I suspect that most Republicans must have immediately “known” that the breastfeeding story was “fake news,” and that most Democrats were immediately sure it was true.

Late last week, the press reported the horrible story of border cops snatching a baby being breastfed by her illegal-immigrant mother. The story however may not be true. The New York Times calls it a mere “anecdote.” U.S. Customs and Border Protection as well as the White House denied it. President Trump, who introduced the “zero tolerance” policy that would be responsible for this outrage, blamed “the democrats.”

The original source of the story was apparently CNN, who quoted an attorney from the Texas Civil Rights Project, who was himself reporting what the mother had told him. The mother is unreachable and is probably in jail. In short, there is no independent, verifiable evidence that the event ever happened—even if the administration revealed that 1,995 children were separated from their illegal-immigrant parents over a six-week period. Attorney General Jeff Sessions views these parents as guilty of “smuggling a child” (besides entering illegally in the country).

All this reminds us of the need to check the veracity of news stories. Looking at the incentives of the individuals involved is a first step. Incentives are the bread and butter of economists. If something is not incentive-compatible, assume it did not happen. So what are the incentives here?

On the one hand, the mother may have had an incentive to exaggerate her story; the same is true for the attorney who was trying to defend her. (I say that the mother may have had the incentive to exaggerate because it may also be prudent for her to lie low before a potentially vindictive bureaucracy.) A popular media outlet also has an incentive to dramatize a story to catch the attention of its readers or flatter their prejudice.

On the other hand, the officialdom’s incentives are to hide, ignore, or deny such a story. (We are lucky that their incentives are not yet to be proud of it.) These reactions would be normal from politicians in the White House and the cops who are responsible or complicit. Christopher Browning’s troubling book Ordinary Men shows how low-level agents—ordinary men—are motivated, even if only by mild group pressure, to go along with the team.

Thus, not all the incentives of the main actors point to the same direction. Except if one wants to do his own investigation, a high-cost venture, one has no alternative but to rely on public or private information sources. The frontier between public and private sources gets blurred if we factor in activist websites or social media accounts where the goal is propaganda as opposed to the search for truth. So the credibility of information sources is important. And—this is my main point—credibility itself depends on incentives.

Other things equal, an established media with strict investigation and reporting standards is more credible than a hairy propagandist. A media outlet whose reputation has more commercial value faces a more pressing incentive to verify the information it publishes. At least as a news source, the New York Times has a higher market value than Breitbart, and the owners of the former will protect their credibility more than will those of the latter. It is remarkable that, despite its known bias, CNN does mention its source with enough information for the reader to see it is not especially solid. The New York Times, as we saw, is even more prudent.

Pursuing in this vein, one should trust the financial press more than popular media, because the former sells money-making information rather than, or more than, entertainment or confirmation bias. (I haven’t seen the breastfeeding story in the Wall Street Journal.)

This case also reminds us to beware of the confirmation bias—which consists in focusing on information that confirms one’s preconceptions. I suspect that most Republicans must have immediately “known” that the breastfeeding story was “fake news,” and that most Democrats were immediately sure it was true. Thinking by oneself is important, and even more so when something seems to confirm what one already believes.

From what is known at this stage, then, the breastfeeding story should be considered false even if, in the current state of America alas, it is not altogether inconceivable.