White House press secretary Sarah Huckabee Sanders smiles as she speaks to the media during the daily press briefing at the White House, Tuesday, Aug. 14, 2018, in Washington. (AP Photo/Andrew Harnik)

The Council of Economic Advisers, a three-person team near the center of the executive branch, was created in 1948 to, among other things, “gather timely and authoritative information concerning economic developments and economic trends” and to “analyze and interpret such information” to assess how policy affects the economy. Its members were meant to be “exceptionally qualified to analyze and interpret economic developments” in order to provide the president with objective, unflinching data and insight.

That process broke on Tuesday.

During her daily press briefing, press secretary Sarah Huckabee Sanders attempted to demonstrate President Trump’s commitment to the black community by sharing some numbers about job growth during his presidency.

“This president, since he took office, in the year and a half that he’s been here, has created 700,000 new jobs for African Americans,” she said. “When President Obama left after eight years in office — eight years in office — he had only created 195,000 jobs for African Americans. President Trump in his first year and a half has already tripled what President Obama did in eight years.”

Those numbers, as we noted, were very wrong. Between January 2009 and January 2017, the number of employed black Americans rose from 15.5 million to 18.4 million, an increase of nearly 3 million. Under Trump, the increase has been 708,000. Trump has been in office for only 20 months, of course, but the point stands: Sanders’s numbers were way off the mark.

Where did they come from? In a tweet, the Council of Economic Advisers owned up to giving her bad data.

Sanders apparently subtracted 636,000 from 831,000 to get 195,000. She didn’t notice that the CEA numbers were including only the first 20 months after a presidential election. Which, you will notice, is a weird way to do this.

The CEA is comparing these three periods:

There are a few reasons this is a questionable way to approach the issue.

The first is that it includes a period during which Barack Obama wasn’t president and attributes a period in which Obama was president to Trump. Those are the lighter-colored months, the periods between the November elections and each new president’s inauguration. The effect of this is to make Obama look worse and Trump better. In those months attributed to Obama, 162,000 fewer black Americans were employed. During those months attributed to Trump, 140,000 more were. Setting aside the very valid question of how much a president affects job growth anyway, Trump’s objective economic team is giving him credit for a period when he wasn’t even president.

They are also saddling Obama with three additional post-recession months (hence the sharp drop in employment). That’s the main reason that the first 20 months after Obama’s 2008 election were so mediocre in job growth, of course: the recession that only began to recede in 2010.

These are the periods the CEA is counting.

Shortly after that first period ended, black employment (and overall employment) began a steady climb.

Notice that in the second period indicated by the CEA, black employment increases were nearly the same as those attributed to Trump: 831,000 vs. 848,000. If we take out the two transitional months from election to inauguration, Trump actually trails, 731,000 to 708,000. Giving Trump a few months from Obama’s tenure helps him do better. Notably, even Sanders saw through this gimmick, sharing only that 708,000-increase figure and not the CEA’s inflated tally.

Sanders, Trump’s spokeswoman, was more objective than the CEA.

That doesn’t excuse her broader point. Obviously, these numbers are a bad way to assess a president’s views on race, Sanders’s goal in releasing the numbers. Less obviously, it’s a bad way to evaluate a president’s performance on economic issues, given how little control a president has over the economy.

But something that may not be immediately obvious bears pointing out: This is precisely the sort of subjective rhetoric-padding that the CEA was supposed to avoid. This is not an objective look at how presidents are faring on jobs issues; it’s data meant to prove that Trump is better than Obama, even though, on this metric, he hasn’t been.

We extend the point to ask this question: Are the data and advice the CEA is giving Trump in private less dependent on proving what he wants to hear?