Writing in the New York Times in November 2003, Austan Goolsbee, then a professor at the University of Chicago, flamboyantly accused the government of "cooking" the books regarding unemployment.

"The situation has grown so dire," he said, "that we can't even tell whether the job market is recovering. The time has come to correct the official unemployment statistics to account for those left out."

Professor Goolsbee is now a top economic advisor to President Obama. Would he admit that the official jobless of 9.5% grossly underestimates the pain of job losses in America and do something to correct the situation?

Findings in recent IBD/TIPP polls suggest that now would be a good time to undertake such a project.

According to Labor Department data, the civilian labor force in June totaled 153.7 million people, 14.6 million (9.5%) of whom were unemployed. But in the latest IBD/TIPP poll conducted last week, 28.6% of respondents said at least one member of their household is unemployed and looking for work. This number for June was 27.8% and for May 28%.

When we project our household job-seekers rate and calculate the share of Americans who are unemployed and looking for work, we get a job-seeker rate of 24.1% for July for a total of 37 million Americans vs. the government's aforementioned 14.6 million.

The difference between our crude job-seeker rate of 24.1% and the Labor Department's jobless rate of 9.5% is night and day. The difference between our job-seekers (37 million) and Labor's unemployed (14.6 million) is a staggering 22.4 million. How does one account for 22 million people? Which is the reality?

To come up with its monthly jobless rate, the Labor Department surveys some 60,000 households. But the popular unemployment measure that results, dubbed U-3, is not a good indicator, because the department counts as "unemployed" only those who report actively looking for work in the past four weeks.

Some 8.6 million Americans responding to a tight job market have taken part-time jobs, though they'd like to be working full-time. In calculating U-3, the government counts these "underemployed" people as "employed," which helps reduce the unemployment rate.

Another 6.5 million individuals in the government survey say they want a job, but they are counted as not being part of the labor force ("persons not in the labor force but who currently want a job"). This category includes 2.6 million who are marginally attached persons. Of these passive job-seekers, some have other obligations but would take a job if offered.

True, the Labor Department puts out a broader unemployment gauge, known as U-6, that includes "underemployed" and "discouraged." For June, it stood at 16.5%. This alternative measure has been in existence since the mid-'90s and is comparable to methods used by Canada, Japan and Western Europe. But even U-6 doesn't capture the nation's real pain.

The underemployed (8.6 million) and those not in the labor force but who want a job (6.5 million) bring down our difference with the Labor Department from 22 million to 7 million.

Labor also has a large category of 18 million workers classified as "part-time not for economic reasons." These are people who the department believes have voluntarily opted for part-time employment. If we assume 7 million of the 18 million voluntary part-timers are looking for full-time employment, we can reconcile the statistics globally.

Americans have lived through a deep recession and are feeling real pain. But if we continue to measure that pain by the government's official unemployment rate, we are living in denial. It's time to come up with a statistical model that tells the truth.

• Mayur is president of TechnoMetrica Market Intelligence, which directs the IBD/TIPP Poll that was the most accurate in the last two presidential elections.