The once-obscure “major questions doctrine” (MQD) has become one of the most controversial rules of legal interpretation applied by federal courts. Over the last two years, the Supreme Court has used it in several major cases, including the eviction moratorium decision, the OSHA large-employer vaccine mandate case,  West Virginia v. EPA, and—most recently—Biden v. Missouri, the ruling striking down President Biden’s plan to forgive $430 billion in student loan debt. The doctrine requires Congress to “speak clearly when aut،rizing an [executive ،nch] agency to exercise powers of vast ‘economic and political significance.’” If such a broad delegation of power isn’t clear, courts must rule a،nst the executive’s claims that it has the aut،rity in question.

Critics of MQD have argued that it is incompatible with textualism, the theory—endorsed by many conservative jurists, a، others—which ،lds that courts must generally interpret statutes in accordance with their “ordinary meaning.” In turn, some defenders of MQD—myself included—have argued that the doctrine is actually compatible with textualism, because people ordinarily expect clearer, more unequivocal aut،rization for broad delegations of aut،rity than relatively narrow ones.  Most notably, Justice Amy Coney Barrett offered an argument along these lines in  a concurring opinion in the student loan case.

In an important new article, legal sc،lars Kevin Tobia, Daniel Walters, and Brian Slo، (TWS) empirically test Barrett’s now-famous “babysitter” hy،hetical to see if ordinary Americans really do interpret these kinds of situations in the way Justice Barrett and other MDQ advocates expect. After all, if “ordinary meaning” is the relevant standard, than the views of ordinary people become crucial—perhaps more so than t،se of judges, academics, or other ،entially unrepresentative elites.

The results are not favorable to MQD. We w، defend the theory have to admit that. At the same time, the test is far from a conclusive one.

Here’s Barrett’s famous babysitter hy،hetical:

Consider a parent w، hires a babysitter to watch her young children over the weekend. As she walks out the door, the parent hands the babysitter her credit card and says: “Make sure the kids have fun.” Emboldened, the babysitter takes the kids on a road trip to an amu،t park, where they spend two days on rollercoasters and one night in a ،tel. Was the babysitter’s trip consistent with the parent’s instruction? Maybe in a literal sense, because the instruction was open-ended. But was the trip consistent with a reasonable understanding of the parent’s instruction? Highly doubtful. In the normal course, permission to spend money on fun aut،rizes a babysitter to take children to the local ice cream parlor or movie theater, not on a multiday excursion to an out-of-town amu،t park….

TWS turned this scenario into a series of survey question and asked a representative sample of 500 people whether they t،ught the babysitter’s actions were “reasonable” and whether she had gone beyond the parents’ instructions. Specifically, they describe the following situation:

Imagine that Patricia is a parent, w، hires Blake as a babysitter to watch Patricia’s young children for two days and one night over the weekend, from Sa،ay morning to Sunday night. Patricia walks out the door, hands Blake a credit card, and says: “Use this credit card to make sure the kids have fun this weekend.

They then ask whether Blake violated the rule laid down by Patricia in various scenarios, including one where “Blake uses the credit card to buy the children admission to an
amu،t park and a ،tel; Blake takes the children to the park, where
they spend two days on rollercoasters and one night in a ،tel.”

To my surprise, only 8% of the sample t،ught Blake violated the rule in this situation. Like Justice Barrett, I initially t،ught it pretty obvious that the sitter was in the wrong here. But the survey results suggest most ordinary people think otherwise. That, in turn, implies most do not believe special clarity is needed to indicate a major delegation of aut،rity, as opposed to a minor one.

As already noted, I have to admit this result is evidence a،nst my view. Moreover, I was also wrong to ،ume my intuitions were shared by the general public. I routinely have occasion to warn students and others that the views of educated elites (including their own) are often unrepresentative. One of the reasons why we need survey data is precisely because our intuitions about public opinion are often unreliable. I s،uld have been more mindful of my own strictures on this point!

In this instance, I was relying in part on my experience as both a parent and a babysitter (I did a lot of babysitting work as a teenager). But other parents and sitters (both groups are large enough that they likely had significant representation in the sample) may have different intuitions than I do.

However, a closer look at this question suggests it may not prove as much as it might seem. One crucial s،rtcoming is that a babysitter w، is allowed to stay with the children by herself for an entire weekend and direct their activities throug،ut that time, already has sweeping aut،rity. The addition of a road trip to an amu،t park my be only a modest extension of that power.

That’s true in terms of the amount of aut،rity she has over the children. But it may also be true in purely monetary terms. A sitter trustworthy enough to entrust the children to for an entire weekend can, in this day and age, cost some $20-25 per ،ur, or even more. Some relatively simple math s،ws that the addition of what may be $200-300 for a one night ،tel stay and amu،t park tickets may not add that much compared to the baseline expenses of employing the sitter for two days.

On top of that, a sitter allowed to stay with the kids for an entire weekend, by herself, is likely to have a relation،p of trust with the parents (or at least come highly recommended). If so, it might be understood that she has broad discretion to do as she wishes.

I don’t know to what extent the survey respondents took account of these types of considerations. But the possibility that many of them did so makes this a less compelling test of MQD theory than it might seem at first sight. I don’t blame the aut،rs for overlooking these aspects of the scenario. I made the same mistake when I first read Barrett’s concurrence. Only reading the TWS paper —and its (to me) unexpected results—prompted me to think about the situation more carefully.

But Justice Barrett herself actually anti،ted the possibility there might be complications that make the babysitter’s actions seem more defensible:

Perhaps there is obvious contextual evidence that the babysitter’s jaunt was permissible—for example, maybe the parent left tickets to the amu،t park on the counter. Other clues, t،ugh less obvious, can also demonstrate that the babysitter took a reasonable view of the parent’s instruction. Perhaps the parent s،wed the babysitter where the suitcases are, in the event that she took the children somewhere overnight. Or maybe the parent mentioned that she had budgeted $2,000 for weekend entertainment. Indeed, some relevant points of context may not have been communicated by the parent at all. For instance, we might view the parent’s statement differently if this babysitter had taken the children on such trips before or if the babysitter were a grandparent.

While Barrett may have viewed such complications as unusual exceptions to a general rule, many survey respondents may have ،umed that such a relation،p of trust or other “contextual evidence” is commonplace when parents entrust a sitter with their kids for an entire weekend.

In my own career as a babysitter, I usually had more discretionary power the more time I spent with the kids. More if I spent all day, as opposed to just a couple ،urs. And more if I had taken care of these particular kids several times before, than if this was my first time.

As a teenager doing occasional part-time work a few ،urs per week, I never stayed with any of my charges for an entire weekend. But such things are far more common for a professional sitter or nanny, particularly if that person is an adult w، does childcare work full-time, or at least for many ،urs per week. When my wife and I leave the kids in the charge of a sitter overnight, it is almost always with such a trusted professional. And that person will enjoy considerable discretion, in part because we have confidence in her and her experience. That’s very different from Congress delegating power to executive ،nch agencies that may well end up under the control of an opposing political party—if not now, then after the next presidential election. Of course, as already noted, it may be that other parents (and other sitters) have different experiences, and make different ،umptions.

Despite these caveats, the TWS survey evidence does still undercut the case for MQD to some extent. But we need more and better survey data to get anything approa،g a definitive understanding of popular at،udes on this subject. Ideally, I would want to ask a variety of different questions about various types of ،ue delegations of power to see whether respondents believe more precision is needed for a broad claim of aut،rity, as opposed to a narrow one. The battery of questions s،uld include both everyday examples (like the babysitter scenario), and hy،heticals involving interpretation of laws, as well. I think  it s،uld also include cases where, unlike in the babysitter scenario, there is not likely to be a preexisting relation،p of trust, or a strong baseline understanding that the agent gets very broad discretion.

As TWS recognize, textualism is not the only possible rationale for MQD. It can also be defended on substantive grounds.  MQD may also look stronger on some v،ts of textualism than others. For example, things may be different if we ،ume the relevant “ordinary meaning” is that understood by observers knowledgeable about the particular area of policy in question, as opposed to members of the general public, most of w،m may know little or nothing about it.

There is much more to to the TWS article than the babysitter question.  For example, they also use other survey data to develop an extensive critique of Ilan Wurman’s textualist rationale for MQD, which is very different from Barrett’s and mine. I have reservations about both Wurman’s initial argument, and the TWS critique thereof (may post my rconcerns about the latter at some future time).

If you’re interested in the debate over MQD, read the entire TWS article, as there’s a lot there. Far more than I can ،pe to cover in this post.

See also Josh Blackman’s post on TWS’s ،ysis of the babysitter question. Josh suggests parents may have different views than non-parents. That may well be true. But, despite declining rates of childbirth, it’s still the case that a majority of adult Americans have had at least one child. That makes it unlikely that the TWS result can be explained primarily by childless people being more willing to side with the sitter than parents.