Thus it may come as a surprise that the first conspiracy theory about the unemployment rate, which emerged in the 1960s, was that the Labor Department was seeking to artificially inflate unemployment—to advance communist objectives, naturally.

Ironically, the claims largely revolved around efforts to measure “discouraged” people who weren’t actively seeking work. Today, the exclusion of the discouraged is often presented as Exhibit A that the numbers are manipulated. The original theory held that their inclusion was the method of manipulation.

First, some historical context on unemployment in the 1950s and 1960s when the conspiracy emerged.

The unemployment rate was relatively new. Then, as now, the survey was conducted by the Census Bureau and the data tabulated and analyzed by the Bureau of Labor Statistics, a largely independent Labor Department agency. Since 1916, the BLS had produced measures of how many people worked in factories, for example, but during the Great Depression two things became obvious: 1) many people were seeking work and couldn’t find it and 2) nobody had a clue how many.

Everyone recognized it would be useless to simply measure unemployment by taking the total population, and subtracting the number of employed people. Many women, many elderly, many students, and at least some injured or ill men weren’t working, nor were they trying to. What needed measurement was the vast pool of job seekers. So in the 1930s, the Census Bureau developed the still-novel technique of random sampling to survey how many people were without work and seeking it.

By the 1950s, the BLS’s monthly report was gaining prominence. As happenstance would have it, the jobless rate was low or falling during the 1954, 1956 and 1958 elections. The figures were published around the 10th of each month, but when President Dwight Eisenhower learned each year that favorable numbers for October were set to come out, he announced the figures in advance of the election, and ahead of their typical publication schedule.

But 1960 was different. Unemployment was rising that year, and the Labor Department announced no figures would be available before the election. The politics were obvious. Democratic National Chairman Henry “Scoop” Jackson told the Washington Post:

“This is one more flagrant case of the administration holding back distasteful information. Those employment reports belong to the people of this country. They pay for them. They are not the private property of the Republican Party to use or not use for political purposes.”

Then, in a scoop today’s reporters can only dream of, Post reporter Bernard Nossiter obtained a leak. On Nov. 5, 1960, three days before the election, the Washington Post published those leaked BLS figures on its front page. And they were bad. Unemployment climbed by 200,000 and the unemployment rate skyrocketed to 6.4% from 5.7% in one month. After the election, the official report was released. The scoop was correct.

ENLARGE

The story that ran on the front page of the Washington Post on Nov. 5, 1960.

There’s no evidence Mr. Eisenhower sought to manipulate the numbers, only to delay their release, but the figure was now politicized. Charges of manipulation would come soon.

It was Reader’s Digest that looked into the details of how unemployment was calculated, and smelled commies. A September 1961 article, heavy with scare quotes, began:

“For months the U.S. Bureau of Labor Statistics has been pouring out a stream of doleful figures depicting the worst ‘unemployment crisis’ in the United States since the Great Depression of the 1930′s. Almost daily some administration official tells us that nearly seven percent of our labor force is out of work. Meanwhile, Congress has passed one emergency spending bill after another on the ground, in part or in whole, that it will help employment…. All this unemployment news out of Washington provides excellent fodder for the communist line, of course.”

The article was an unexpected national sensation. Judith Innes, a professor at the University of California, Berkeley, who studied the development of public policy indicators wrote: “though academic discussions had considered these issues, this was the first commentary to appear in a mass magazine on the implications of measurement decisions.” Despite a “seemingly dry topic,” Reader’s Digest sparked a wave of “articles and editorials nationwide, and a massive official evaluation of unemployment data.” The Joint Economic Committee held hearings. A presidential committee was appointed to investigate.

The Digest article reserved particular scorn that people who voluntarily quit their jobs “from dissatisfaction and have become floaters” could be counted as unemployed and lamented the “bureaucratic compulsion to classify people as unemployed.” The article suggested real unemployment was much lower and explicitly rejected the idea that discouragement was a valid reason to stop a job hunt:

“Originally, a person was classed as ‘looking for work’ only if he actually looked. Then, in the 1930s, it was decided that this was unfair to non-lookers, because they were discouraged by the ‘lack of job opportunities.’ In 1940, therefore, Washington started classifying anybody who wanted a job as looking for work, even if he said he had not looked because he was sick or ‘guessed’ that no work was available.”

This is precisely the opposite of today’s theories: that the BLS is deliberately and dishonestly suppressing people that give up looking for work. It was also mistaken. Though its questions were not as precise as today, the Labor Department sought to measure those actively searching for work. Those not actively searching were considered a distinct class of discouraged people.

An economist named Robert A. Gordon was asked to lead a committee, including academics and representatives of industry, to review the statistics and the Reader’s Digest claims.

In testimony before the Joint Economic Committee he concluded:

“You will forgive me if I say this article represented an egregious example of irresponsible journalism….After careful investigation, the Committee has unanimously and categorically concluded that doubt concerning the scientific objectivity of the agencies responsible for collecting, processing, and publishing employment and unemployment statistics is unwarranted.”

Despite this exoneration, the Gordon Committee said “that does not mean that the data cannot be improved” and suggested the BLS refine its questions to make them less “subjective.” They also recommended policies (which were subsequently adopted) to prevent the release of the numbers being sped up or delayed for political purposes, as Mr. Eisenhower had done. That’s why today each jobs report is first released as a dry technical summary by the BLS, with no commentary from the White House.

In response to the Gordon Committee report, the BLS began publishing in 1967 a quarterly report that tracked discouraged workers. In 1994, the BLS began publishing the data monthly in the form of alternate measures of the unemployment rate. (We publish these alternate figures monthly in our roundup of the jobs report in charts.)

Much can change and be forgotten in a half century. Outrage that anyone would ever consider subjectively “discouraged” people to be unemployed, gradually turned into outrage that anyone could not consider them unemployed. And throughout it all, the BLS has sought to measure unemployment the same way.