Category Archives: School Finance

What helps poor children more: increasing the EITC or increasing educational funding?


*Note: I’m still working through all this research. If I’ve made a mistake, let me know!


In a previous post, I argued that after a certain expenditure level, call it $15K a child, my guess is that families would be better off with additional funds as cash transfers. So a district that spends $20K a student (D.C, Newark, New York, etc.) would best serve families by spending $15K per student and sending $10K in cash to a two child home.

Here’s  a similar policy question: if taxpayers are willing to spend extra money to help poor children, should they increase the earned income tax credit (giving working low-income parents increased money) or raise educational funding?

Last month, Kevin Carey and Elizabeth Harris wrote a NYT column summarizing the most recent research on increasing education funding. Their conclusion: money probably does increase test scores.

Unfortunately, they did not the review the research on what happens when you give similar amounts of money to families via other transfers, such as the EITC.

The Effects are Within the Same Range

While none of the research is an exact science, research on the EITC (see a summary here) finds that children under the age of 12 see increases of ~.06-.1 SD per $1,000 increase in the EITC (cumulative full schooling impact).

Rothstein finds that low income districts increase their performance by ~.1 SD per $1K increase in funding (cumulative 10 year impact).

And while these rough estimates find a slightly higher impact for education spending, remember that education spending costs twice as much for a two child family.

The EITC estimates are based on $1K per family, while the public spending estimates are based on $1K per child.

Also: families don’t use the EITC every year, so while increasing spending via education expenditures is a constant expense, the EITC is a variable expense that is only used when families are in poverty.

Because of these factors, my hunch is that the ETIC effects are actually higher per $ spent, but for sake of argument, let’s call it a wash.

The EITC is Well Targeted, Education Spending is Not

Assuming equal effects, the reason the EITC is more efficient in that is better targeted: only poor families get the increases.

Most state funding formulas, on the other hand, give increased funding to districts, not individual students.

This means that giving $1K per student in additional funding to low-income districts is spread across all students in the district, not simply low-income students.

Most importantly, it means that low-incomes students in non-low-income districts don’t receive the benefit.

Rothstein notes as much in his study, writing:

Courts and legislatures can evidently force improvements in school quality for students in low-income districts. But there is an important caveat to this conclusion. As we discuss in Section VI, the average low-income student does not live in a particularly low-income district, so is not well targeted by a transfer of resources to the latter. Thus, we find that finance reforms reduced achievement gaps between high- and low-income school districts but did not have detectable effects on resource or achievement gaps between high- and low-income (or white and black) students. Attacking these gaps via school finance policies would require changing the allocation of resources within school districts, something that was not attempted by the reforms that we study.

Unless States Change Their Funding Formulas, Transfers > Increased Spending

In summary: transfers are targeted at all poor families in a jurisdiction, while education funding increases are generally only targeted at poor families living in low-income districts.

Assuming the research holds on both transfers and education spending – and we continue to see similar effects – then transfers seem to be the much better option, as they reach low-income families in all jurisdictions.

More Research Needed

I view the question of wage subsidies vs. universal basic income vs. increased public services to be one of the most important policy topics out there.

Hopefully we can learn more about the cost / benefits with further research.

How to Increase Funding for Public Schooling by ~10 Billion a Year

Facilities are very expensive, and all things being equal, spending less on facilities allows for more money to be spent on instruction.

This report found that in Chicago charters spend 46% less on facilities than does Chicago Public Schools.

I imagine this is a larger deferential than in most districts. And while I don’t I have time to do a full research review, in most jurisdictions I work in I deal with facility costs, and it’s generally the case that charters spend less per student than the district does.


Instead of 46%, let’s consider a lower end estimate of a 15% differential.

Here’s what we spend national on facilities, according to the NCES:

Screen Shot 2016-06-13 at 4.39.32 PM.png

So about ~1K for capital outlay and .37K for interest on debt (which I imagine has a facilities component to it) out of a total of 12.4K.

Let’s call roughly 10% of the per-pupil or 1.2K per student.

Reducing this cost by 15% would save us $180 dollars per student or a 1.5% decrease in total spending.

On an overall budget of $621 billion, we’d save about $9 billion a year.

Let me know if I got my math wrong….


These are all very rough estimates, and while they are fairly conservative, they could be wrong.

But it’s surely plausible that we could shift $10 billion a year from facilities costs to instructional costs by moving to an all charter school system.

Spent well, this could support tutoring, field trips, class size reductions – or whatever educators and families thought best.

To the extent you believe money matters in schooling, it’s worth considering how increasing charter school development can drive more money into educational experiences rather than overpriced buildings.

Accounting Like a State


The phrase “seeing like a state” describes the tendency for governments to try and solve problems with top-down government solutions.

When you see like a state, every problem is solvable with the right technocratic fix.

I’d like to introduce a new phase: “accounting like a state.”

When you account like a state, finances are viewed through the lens of government program solvency, not outcomes.

I think Peter Greene falls for this way of thinking in his piece “The Financial Fantasies of School Choice.”

Peter’s argument is as follows:

  1. Public education benefits from economies of scale; specifically, charter schools are highly inefficient and they end up reducing teacher salaries and pensions to make up for these inefficiencies.
  1. Public school districts have a high fixed costs, so when funds “follow” the student, districts often lose more in revenue than they can save in costs.
  1. Public accountability over taxpayer funds for education is best accomplished through elected school boards (where all citizens can vote for societal ends) rather than choice (where public school parents pursue their own individual ends).

Some commentary:

  1. Efficiency can only be determined with respect to productivity. Peter does not consider academic outcomes in his efficiency argument. At least in the case of urban charter schools, there is evidence that they are more efficient than traditional schools, in that they achieve better academic outcomes for less funds.
  1. With regard to teacher pay, Peter fails to mention that many teacher retirement systems currently risk insolvency. Peter claims choice systems take money away from teachers. I would argue that states have been making promises to teachers that these states may not be able to keep. Or to put in other way: charter schools must balance their books. Many state retirement systems, to date, have not. I’ll leave it up to you to decide which system is based on fantasy. One other note: in some states charter are forced into these retirement systems, which puts their own schools at risk of being fiscally damaged by a problem they did not create.
  1. I agree that school districts have fixed costs. So do many organizations in the world. The solution is not to ban competition. The solution is for the district to build a financial model that allows it to stay solvent even if it loses students. The district might also try to figure out how to deliver an educational experience that keeps families coming back year after year.
  1. Public accountability can be achieved through numerous ends. Voting is one method. So is allowing users of government delivered goods to have the option to use non-profit delivered goods. Peter suggests we put all our eggs in the voting basket and not consider other, compatible forms of democratic accountability. An elected school board that oversaw a system of non-profit operated schools could allow for both forms (voting, choice) of accountability to simultaneously exist.

In sum (and perhaps less generously), Peter’s fantasy is as follows:

  1. School districts are efficient because they use economies of scale to deliver a strong educational experience for students.
  1. States funded teacher pension systems are based on sober predictions of market returns.
  1. The high fixed cost of operating a school district is a good reason to prevent competition.
  1. Democratic accountability is incompatible with giving choice to the users of government services.

As Matt Dicarlo recently pointed out, education reform advocates sometimes do appear to be living in a fantasy world – in that they make academic performance predictions that they will inevitability not meet.

But in this case, I don’t think choice advocates are living in a fantasy world.

It is the anti-choice advocates that are succumbing to the flaws of accounting like a state.