It turns out that a young Max and a middle-aged Max can get away with saying things an old Max cannot.

At least that is the conclusion of a new study by Princeton University researchers aimed at measuring age discrimination, one of the toughest forms of workplace bias to prove.

The subjects of the experiment – 137 Princeton undergrads – were shown a video of a man who would be their partner in a trivia contest. His name was Max – white, neither handsome nor ugly, wore a checked shirt and said he was from Hamilton, New Jersey.

What the students did not know was that there were actually three versions of Max, being played by different actors, a 25-, 45- and 75-year-old.

Each Max followed the same script with one exception. When describing himself, half of the time, Max said he was the kind of person to share his wealth with relatives (the compliant Max); and the other half of the time, Max said he felt no obligation to share (the assertive Max).

The students were then asked their opinion of Max. For those who saw the 25- or 45-year-old Max, it made no difference whether he was compliant or assertive. But students who saw the 75-year-old actor gave the assertive Max a high negative rating.

The results, soon to be published in the Personality and Social Psychology Bulletin, illustrate the subtle bias older people may face in the workforce.

“If you want to be an ageing grey panther, and speak your mind to your manager, that’s fine,” said Susan Fiske, a Princeton professor and a co-author of the study with Michael North. “But expect consequences.”

Age discrimination in the workplace has always been harder to identify and quantify than race and sex bias. Blacks and women have experienced a long history of being underpaid, which researchers can calculate in salary differentials. The maths is less straightforward for older workers. They may have been unfairly demoted, and yet still earn more than their younger co-workers.

There is little doubt that such discrimination exists. When an older man or woman is laid off, it typically takes two to six months longer to find a new job than it takes younger workers, according to the Bureau of Labor Statistics. And the new job is likely to pay considerably less.

During the recent recession, many unemployed older people told a similar story. They sent in their résumé and got called for an interview, but when they walked in, potential employers saw their white hair and that was it.

Feeling discrimination is one thing, proving it another. “It’s simply harder to establish,” said David Neumark, a professor at the University of California, Irvine.

Winning an age discrimination lawsuit has also become harder since a 2009 US Supreme Court case, Gross v FBL Financial Services. Before that, the employee had to show that age was a factor contributing to the layoff. These days, the staff has to show that age was the determining factor for the layoff.

The older generation, those born from 1946 to 1964, accounts for the fastest-growing segment of workplace discrimination claims in the US. In 2012, 22,875 people filed age claims with the federal Equal Employment Opportunity Commission, compared with 15,875 in 1997. That represents 23 per cent of all the individuals filing claims in 2012 versus 19 per cent in 1997.

At the same time, the percentage of people filing race claims has decreased to 33 per cent of all claimants, from 36 per cent; it has held steady at 30 per cent for sex discrimination.

With age discrimination claims on the rise, a growing number of academics are undertaking research projects aimed at better identifying it.

In 2010, the Alfred P Sloan Foundation began a “Working Longer” project, which has awarded 66 research grants focused on age discrimination, including the Princeton study. Kathleen Christensen, the programme director, says it was spurred by a greying population that will be healthier and stay in the workforce longer.

 “Anything that involves Baby Boomers assumes greater significance just because of the sheer numbers,” she said.

One grantee, Sarah von Schrader, a Cornell research associate, says more than 60 per cent of equal employment complaints are related to job dismissals; only 10 per cent involve hiring bias, which, she says, is the most difficult of age cases to document.

Several grant recipients are working to create a model for identifying age bias. One way researchers have been able to measure race and gender bias is to send identical fictional résumés to employers – except that one applicant is black, the other white; or one female, the other male. They then compare – or audit – the success rates of whites versus blacks or men versus women.

However, the audit model is less effective for detecting age bias. Identical résumés for a 25- and 55-year-old would raise concerns for employers. A 25-year-old with three years’ work experience makes sense; a 55-year-old with the same experience would be highly suspect.

Joanna Lahey, an associate professor at Texas A&M, has created an innovative – though limited – audit model for calculating age bias. In a 2005 study, she sent out 4,000 résumés to businesses in Boston and St Petersburg, Florida, for fictional job applicants from 35 to 62.

The applications were for entry-level fields such as data processing and fast-food service – the sort of jobs requiring little work experience. The applicants were all women, since employers might plausibly assume an older woman “has been at home taking care of the family”.

Lahey found that a younger applicant was 40 per cent more likely to be called for an interview than someone 50 or older. In Boston, the younger person needed to send 19 résumés to get an interview; for older workers, it took 27 résumés ; in Florida, the comparable numbers were 16 versus 23 résumés.

As for the three Maxes, the Princeton researchers found there actually was a way to reduce the old Max factor, at least in laboratory conditions. In a second experiment, the subjects read a newspaper article predicting that there would be enough resources to go around between the generations as the population aged.

“Suddenly, the bias effect disappeared,” North said.

The New York Times