Callback rates for older job applicants, and resume changes that raise them

If you're over 50 and sending out applications that vanish, you can't tell from your own inbox whether it's the market, the resume, or your age. One person's search is too small a sample. Researchers have run the bigger version: thousands of applications built to be comparable apart from age, and a count of who gets called back. This article walks through what they found, where in hiring the gap opens, and which resume changes made a measurable difference. We use these studies at FirstSourceTeam because they replace guesswork with counts, and because they tell people over 50 that the silence isn't all about them.

Economists David Neumark, Ian Burn and Patrick Button sent 40,223 fictitious job applications to 13,371 real job ads in 12 US cities. Each ad got a set of comparable resumes, with age signalled by one line: the year the applicant finished high school. Applicants who looked 29 to 31 got a callback 18.69% of the time. Those who looked 49 to 51 got 15.40%. Those who looked 64 to 66 got 12.16%, a rate 35% lower than the youngest group's.

That study first circulated as an NBER working paper in October 2015 and was published in the Journal of Political Economy in 2019. The newer work below, most of it by the same group of economists, looks at where in the hiring process the gap opens and at the wording of job ads. Every number here comes from the papers' own tables.

Callback rates by applicant age

Note: 40,223 applications to 13,371 job ads. Age was signalled by high school graduation year. All differences between age groups significant at p < 0.01 (Fisher's exact test).

Exhibit 1. Source: Neumark, Burn and Button, "Is It Harder for Older Workers to Find Jobs? New and Improved Evidence from a Field Experiment," NBER Working Paper 21669 (rev. 2017), Table 4, Panel A; published in Journal of Political Economy 127(2), 2019

A field experiment of this kind is called a correspondence study. Researchers write realistic resumes, assign ages at random, apply to live postings and count who gets called back. Because age is the only thing that differs on average, a gap in callbacks is a measure of how employers respond to age.

Neumark, Burn and Button built their resumes from real ones scraped from a job site, so the older applicants would look like real older applicants. Most of the older resumes carried experience that matched their age; some were trimmed on purpose to test whether that changed anything, and the combined rates above include both kinds. All differences between the three age groups were significant at the 1% level.

The first thing we look at on a client's resume is the bottom of it, because that's where the dates sit. The jobs were administrative assistant, retail sales, security guard and janitor. None of the field experiments in this article tested manager, director or executive roles, so read the numbers as the direction and rough size of the age effect at the screening stage, not as a forecast for a senior search.

Callback rates by age and gender

29 to 3149 to 5164 to 66

Note: Sample sizes: administrative 24,350; sales men 5,348; sales women 4,707; security 4,138; janitor 1,680. For security, no age difference is significant at the 5% level.

Exhibit 2. Source: Neumark, Burn and Button, NBER Working Paper 21669 (rev. 2017), Table 4, Panels B to F

If you're a woman in your late 50s or 60s and the search has felt harder than it did for male peers, the data backs you up. The gap was widest for women. In administrative jobs, women aged 64 to 66 got a callback 7.58% of the time against 14.41% for women aged 29 to 31, just over half the rate. In retail sales, the oldest women got 18.43% against 28.68%.

For men the picture was mixed. Older men in sales got 14.70% against 20.89%, and older janitors 25.89% against 32.08%. In security jobs, the callback rates for men of all three ages were close (24.28%, 21.55% and 21.74%), and none of the differences was significant at the 5% level. The authors report consistent evidence of age discrimination against older women, "especially those near retirement age," and "considerably less evidence" against men once they corrected for statistical biases found in earlier studies.

Applications needed per interview, by age

Joanna Lahey ran an earlier experiment, published as a working paper in 2005, for entry-level jobs in the Boston area and around St. Petersburg, Florida, sending pairs of resumes for women aged 35 to 62 to 3,996 firms. Her working paper turns the gap into a count of applications.

Area and response	Younger (35, 45)	Older (50, 55, 62)
Massachusetts, any callback	11	13
Massachusetts, interview request	19	27
Florida, any callback	9	11
Florida, interview request	16	23

Note: Paired resumes sent to 3,996 firms for entry-level jobs. Figures are from the working-paper version.

Exhibit 3. Source: Lahey, "Age, Women, and Hiring: An Experimental Study," NBER Working Paper 11435 (2005), section on results, pages 18 to 19; published in Journal of Human Resources 43(1), 2008

In Massachusetts, a younger applicant needed 19 applications to get one interview request and an older applicant needed 27. In Florida the numbers were 16 and 23. Put as rates, a younger applicant was 42% more likely to get an interview request in Massachusetts and 46% more likely in Florida. For any callback the gap was smaller, 19% and 18%, so more of the difference showed up at the step where an employer commits to meeting someone. These percentages are noisy: Lahey's own confidence ranges for the interview figures run from about zero to more than double, though her main statistical test for older applicants was significant.

Age bias at the interview stage

The most recent evidence comes from a different kind of data. In a 2024 Journal of Human Resources paper, Neumark used hiring records from a restaurant chain, drawn from an age discrimination lawsuit against it. The chain switched from paper applications in the store to online applications during the period studied. On paper, the manager saw the applicant's age along with everything else. Online, the first cut for interviews was made from a detailed application that didn't show age, and age only became visible at the interview.

Note: Under-40 baseline rates: 14.1% job offers on paper applications; 33.4% interview selection and 30.0% offers after interview on online applications. The -1.2% interview-selection difference is not statistically significant.

Exhibit 4. Source: Neumark, "Age Discrimination in Hiring: Evidence from Age-Blind vs. Non-Age-Blind Hiring Procedures," NBER Working Paper 26623 (2020), Table 2 column 2 and Table 3 columns 3 and 4; published in Journal of Human Resources 59(1), 2024

With paper applications, people aged 40 and over got job offers at a rate 67.9% below the 14.1% rate for younger applicants. With online applications, the age-blind screen picked older applicants for interviews at about the same rate as younger ones; the 1.2% gap with full controls isn't statistically significant. After the interview, older applicants' offer rate was 42.8% below the 30.0% rate for younger interviewees. Neumark calls this "strongly consistent with age discrimination in hiring," since the gap appears only once the hiring manager can see the applicant's age.

Many of the people we coach describe exactly this: a strong first screen, a friendly interview, and then nothing. It's one employer in one industry, and the jobs were restaurant jobs. Lahey's numbers lean the same way: her age gap was larger for interview requests than for any callback.

Resume changes tested in the studies

The big experiment also varied the resumes themselves, so some common resume choices were tested directly.

Applicant and resume type	Change in callback probability (points)
49 to 51, experience matching age	-3.3
49 to 51, experience matching age, applying to a lower-level bridge job	-2.7
49 to 51, experience trimmed to match the young applicant	-3.5
64 to 66, experience matching age	-5.8
64 to 66, experience matching age, applying to a bridge job	-5.4
64 to 66, experience trimmed to match the young applicant	-6.2

Note: Probit marginal effects. Callback rate for the young group is 18.69%. Trimmed versus full experience: p = 0.80 (49 to 51) and p = 0.56 (64 to 66), so no detectable difference. The 64 to 66 bridge row is the SHB(L) resume type in the paper.

Exhibit 5. Source: Neumark, Burn and Button, NBER Working Paper 21669 (rev. 2017), Table 6, column 1 (combined occupations, full controls)

Trimming experience didn't help, which is why we push back when people tell older job seekers to cut everything before 2010. Neumark, Burn and Button sent some older applicants with work histories cut to match the young applicant's. For the 64 to 66 group, the callback penalty was 5.8 points with full experience and 6.2 points with trimmed experience, and the authors couldn't tell the two apart (p = 0.56). Presenting the application as a step down into a less demanding "bridge" job didn't change the penalty either.

Listing computer skills helped older applicants a little more than younger ones. For administrative jobs, listing computer skills raised older applicants' callback probability by 3.4 points more than it raised younger applicants'. The authors note the overall age gap in those jobs was 9 points, so the skill line narrows the gap without closing it. A typo-free resume is weaker evidence. The paper's estimate for older administrative applicants was 3.1 points better than for younger ones, but its estimate for younger applicants was oddly negative (minus 1.9 points), so the net gain for older applicants was small. College degrees, volunteer work and typing speed tended to do less for older applicants, though only weakly, and not significantly for volunteer work.

Job-ad wording predicts the gap. Burn, Button, Munguia Corella and Neumark scored the text of the job ads from the big experiment against stereotypes of older workers, using machine-learning methods. For men, ads with more language tied to physical ability, memory or experience were linked to more discrimination. For janitor jobs, men aged 49 to 51 were passed over while the younger applicant got a callback on 9.6% of ads, and ads scoring one standard deviation higher on physical-ability language were associated with 3.8 points more. A follow-up experiment by Burn, Firoozi, Ladd and Neumark posted ads they wrote, with and without that kind of language, and found it deterred older people from applying at all. The authors conclude the deterrent effect can be "roughly as large" as direct discrimination in hiring.

State law made a small and uncertain difference. A 50-state experiment for retail sales jobs found "some evidence" of less discrimination in states whose age discrimination laws allow larger damages, but it didn't hold up in every test. Stronger laws did not make employers less willing to hire older workers.

Recommended changes to your job search

The studies point to a few concrete moves, each limited by what was tested.

The age signal in the Neumark, Burn and Button experiment was the high school graduation year, and 56% of the real resumes the researchers collected listed one. Leaving the year off removes the cue employers in the experiment reacted to. None of these studies measured what happens to callbacks when it's removed, so treat that as a reasonable step, not a tested one.

Name your current tools and software on the resume. In administrative jobs, computer skills were the clearest item that did measurably more for older applicants than younger ones. Proofread too, though the evidence on typos is weaker.

Read the ad's language before deciding how to spend an application. The research links stereotype-heavy wording with more discrimination against older men, and the ad text is free to read.

Plan for the interview as the hard step. In our coaching, interview practice is where senior people most often underinvest, because they've been on the other side of the table for years. In the restaurant chain's data, older applicants passed the age-blind written screen at the same rate as younger ones and fell behind only after meeting a manager. Interview practice is where more of the effort can go.

Two resume headers that differ only by graduation year

Name (left)

Karen M. Olson

Name (right)

Linda J. Olson

Target role (both)

Administrative Assistant

Most recent job (both)

Office Coordinator, Lakeside Property Group, 2011 to present

Skills line (both)

Microsoft Office, 65 words per minute, scheduling, vendor billing

Education (left)

Roosevelt High School, Diploma, 2002

Education (right)

Roosevelt High School, Diploma, 1968

Margin note

56% of the real resumes the researchers scraped listed a high school graduation year.

Refusing to hire someone because they're 40 or over is illegal under the Age Discrimination in Employment Act, which covers hiring at employers with 20 or more workers, per the EEOC. These experiments measure the pattern across thousands of applications; showing it happened in one hiring decision is much harder. If you think a specific employer turned you down because of your age, an employment lawyer can tell you whether the facts support a charge and which filing deadline applies to you.

Working with FirstSourceTeam

At FirstSourceTeam, a family-owned firm in Minneapolis, we've coached job seekers since 2006, and we see this situation often. We usually start with the resume's first third and the dates that give age away, then rebuild the LinkedIn profile to match. Next we work on warm introductions to recruiters and hiring managers, and on interview preparation, since that's where the gap reopens. Where it fits, we help you think through severance, benefits and retirement timing too.

If you want a low-key first step, our free one-hour webinar runs every Wednesday at 11 AM Central with John Schrup and Misty Schnuelle.

Sources

1. David Neumark, Ian Burn and Patrick Button, National Bureau of Economic Research, Is It Harder for Older Workers to Find Jobs? New and Improved Evidence from a Field Experiment (NBER Working Paper 21669; Journal of Political Economy 127(2), 2019), accessed Oct 12, 2026.
2. Joanna N. Lahey, National Bureau of Economic Research, Age, Women, and Hiring: An Experimental Study (NBER Working Paper 11435; Journal of Human Resources 43(1), 2008), accessed Oct 12, 2026.
3. David Neumark, National Bureau of Economic Research, Age Discrimination in Hiring: Evidence from Age-Blind vs. Non-Age-Blind Hiring Procedures (NBER Working Paper 26623; Journal of Human Resources 59(1), 2024), accessed Oct 12, 2026.
4. Ian Burn, Patrick Button, Luis Felipe Munguia Corella and David Neumark, National Bureau of Economic Research, Older Workers Need Not Apply? Ageist Language in Job Ads and Age Discrimination in Hiring (NBER Working Paper 26552), accessed Oct 12, 2026.
5. Ian Burn, Daniel Firoozi, Daniel Ladd and David Neumark, National Bureau of Economic Research, Help Really Wanted? The Impact of Age Stereotypes in Job Ads on Applications from Older Workers (NBER Working Paper 30287), accessed Oct 12, 2026.
6. David Neumark, Ian Burn, Patrick Button and Nanneh Chehras, National Bureau of Economic Research, Do State Laws Protecting Older Workers from Discrimination Reduce Age Discrimination in Hiring? Evidence from a Field Experiment (NBER Working Paper 25369; Journal of Law and Economics 62(2), 2019), accessed Oct 12, 2026.
7. US Equal Employment Opportunity Commission, Age Discrimination, accessed Oct 12, 2026.