Your Theory of Action Matters, Actually

It's not about the schools and it shows

Author

James Shanahan
May 13, 2026

Hi! This is Scholastic Alchemy, a twice-weekly blog where I write about education and related topics. Mondays usually see me posting a selection of education links and some commentary about each and Wednesday posts are typically a deep dive into an education topic of my choosing. If Scholastic Alchemy had a thesis, I suppose it would go a little like this: We keep trying to induce educational gold from lead and it keeps not working but we keep on trying. My goal here is to talk about curriculum, instruction, policy, public opinion, and other topics in order to explain why I think we keep failing to produce this magical educational gold. If you find that at all interesting, please consider a paid subscription here, or at the parallel publishing spot on Beehiiv. (Some folks hate the ‘stack, I get it.) That said, all posts are going to remain free for the foreseeable future. Thanks for reading!

Author’s Note: Hi all! Glad to have found some free time for personal reading and writing again. I’ve been bogged down with various responsibilities, family needs, sickness, and some conference travel but now that things are winding down and we’re headed into the summer, I should be back on the wagon. Thanks for your patience!

-James

In the two weeks since I last posted, there have been several education related posts from authors and outlets that I frequent.

Firing bad teachers, theories of action

Yesterday over at The Argument, Bagley and Gordon, a pair of authors with lots of policy and advocacy experience, argued that Democrats should fire bad teachers. They also mention firing bad cops but that’s not really in my wheelhouse. They also barely mention the cop stuff and totally drop it by the end of the piece. It almost feels like a smokescreen, so they don’t totally single out teachers, therefore I won’t comment on that part. Their overall position is an anti-union one and they think that Democrats would do well to throw unions back under the bus and distance themselves from teachers unions. However, they don’t have a lot to say about unions impact on schools besides noting that unions don’t necessarily help schools attract the most talented teachers while also being terrible for municipal budgets.

What I want to dig into here is something called a Theory of Action. That’s a fancy term for “what you think will happen when you do something” but it’s one way we can analyze policy arguments like those put forward by Bagley and Gordon. The way I see it, there are really two or three different theories in operation here. Let’s lay them all out and see if we can’t make some sense of this muddled argument.

First, let’s start with the headline’s point and the theory of action around it. Democrats should fire bad teachers. This is the middle part of a logical series of events. Before Democrats can fire bad teachers, they must 1) identify bad teachers, 2) possess the ability to fire them. How do we identify bad teachers? They don’t say. Despite being an important precondition, it’s given zero explanation or thought. Turn your attention instead to the need to be able to fire them.

Under collective bargaining agreements, government dismissal processes are often incredibly onerous. Across multiple rounds of review, the government must extensively document and defend multiple efforts to help employees succeed.

After Democrats fire bad teachers, we have several outcomes listed. Students will earn more money over their lifetimes.

Students with the least effective teachers earn less as adults, save less for retirement, and are less likely to attend college. They even have more kids as teenagers. Raj Chetty and his coauthors found that replacing the worst 5% of teachers with an average teacher would raise the present value of an average classroom’s lifetime earnings by more than $250,000.

Poorer schools will see more stable teacher turnover.

Collective bargaining agreements often mandate that seniority should be the only thing that matters for layoff decisions. This means that an outstanding new hire will lose out to an indifferent veteran waiting for their pension to vest. When schools face layoffs, poorer schools with more junior teachers see the most turnover — exactly where you would hope for more stability.

New teachers will earn more money (because of changes in retirement plans, more on that below).

Consider public pensions. A lot of research suggests that a dollar spent on pensions is less effective at recruiting and retaining staff than a dollar spent on upfront pay. The reason is a familiar one: people care about the present more than the future.

Matthew Yglesias recently noted an incredible fact: “while New York pays 70 percent higher teacher salaries than Louisiana on average, its entry-level salaries are only 7 percent higher.”

Democrats will become more popular.

David Broockman and Josh Kalla recently tested how Democrats and Republicans changing their views on issues might move voters. Across 29 topics, the authors found that embracing teacher accountability ranked third for winning voters over. It’s a slightly better issue for Democrats, the study found, than moving to the middle on transgender kids in sports or on immigration.

And, in parallel, Republican candidates got hammered especially hard for expressing opposition to the view that bad teachers should be dismissed.

Wow. How many millions more words have Democrats spoken on gender rights than teacher accountability over the last year? That gap exists because getting crosswise with teachers’ unions is more politically costly than fighting with gender-equity groups.

Second, there is a budgetary argument that is less related to firing bad teachers and more related to the impacts of public sector pensions on city finances. The theory of action here begins with breaking union’s political power. Before blue cities like Chicago and New York can fix their budgets, they have to get around teachers unions and their pesky collective bargaining for defined benefit pensions.

How did so many governments end up with retirement plans rarely found in the private sector? Federal law leaves states and cities far more freedom than private employers to underfund pensions. And that means unions fighting for their members can secure promises of future payouts that current leaders need not fulfill. Politicians win praise for the promises, then bill their successors.

The result of fighting teachers unions is pension reform. They use Rhode Island as an example.

Cities interested in effective government services should adopt pay scales that appeal to talented people early in their careers. Compensation would still grow, but not at the same rate.

Rhode Island traveled some of that path beginning in 2011, when its then-treasurer, Gina Raimondo, raised the alarm about pension costs, shifted workers into a hybrid between traditional pensions and 401(k) plans, and was able to cut the pension pressures that hung over her state’s education spending. For that, she earned the opposition of the state’s public sector unions — and easy election as governor in 2014.

And where does this lead? What do we get for reforming pensions. For one, we get higher new teacher pay as noted above. This is because pension dollars are seen as a trade-off with simply paying teachers more money up-front, with up-front pay being more effective in terms of teacher retention. The bigger trade-off, though, is with other aspects of running a municipal government.

While all of this has unfolded, both Chicago and New York City face tight budgets. The month before Pritzker signed the pension sweetener, Chicago Public Schools cut almost 1,500 staff, including more than 400 teachers. New York Mayor Zohran Mamdani is now in court fighting against a legally mandated expansion in housing vouchers that he supported as a candidate.

Hiking pensions while cutting schools and housing makes no sense. It is possible only because no one outside of state capitols or public sector unions is paying much attention.

Well-functioning municipal government is the goal. Well, that and getting the Democrats to be more popular.

Democrats need to figure out how to better spend the money they have. And it is only broken, transactional politics that prioritizes early retirement for government workers over day care slots — or that pays for bad government employees while failing to attract great ones.

To summarize, there are two interrelated theories of action here. Both lead to the public trusting and liking Democrats more and maybe winning some elections. Both require diminishing the power of teachers unions to protect teachers and require a defined benefit retirement plan. Along the way, we get higher new teacher pay and more stability at poor schools plus higher performing teachers leading to higher pay for the kids as they grow up. We also get solvent municipal budgets so that other public programs don’t face cuts and services are provided more effectively.

Paying good teachers more, theory of action

The other post is from last week. Matt Yglesias thinks we should pay more for the best teachers. Matt’s post, despite being longer and winding through more analysis of voters opinions on things, contains a much clearer theory of action and it doesn’t have odd crossovers with policing or focus on budgets. Matt says, hey, voters care enough about schools that changing school policies to include higher pay for better teachers will meaningfully convince them to elect Democrats. So, let’s break it down. Before we can pay better teachers more money, we need two things: 1) we need to identify the best teachers; 2) we need to be able to pay them more. Unlike Bagley and Gordon, Matt actually spends time on both necessary preconditions and even complicates his own thinking with some careful qualifiers.

He says that one reason performance pay floundered so much in the Obama era was because nobody was implementing programs rigorously.

My qualitative understanding of why is that the incentive program was almost too effective. States really wanted the money. Teachers unions wanted states to get the money. So everybody duly submitted their grant applications and said the magic words, but then in most cases they did not actually create rigorous and well-implemented evaluation systems that meaningfully changed anything.

Often identifying the best teachers was simply based on students’ test scores or highly subjective observations. That certainly matches my experiences under merit pay. Matt says you need good implementation and offers three examples: Dallas, DC, and Andhra Pradesh, India. While each program differed, overall the extra pay in all three was based on two things. First, a holistic system for determining quality that relied on multiple measures, observations, and other outcomes. Second, incentive pay to work in struggling schools.

Like Bagley and Gordon, Matt sees teacher incentive pay as something unions will oppose on principle so implementing some kind of merit pay program will require fighting against teachers unions. Matt draws on some survey data to make the point that the public is generally supportive of merit pay and doesn’t seem to notice that this is related to teachers unions at all. He closes by telling politicians not to be afraid of the unions because merit pay is broadly popular.

A Democrat who just says that teachers should be paid according to ability is appealing to swing voters and primary voters simultaneously. You know and I know that it will make teachers’ unions mad, but I think going out of your way to have a fight about unions distracts from the more important point that this is just an overwhelmingly popular position that is well-supported on the merits.

Just say and do the thing. Don’t be scared off by the unions. If they come at you, punch back by just restating the compelling core thesis: Teachers’ work is important and should be well-compensated, but that means paying to retain the people who are good at it, not just arbitrarily paying more for whoever happens to have been in the job longest.

Like Bagley and Gordon, Matt’s theory of action is ultimately about getting Democrats elected. So the theory of action is: tell the public that you will identify the best teachers and pay them more, resist union backlash, get elected and, presumably, do the thing you said you would do. Unlike Bagley and Gordon, Matt is not promising much else. Indeed, he seems to think that the pay changes would have limited impact on student outcomes.

I’m a little torn here because I don’t want to overpromise.

As a D.C.P.S. parent, I can tell you that the system does not work miracles. The main thing that parents in the real world care about is peer effects and the demographic composition of their kid’s school, not sophisticated value-added measures of student performance.

What great teachers in a diverse public elementary school do is have an unusually high success rate at taking really low-performing kids up to a baseline of basic competence. From both an objective statistical viewpoint and also my eyeball experience of seeing a group of kids with mostly very high-scoring teachers, it’s clear that the actual impact of teacher quality is modest compared to innate ability and out-of-classroom stuff.

I think that’s a pretty refreshing take and it gets to the larger point of both pieces which is that none of this is about academic outcomes. It’s about getting Democrats elected. So long as there appears to be some political benefit to making arguments for merit pay or for firing bad teachers, all of the authors think politicians should make those arguments and advance policies along those lines.

Do these theories of action make sense?

Both posts are responding to some polling that was done by Brookman and Kalla about moving “to the center” on various issues. Among their findings was that the public generally supports some kind of mild teacher accountability which they characterize as

It should be difficult to fire teachers, but better teachers should get paid more than worse teachers.

I think it’s really interesting that Bagley and Gordon interpret this as “fire bad teachers” while Matt Yglesias interprets this more strictly in alignment with what the actual popular centrist statement is. I think that gets to some larger structural problems with Bagley and Gordon’s argument. It’s really striking that they don’t attempt at all to discuss how to identify good or bad teachers who need to be fired. In fact, Robert Gordon is one of the authors of a highly cited paper about identifying effective teachers! This is, perhaps, because the value-added measures promoted by that paper were exactly the kinds of measures that failed to reliably reform the teacher workforce. As Matt points out, a lot of what happened under Obama using VAMs was sloppy and poorly implemented. Gordon even owns up to the limitations of VAMs in the comments.

I coauthored a paper about value-added metrics that has been cited like 900 times. Gordon, Kane, Staiger 2006. You could look it up.

That said, I agree VAM has lots of limitations, and I have even more regrets about the too-fast scaling of VAM under the Obama Administration--and a bit of my own influence. I wish we’d gone slower and put more emphasis on local ownership and common sense (though then folks would have criticized the system as “subjective”).

I think he’s also right that people would have criticized local ownership of determining who was good or bad as being subjective. But this gets to the meat of what’s going on here. It’s really hard to figure out who is or is not a good teacher. That goodness or badness can change based on changes in school population or where the teacher is working. One big problem with the last round of “K-12 Accountability” is that it produced so much focus on the measures used to determine if teachers kept their jobs or schools closed or if incentive pay was awarded that it made schools illegible for parents. Everything in schools was reoriented around standards and testing and parents turned against it. That’s part of the technocratic backlash we’re experiencing around schooling.

In part, I think what I’m noticing is that these posts are really about schools. Just like with literacy, the battles around schooling are not about schools or kids or outcomes. What both of these posts are really doing is making a case for centrist political positioning so that Democrats can get elected. Matt points out that the survey evidence says this positioning matters to voters.

The authors also find consistently in this experiment that voters with conservative views on a given topic appreciate it when Democrats move to the center on that issue (and vice versa for voters with liberal views when evaluating Republicans). The upshot is that there is a lot of electoral upside to offering the pretty tepid take that teacher compensation should be based on performance rather than pure seniority, even without any compromise on basic job protections.

And, fine, that can be the case. I have my own ideas about what the voters really want from schools but I’m not writing today to object to everything in detail. What I’m trying to figure out is why some of the reasoning is so scattered. Bagley and Gordon leave out important preconditions. Matt says he doesn’t think the merit pay policies will accomplish much. It’s because none of this is really about schools, making schools better, improving kids’ outcomes, or meaningful education reform. It’s about getting Democrats elected, full stop.

This is all really weird since one author is a prime-mover on the whole merit-pay thing under Obama and another pens posts lamenting how America stopped caring how poor kids do in school. I guess at the moment, political concerns trump everything else.

Thanks for reading!