Policymakers like being consumers of research. Part of their job is to take in new information and work it into the policymaking process. What they have less experience with, however, is being subjects of research. Despite serving as crucial decisionmakers on a range of issues, there are very few systematic studies on how security and development bureaucrats do their jobs on a day-to-day basis. For the next two editions of Deep Dive, we’ll look at some of the research that shows what happens when we turn the microscope around and look at the policymaking community.

There has been a major shift in recent years toward training policy professionals to work with data. If you are a security or development bureaucrat who went to graduate school recently, you almost certainly had to take some data science and coding classes. The theory is that policymakers should be unbiased in their policy evaluations, and that data and data training should reduce bias. But does it actually?

In a new article for the World Bank Economic Review, Sheheryar Banuri, Stefan Dercon, and Varun Gauri show that the answer is, at best, sort of. Banuri, Dercon, and Gauri ran survey experiments with development professionals at the World Bank and the UK’s Department for International Development (DfID) that tested their political bias in evaluating the results of a study and their susceptibility to positive and negative framings in determining a study’s success.

In the political bias test, respondents were asked to look at two studies. One of the studies investigated the effectiveness of a skin cream on treating an unnamed new disease (a less political example) and the other investigated the effectiveness of raising the minimum wage (among economists, a very political example). The respondents were asked to say whether each study supported the policy intervention being tested — that is, prescribing the skin cream or raising the minimum wage. The results of the two studies were the same in terms of numbers of positive outcomes found, but many policymakers perceived them differently. Respondents — many of whom are trained economists — were overall about 20% less likely to accurately describe the outcome of the minimum wage study than the skin cream study.

The framing study was even more stark. Respondents were presented with a treatment for a deadly disease that would definitely save only one-third of all patients versus a treatment that has a one-third chance of saving every patient and a two-thirds chance of not saving anyone. Policymakers were 45% likelier to choose the “safe” choice — definitely saving a third of patients — if it was presented as definitely saving some people rather than as definitely letting most people die. The two options are functionally exactly the same, so the huge discrepancy suggests that policymakers are definitely influenced by how a policy option is framed.

There was some positive news for exponents of the idea that data training reduces bias. Respondents with better math skills were more likely to get the political bias question right, as were respondents with economics degrees. But the overall results suggest that there is still a long way to go, and that policymakers should open themselves up to more research if they want to know if they are living up to their ideals.

The study also seems to expose a funny cultural disparity between the World Bank and DfID. World Bank staff were offered a free coffee mug in exchange for their participation in the study, while DfID staff got nothing for filling out the survey. Yet when results came back, the response rate at the World Bank was 43% and at DfID it was 72%. Either economists don’t respond to incentives the same the world over, or the mugs were ugly.

Critical State’s own Amanda Leigh Lichtenstein traveled to a Midwestern ashram where a guru known as Amma is attempting to change the world through hugging. Amma’s followers value her hugs as a source of divine healing, and the small town of Elburn, Illinois — the site of her seventh North American ashram — values Amma and her followers as a source of economic renewal.

Jon Letman told the story of an American beer that’s in bad taste, and not because it’s an IPA. Manhattan Project Beer Company makes its money selling beer branded with names from nuclear weapons history. None of the beer names are great, but one in particular — “Bikini Atoll” — has drawn the ire of people from the actual Bikini Atoll, a region of the Marshall Islands where the US tested nuclear weapons in the 1940s and 50s. Radiation from the tests caused widespread illness among people living in the area, who had no say in whether or when the tests took place.

Monica Campbell spoke to migrants trying to enter the US who are stuck waiting in Tijuana, Mexico, while US courts consider their asylum claims. Asylum-seekers used to be able to remain in the US while their claims were considered, but a Trump administration policy has forced more than 14,000 asylum-seekers to remain in Mexico while the court system slowly does its work.

You know that thing where before and after photos show how quickly being president can age a person? That, but for being CENTCOM commander — in charge of US military forces in the Middle East — during the Trump administration.

The bond curve inverted last week, meaning the market believes a recession is imminent. A lot of people are alarmed, but at least someone is into it.

Pro-government Chinese rappers are always on call to lend their talents to the Chinese state. They’ve weighed in on the escalating crisis in Hong Kong, and the power of their artistic voice has staggered observers.

As part of its long-running effort to be the parody version of itself, the US military uses the simple mnemonic DOTMLPF-P to remember the elements of operational planning: Doctrine, Organization, Training, Materiel, Leadership and education, Personnel, Facilities, and Policy. Anyway, here’s Madonna.

As an addendum, the tendency for militaries to become the parody version of themselves isn’t limited to the US. The British version of DOTMLPF-P is, incredibly, TEPID OIL.

The world’s greatest tennis player is now also its leading drone countermeasure.

If you’re looking for something to commemorate the enlistment of the first uniformed Americans born after 9/11, check out this haunted 2001 McDonald’s Easter plate, available on eBay.