With virtually every state’s statewide elections coming up, let’s talk about state attorneys general.

No, really.

How many state AGs can you name? If you’re lucky, you might be able to get two or three, most likely. Even the most politically plugged in people are unlikely to come up with more than seven or eight. I’ve been working on this post for multiple months and I could only name 22, or less than half, without prompting.

So what, right? State AGs just do some arcane legal thing that doesn’t really have anything to do with day to day life. Sure, it would be nice to have their party registration match yours, but how much does it matter?

As it turns out, it can matter a ton. If you care at all about criminal justice reform, the federal government is not where most of this work is done. Although there has been a surge in the number of people incarcerated in the federal system, state systems still hold nearly 90% of all prisoners in this country. State AGs therefore have a large say in holding corrections systems accountable. (As illustration, the AG’s office is usually the one defending prisoner § 1983 suits regarding conditions of confinement, assaults by prison employees, etc.) The parties’ positions on CJ reform are converging, but many dangerous myths persist (or are cynically wielded for political gain).

And in addition to this power, most other areas of policy have not been fully preempted by the federal government, and thus state AGs can choose to operate in all sorts of domains if they would like. Former Maine AG Patrick Tierney has a website, StateAG, that outlines some of the possible ways in which a state AG can touch on things that influence our daily lives, and includes this graphic:

You’d expect, given the amount of power that the AG wields within state executive branches, that the AG is going to often be an appointed position. And indeed, so that a government is not working against itself, that would make the most sense. But only five states (Alaska, Hawaii, New Hampshire, New Jersey, and Wyoming) have AGs appointed by the governor. (In Maine, the AG is chosen by the state legislature, and in Tennessee, the AG is chosen by the state supreme court.) This leaves 43 states where the attorney general is elected, and thus where your downballot vote has a substantial impact on policy. (Thirteen of those 43 states have state AGs of a different party than their governor.)

Indeed, many of the Supreme Court cases that have come up in recent years have been brought by Republican AGs trying to chip away at portions of Obama’s legacy—for example, the case invalidating DAPA, United States v. Texas, was originally brought by 27 state AGs who objected to the policy. (Not coincidentally, there are currently 27 states with a Republican state AG.) Most of the large lawsuits against things like the ACA were also joined by most, if not all, Republican state AGs.

Given these stakes, it would make sense for a typical state AG to profile as somebody relatively late in their legal career, somebody with experience managing a diverse practice and working regularly with complex regulatory schemes. But of the current state AGs, eight became AG when they were in their thirties, and an additional 22 became AG before they were 50. (Keep in mind that it is highly unusual to be admitted to the bar before you are 25, just based on the amount of time it takes to go through law school.)

But my goal here is not to argue that this state of affairs needs to change. Indeed, while it may be advantageous to not have an incredibly inexperienced AG, the state AGs’ offices should run substantially like the federal Department of Justice: with civil servants who employ the policy that the AG sets. Any attorney with a sound head for policy and law office management should be considered qualified.

But who tends to end up in these positions? There’s not as standard a career path to this position as there is to, say, judgeships and US Attorney jobs on the practice side, or governorships and Senate seats on the political side. Or at the very least, there isn’t a career path that people, off the cuff, would consider “standard”.

To answer this question, I looked at the backgrounds of every person who has occupied a state attorney general’s office on a non-acting basis in the last thirty years (since 1988), with the caveat that they had to have first assumed the position after 1978. Of the 284 qualifying state AGs, I attempted to approximate the number of years each had spent in each of seventeen possible employment tracks:

1. U.S. Attorney;
2. Assistant U.S. Attorney;
3. District attorney;
4. Assistant district attorney;
5. Assistant Attorney General (both in senior leadership and lower-level positions);
6. Judicial clerk;
7. Law firm (both solo practices and larger firms);
8. In-house counsel;
9. Unelected government position not covered above;
10. Professor (counting only full-time professorships);
11. Criminal defense attorney (excluding defense work done at law firms UNLESS the law firm was focused on criminal defense);
12. Elected office (for part-time elected state legislative positions, those years were double-counted if I could determine a second job for those years);
13. Active-duty military service;
14. Public interest work, aside from work done at law firms or criminal defense organizations;
15. Thinktanks or similar organizations;
16. Work directly for a political party;
17. Work as a judge.

I will note that for every single state AG’s background I could find, every job that they held after law school fit into one of these 17 categories.

For seven of the AGs in the sample, I could not find any reliable data with which to approximate their credentials, and as such, I had to ignore them. Of the remaining 277, 61 were from states with appointed AGs, leaving 215 elected AGs. The data for each of these sets follows:

So what are the takeaways here?

First, it’s interesting to note that the 277 AGs averaged roughly 15 years of law practice experience, but that number was highly variable, with a high of 42 years decreasing all the way to seven AGs clocking in at zero years of practice experience (with plenty more having under five years).

Second, while there is a lot of experience in AG’s offices represented, firm experience reigns as by far the most common category, followed among elected AGs by elected office and other government work. I will note that most of these firm jobs were not global corporate firm (biglaw) jobs but rather solo, small, or regional firms, so there is more diversity of experience represented here than might meet the eye.

Third, state AGs have prosecution-heavy backgrounds, just as much as DAs and judges. Roughly 1/3 of state AGs and almost 40% of elected state AGs have been state prosecutors, while roughly a quarter of appointed state AGs have been federal prosecutors. By comparison, only seven of the 277 state AGs have ever been criminal defense attorneys. (In fact, all 277 state AGs, between them, have only 23 combined years of criminal defense experience, or less than Larry Krasner.)

To the extent that criminal justice reform matters to you, this last one should be a huge red flag. Aside from the direct criminal justice impacts noted above, we should be wary of creating pipelines to powerful jobs that run through prosecutors’ offices, which has an effect not only on the policy positions of those running for these jobs but also the incentives for politically ambitious lawyers to be drawn to prosecutors’ offices.

And importantly, the question of creating pipelines is highly relevant to the office of the state attorney general itself. Of the 175 state AGs who were elected in their own right, 117, or two-thirds, have sought or are seeking higher office. Dozens of them are doing so, or did so, with no other electoral experience. That is, state attorney general was their springboard (or attempted springboard) to marquee elected offices. You may recognize some of these names:

• Christine Gregoire, a career AAG, became a two-term Democratic Governor of Washington
• Sheldon Whitehouse, a career prosecutor, is now Rhode Island’s junior senator, with an unsuccessful gubernatorial run in between
• Jennifer Granholm, one of only two Democrats to be reelected governor of Michigan in the last 60 years, ran from the state AG’s office
• Janet Napolitano, former Secretary of Homeland Security and twice-elected Governor of Arizona, spent one term as AG before running directly for governor

However, the AG pipeline has also turned out some subpar candidates as well:

• Martha Coakley, a career prosecutor, who ran first for U.S. Senate and then for governor of Massachusetts from the AG’s office and lost both races
• Jack Conway, who ran first for U.S. Senate and then for governor of Kentucky from the AG’s office and lost both races, with the 2015 gubernatorial election being only the third Republican gubernatorial win in 70 years

And of course, the AG pipeline can swing both ways:

• In West Virginia in 2012, longtime Democratic WV AG Darrell McGraw was defeated by out-of-state lawyer Patrick Morrisey by 16,024 votes. As compared to the 2012 presidential balloting, though, there were nearly 27,000 undervotes. Morrisey is now a candidate for U.S. Senate.
• In Nevada in 2014, Republican Adam Laxalt, who had never held elected office before, won his state AG race by 4,868 votes. The ballot line for None of These Candidates gained 15,643 votes. Laxalt is now the leading candidate for governor of Nevada and likely has presidential aspirations down the line.

What does this mean going forward, though?

First, I think it’s important for us to incorporate AG races when we’re discussing marquee elections, to ensure that people we know don’t leave that ballot line blank. It’s not the easiest sell, but I don’t think it takes all that much time to make a sell for why the AG affects day-to-day life. (“They’re the state’s lawyer. If the Republicans hold that position, they get to take pro-life, anti-immigrant, anti-environmental, and anti-minority positions when they represent the state.”) This is especially easy to do if you’re already knocking on doors for a separate election.

Second, I think it’s important to pay attention to AG primaries. While the gulf between a left and a right candidate for this type of position is substantial, the gap between a center-left and a left-left candidate might have an outsized effect on poor communities and communities of color in particular.

Third, I think that the recruitment pool for state AG candidates is ripe for expansion. While Larry Krasner’s election in Philadelphia signals that a change may be coming in the way criminal justice is administered, I would argue that progressive reformers are actually an easier immediate sell in a position that doesn’t directly implicate low-level criminal prosecutions. And while I think civil legal services lawyers and public defenders aren’t often thinking about running for office, attorney general is less of a theoretical stretch than either a prosecutorial or legislative position. And given the profiles of those who get elected to these positions, public defenders and civil legal services lawyers in their thirties should consider themselves qualified to run for these positions.

I do think that focusing attention on state AG races is going to be tough to do by splitting your finite donation resources even further; while I don’t want to discourage donations to candidates that are worthwhile, I think the more important work is grassroots in nature. Without a bottom-up conversation about the importance of state AGs, the undervotes will chronically remain, even with a candidate that might otherwise be exciting.

There is too much word yet to spread in order to make progress, but having conversations is the easy part.

We can change things. But as we tend to the fruits of our labor, we need to plant more seeds.