My experience says “harder,” and that experience is reinforced to some extent, but also challenged to some extent, by an analysis conducted by researchers from UCLA and the Max Planck Institute. Steven Clayman and Mathew Fox (2025) conducted a review of video from more than 100 question and answer sequences between judges and prospective jurors in four criminal trials in Arizona and one civil trial in the federal district of Northern California. 

The researchers’ question: Given that we know the pressures for “social desirability” in voir dire, or the tendency to give the perceived “correct” or “positive” answers, and the consequent under-reporting of stigmatized attitudes that actually would be harmful to one party or the other, how are judges reacting to that? The research is descriptive rather than prescriptive, but it does point to three broad tendencies observed in the analysis, with two being potentially productive in making it easier for prospective jurors to acknowledge bias (if they go far enough), and one making it harder. 

1. Judges refer indirectly to bias. Instead of asking directly about “bias,” they tend to use stand-ins for that term – “sympathy or animosity toward one side,” “affect your ability to be fair,” etc. – at least in their initial questioning. My own view is that judges and attorneys should go even further in “detoxifying” the question, to make it easier for jurors to acknowledge: e.g., would any party “start out a few steps behind the other” at the start of the case? 

2. Judges diffuse responsibility for bias. They use language that ties the bias circumstance rather than to the juror personally. For example, rather than the jury “being biased,” it is the past experience, or previous role that has "affected your ability" to be fair. Putting the responsibility on an external factor makes it less stigmatizing, and easier to admit.  

3. Judges suggest that bias is minimal or unlikely. This is the one that is actively harmful, given that we know from the social sciences that cognitive biases are anything but minimal or unlikely – they’re actually very common (e.g., see Salerno et al., 2020). But in addition to the "any" rather than "some" phrases the researchers point to (with the former linguistically minimizing these experiences), I’ve found that many judges seem to assume that most jurors can be fair, it is the rare exception that calls for an extreme solution in the form of excusing the juror, and that assumption creeps into their language when they describe this duty to jurors. 

The bottom line, there are often pragmatic reasons why judges are trying to limit the number of excused potential jurors. Advocates should ensure that their questioning is making it easier, not harder, to acknowledge the biases that may be there.