On Thursday, the government released a flood of data about the stimulus, showing how 9,000 federal contractors spent their stimulus dollars — including the value of the contract, each project’s status, and how much each of the contractor’s five highest-paid officers were paid.

But when it came to presenting that data, Recovery.gov, the government’s official site for stimulus information, highlighted one number in particular, posting it on the site’s main page in large font: "JOBS CREATED/SAVED AS REPORTED BY FEDERAL CONTRACT RECIPIENTS: 30,383." To make extra certain of getting viewers’ attention, the number itself appears in bright green.

As the economy continues to shed jobs, it’s easy to see why the administration is keen to highlight the number of jobs created by the stimulus. When the numbers were released, Jared Bernstein, the administration’s chief economist, said the job count "exceeds our projections," adding that it supported the conclusion "that the Recovery Act did indeed create or save about 1 million jobs in its first seven months."

But do the 30,000 jobs represent a good return? And since the federal contracts for which data was reported this week represent just a sliver of the overall stimulus package, what do they really say about the impact of the stimulus as a whole?

Let’s start with the 30,000 jobs themselves. The federal contracts in question represented $16 billion in stimulus spending. Assuming the number of created or saved jobs reported by each contract recipient was accurate—which, as we’ve reported before, is still an open question—that breaks down to $533,000 for each job. That’s more than five times the projection of the president’s own Council of Economic Advisers, which estimated in May that every $92,136 in government spending would create one job for one year.

Five hundred thousand dollars per job might sound like a lot of money, but wait: The data released this week covers only the jobs directly created by federal stimulus money so far. It doesn’t cover indirect jobs — the people who make the materials that contractors need to complete their project, or make the sandwich when a construction worker buys lunch from the proverbial roadside diner.

So, if the $16 billion in federal stimulus contracts generated 30,383 direct jobs, how many indirect jobs were created or saved? We asked the White House, which told us they believe that for each direct job created or saved, there is one indirect job. Assuming that’s right, that $16 billion created or saved some 60,000 jobs — which still clocks in at $267,000 per person.

What about the second question — the relationship between the 30,000 direct jobs reportedly created or saved by that $16 billion, and the job impact of the stimulus as a whole? We asked the White House for the logic behind Bernstein’s statement that this week’s numbers "point to the conclusion that the Recovery Act did indeed create or save about 1 million jobs in its first seven months."

Their response: Because the $16 billion in federal contracts represents about 5 percent of the $339 billion spent so far, they multiplied the 30,000 jobs by 20. The result is 600,000 direct jobs; and, relying again on the assumption that each direct job produces one indirect job, the White House doubled that number to 1.2 million.

Of course, that assumes that for every part of the stimulus will have roughly the same job-creating impact as federal contracts. That’s contradicted by the Council of Economic Advisers’ own report, which said that while every $92,136 in government spending creates one job for one year, it takes $145,351 in tax cuts to achieve the same result. As much as 28 percent of the stimulus is going to tax cuts; if the Council’s estimates are right, then the White House’s assertion — that every part of the stimulus will produce the same job impact as federal contracts — starts to look a little less certain.

The bottom line, it seems, is that knowing for certain how many jobs the stimulus is producing remains, for now, as tricky as ever.