqualifier. The catch is volume. There are far fewer returnship seats than there are people who want one, and they cluster in large employers and a handful of functions. Apply to them, but do not make them your only plan.

The order that costs least

Spend nothing until you have tested the free changes, because the evidence says one of them outperformed everything else in the Kristal experiment. First, reformat: list each role with its duration in years rather than a start and end date. That was the only intervention in a 9,022-application field test that beat a resume with no gap at all. Second, stop writing the justification paragraph, since the explained-gap condition performed no better than the unexplained one. Third, send a batch of 20 to 30 applications in the reformatted style and count responses before you conclude anything, because the base interview-request rate across Weisshaar's entire sample was 9.45 percent and a run of ten silences tells you nothing. Only then consider paying for an exam, and only if you can point to job adverts within your commuting distance that name the specific tool by name. If you cannot find those adverts, the exam is not the constraint. Two situations argue for buying nothing at all: if a licence or degree, rather than a skill, is what the roles you want require, and if your money is short enough that an exam fee competes with a bill. In both cases the returnship route and the free reformat are better uses of the next month than a certificate.

What this does not fix

This page will not claim that a certification raises your callback rate after a caregiving break. Every experiment cited here measured how employers respond to fictitious resumes, and none of them added a credential to a resume with a gap and remeasured. The penalty Weisshaar found is a judgement about commitment that a certificate does not address. What a credential changes is the date on your most recent line, the verifiability of the claim, and the keyword you are screened on.

Where an exam fits

• Microsoft Office Specialist (MOS), if the roles you are answering are administrative or coordinator roles that name Excel or Word in the advert and you need one verifiable item dated this year
• Power BI Data Analyst (PL-300), if local job adverts say Power BI by name and you can absorb a credential that must be renewed every 12 months
• Google Analytics Certification, if you are aiming at marketing roles and want a zero-fee credential before you spend anything
• HubSpot Certifications, if the marketing or sales roles near you name a CRM and you need to show recent tool exposure without a fee
• Salesforce Administrator, if you can find Salesforce admin openings within commuting distance, since the credential is worth little where nobody runs the platform
• CAPM (Certified Associate in Project Management), if you coordinated work before the break but never held the title, since CAPM asks for 23 hours of education and no work experience

Common questions

Sources

1. 01Interview request rate for applicants whose resume said they left work to care for children, compared with continuously employed applicants, in a randomised audit study 4.9 percent for opt out mothers and 5.4 percent for opt out fathers, against 15.3 percent for employed mothers and 14.6 percent for employed fathers
   Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28
2. 02Interview request rate for applicants who said they were laid off due to downsizing, same study, same 18-month gap length 9.7 percent for mothers and 8.8 percent for fathers, roughly two-thirds the employed rate and about twice the opt out rate
   Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28
3. 03Scale and design of the Weisshaar audit study One application sent to each of 3,407 job openings in 50 US metropolitan areas between August 2015 and January 2016, across five occupations requiring a degree but no licence (human resources managers, marketing directors, accountants, financial analysts, software engineers); every applicant was a college-educated parent with two jobs since college and about 9.5 years of experience, implying an age of approximately 32 to 34, with the employment lapse fixed at 18 months in both lapse conditions
   Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28
4. 04Overall employer response rate across the entire Weisshaar sample, all conditions combined 9.45 percent received an interview request and 8.34 percent received a formal rejection; the remainder went unanswered
   Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28
5. 05Callback ratio between childless women and equally qualified mothers in the audit-study component of Correll, Benard and Paik Childless women receive 2.1 times as many callbacks as equally qualified mothers. The summary describes the audit design (matched same-sex pairs of one parent and one non-parent sent to real openings) but does not publish the sample size or a corresponding callback ratio for men, so neither is asserted here
   Correll, Benard and Paik, Getting a Job: Is There a Motherhood Penalty?, American Journal of Sociology 112(5), 2007, as summarised by the Harvard Kennedy School Gender Action Portal checked 2026-08-28
6. 06Whether explaining an employment gap on a resume improves callbacks, tested in a field experiment with real employers Across 9,022 applications sent to real UK employers from October 2019 to March 2020, the explained gap condition performed no better than the unexplained gap; both scored significantly below a format listing job durations in years rather than calendar dates, which beat all other conditions including the no-gap resume. Where a gap was used it was set at 2.5 years, the average time out of the workforce for women who leave paid employment. The authors state they found no empirical support that explaining the gap works better than leaving it unexplained
   Kristal, Nicks, Gloor and Hauser, Nature Human Behaviour, 2022 checked 2026-08-28
7. 07People aged 16 to 64 in the UK economically inactive because they are looking after family or home 1,518,000 in April to June 2026, of whom 1,306,000 are women, out of 9,111,000 economically inactive in total
   Office for National Statistics, dataset INAC01 SA, economic inactivity by reason, seasonally adjusted, released 18 August 2026; figures read from the Apr-Jun 2026 row of the People and Women sheets of the current edition spreadsheet linked on this page checked 2026-08-28
8. 08US labour force participation rate for women aged 16 and over 56.4 percent in July 2026, down from 57.2 percent in January 2026, seasonally adjusted
   US Bureau of Labor Statistics, series LNS11300002 checked 2026-08-28
9. 09Share of returnship participants who convert to a full-time role, as stated by the organisation that runs the programmes More than 80 percent of returners who complete a returnship convert to full-time positions; typical programme length is three to four months, ranging from eight weeks to six months
   Path Forward checked 2026-08-28
10. 10CAPM eligibility and exam format Secondary degree such as a high school diploma or GED, plus at least 23 hours of project management education before the exam, with no work experience requirement; 150 questions in 180 minutes, 15 of them unscored; up to three attempts in one year; 15 professional development units per three-year cycle to maintain
    Project Management Institute, CAPM certification page checked 2026-08-28
11. 11PL-300 exam duration, pricing basis and renewal requirement 100 minutes to complete; price set by the country or region in which the exam is proctored rather than a single published figure; renewal frequency of 12 months, done through a free assessment on Microsoft Learn
    Microsoft Learn, Microsoft Certified: Power BI Data Analyst Associate checked 2026-08-28
12. 12Ratings of stay-at-home applicants against continuously employed applicants in Weisshaar's companion survey experiment Fielded by YouGov to 1,000 US respondents in January 2014. Opting out produced a penalty of about 0.2 standard deviations on capability, reliability and deservingness (-.187, -.158, -.203) and its largest penalty on perceptions of commitment (-.459 standard deviation units, about 0.46). The opt out effect was significantly more negative than the unemployed effect on commitment, deservingness and reliability
    Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28
13. 13Whether the penalty Weisshaar measured differs by gender In the audit study main effects there were no statistically significant gender differences, with no evidence of a motherhood penalty and no evidence of a fatherhood opting-out penalty. Weisshaar does report gender effects elsewhere: opt out fathers took a larger hit than opt out mothers in the survey experiment, and in competitive local labour markets penalties emerged for continuously employed mothers and were greater still for opt out fathers
    Katherine Weisshaar, From Opt Out to Blocked Out, American Sociological Review 83(1), 2018, free full text hosted by the publisher via the author's open-access eprint link checked 2026-08-28

Related