What Trump should look for in a predictable Supreme Court Justice

President Trump is planning to announce his second Supreme Court appointment of his presidency tonight, filling the seat vacated by outgoing Associate Justice Anthony Kennedy, who announced his retirement the week before last. Trump has whittled down his choices to four finalists, with a top priority reportedly being to avoid selecting a new justice who proves similar to the retired David H. Souter–an Associate Justice appointed by a Republican president who wound up siding with his politically liberal colleagues more often than many had expected.

According to the Wall Street Journal, some in the White House are pushing for a nominee with a deep archive of written legal opinions as proof of their views, while the President himself may be factoring in personal chemistry. If the goal is indeed to appoint a new associate justice to the Court whose future performance will be predictable–particularly in the decision-making domains that jurists are concerned with–here’s what he needs to know.

We can be irrational and consistent simultaneously

For starters, even smart people are biased in their judgments, making rapid or automatic decisions that reflect their subjective biases and preferences. To make matters worse, people rarely change their minds about things, even when presented with contrary facts and evidence. When it comes to decision-making, in other words, reality distortion is the norm–even among the most rational jurists.

However, the biases and flaws polluting our judgment tend to be pretty consistent, meaning that we’re both irrational and predictable. This applies to the mistakes we all make, like ignoring information that is ignoring information that conflicts with our beliefs or falsely assuming that someone is an average or representative member of their group or class. Obviously, these basic psychological tendencies legal profession suffuse the legal profession much as they do every other field, but the good news for the President and his staff is that there often is a discernible pattern to the ways people think, even if their thinking is marred by bias and judgment errors.

Ideology matters, with a caveat

Ideology, including someone’s political views, is generally a strong predictor of their decisions. Like any other moral code, ideology simplifies decision-making by providing higher-order principles to classify events into right or wrong, turning ill-defined problems and ambivalent events into seemingly coherent matters.

However, not everyone is equally ideological, or even ideological at all. In fact, a large proportion of people are more “grey” than “black-and-white” thinkers, meaning they’ll often make ideologically inconsistent decisions–something Trump may want to avoid in choosing a new associate justice. One way to identify a more predictable one, however, is to pick a more ardently ideological jurist. Since people are more likely to think freely when they’re less constrained by ideology, those with more weakly-held ideologies tend to make more nuanced–and less predictable–decisions.

But there’s a critical caveat here: your ability to persuade others that you’re right improves when you’re more ideological–except when the people you’re trying to convince espouse different ideologies. In other words, a very ideologically conservative justice could help bring other conservatives on the Court into line but isn’t likely to persuade more moderate and liberal justices to share their views.

Personality is underrated

President Trump might be on to something in weighing the finalists’ personality traits, which can actually explain systematic differences in people’s reasoning more than many of us might expect.

For example, people with high openness to experience tend to be liberal, intellectual, and curious, which generally results in considering much more data–including uncomfortable facts–in their decision-making. In contrast, people with low openness to experience judge and decide in more authoritative ways are more likely to be conservative. They’re also less likely to question rules and the status quo, focusing instead on preserving norms. Researchers have also found that a person’s humility can predict how well they accept mistakes and take responsibility for bad judgment calls, which arrogant people tend not to do.

Another important factor is personal values. For instance, altruistic people often take into account others’ views and feelings when deciding things, whereas self-centered people are typically more motivated by their own interests and feel less remorse about the potential for their decisions to harm others.

Track records are good guides, but not guarantees

Social science relies on statistics, especially the probability that one event follows another. Even when correlations are strong, there’s no guarantee that a particular outcome will happen. However, when you have enough data on a person, it’s easier to predict what they’ll do most of the time, which includes their default judgments and decision-making patterns. That’s why some of the advisors who are weighing in on the President’s nomination process are emphasizing judges’ track records and prior decisions; it’s a sensible strategy.

Indeed, it’s possible to identify predictive correlations even if there’s not a lot of data to go on. Researchers do this all the time by extrapolating from lookalikes, essentially betting that an individual will probably act like individuals who share some other highly predictive quality, like personality, values, ideology, and so on. Still, this doesn’t change the fact that there’s no scientific formula for being certain about what a given person will decide on a specific instance. Probability–rather than certainty–is as good as it gets.