Getting all “bubbly” over that spending “bubble?”

Let me just say that I hate this graph!


Why? Well, this is my own version of the graph… but it is a graph that has been used many times over, of late, to make the argument that American public schools have simply been drowning in an excess of public funding for decades and that public school districts nationally have leveraged all of that additional money over time to flood classrooms with additional staff. Of course, this framing is being used to set up the argument that it’s time for a logical correction – a return to sanity – a return to reasonable class sizes, etc. etc. etc.

First of all, even this national graph – which isn’t particularly meaningful – shows a reduction of slightly over 2 students per teacher over the past 25 years. Chop off the period prior to 1985 and the most dramatic shifts in pupil to teacher ratio are wiped away.

Second, national averages just aren’t that meaningful.  Claims like this one, from Mike Petrilli of Fordham Institute, misunderstand entirely the variation in resources across states and variation in effort across states:

The tough-love message to superintendents and school boards nationwide should be clear: The day of reckoning has arrived; let the de-leveraging begin. The spending bubble is over. No more adding staff at a pace that outstrips student enrollment; no more sweetheart deals on pensions or health insurance; no more whining about “large” class sizes of twenty-five. It’s time to live within our means.

The new “reform” story line is that all states have been spending out of control they have taxed their citizenry to death and have spent most of that money hiring more and more teachers even while student populations have stagnated. Further, we have done all of this out of control spending and class size reduction while seeing absolutely no return for our dollar! Pretty simple!  Very compelling, eh? NO.

When one takes a look at individual states – in terms of relative tax burden, changes in tax burden over time – in terms of changes in pupil to teacher ratios over time – and in terms of student outcomes in relation to pupil to teacher ratios and tax burdens – it’s pretty damn hard to find states which actually fit that story line. And public K-12 education is largely a state and local endeavor, not a federal one.

So, let’s take a state by state look. Let’s start here – with the total revenues per pupil – state aggregate – including state, local and federal sources. I’ve adjusted these for inflation (Employment Cost Index – Gov’t Workers), but not for regional variation. For more information on comparing across states, see THIS POST. (<–See this link if you want to know how states rank compared to one another on state and local revenues).


What this shows us is that there’s quite a bit of variation in revenues for local public school districts across states, with some states investing a lot – like New Jersey, Vermont and Wyoming and some not so much – like Utah, Arizona and California. This graph also tells us that many states really didn’t see consistent spending/district revenue growth throughout the period. No massive… long term… uniform… spending bubble… from which they all benefited.

The patterns of revenue increases mirror patterns of pupil to teacher ratio change over time. Yes, some states like Vermont, Wyoming and New Jersey did see pupil to teacher ratios decline. But, not so for Utah, California (after 1998) or Arizona. In fact, in Arizona pupil to teacher ratios increased over time!


And, it is similarly foolish to assert that all states put themselves over the brink in terms of taxes to support all of this supposedly lavish spending. Here are the state and local direct expenditures as a percent of personal income across states:


Look at New Jersey, which has provided relatively high levels of funding, with increases in the early 1990s and again from 1998 to 2003 (though lagging in the middle). New Jersey is not among highest at all on state and local direct expenditures as a percent of personal income – because incomes in New Jersey are quite high. Arizona is also relatively low and Utah and California only average – not high. The “high” state and local direct spending burden states in this mix are Wyoming and Vermont.

So then, what about that story line we’re being told applies across our nation’s schools –

  1. that they’ve been swimming in money for decades – a huge ongoing spending bubble – and
  2. that they’ve spent it all on pupil to teacher ratios – increasing teacher quantity not quality – and
  3. that they’ve taxed their state residents to death in the process – and
  4. got nothing for it in improved student outcomes.

Well, the only state that seems to come close here is Vermont (and perhaps Wyoming) – which does have high and growing public expenditure burden (and the highest “effort” index at, increased education spending per pupil over time and continued decline in pupil to teacher ratios.

Of course, that last part of the story line about outcome failure doesn’t fit so well, because Vermont does very well on outcomes (See this article for discussion of empirical research on results of Vermont school finance reforms:

It also turns out that even Vermont’s “bubble” story isn’t so simple. Vermont really hasn’t been adding significant numbers of additional staff and spending a lot more in recent years. Rather, Vermont is experiencing a significant decline in student population – but has not adjusted its public education system accordingly – reorganizing into more efficiently organized schools and districts. You see, if you spend the same amount and retain similar numbers of teachers, when enrollments decline, pupil to teacher ratios decline and per pupil spending goes up. Yeah… same effect as a spending bubble – but very different cause. And yes, this is something that should be addressed. But again – somewhat different issue! Except for an infusion of funding in the late 1990s in response to school funding equity litigation – Vermont has not necessarily been on a wild education spending binge to reduce class sizes – Rather, the state is a victim of declining school-aged population, resulting in declining enrollment and declining pupil to teacher ratios.

Here’s Vermont enrollment over this same period.


This map shows the locations and enrollment of Vermont schools. Small red dots are schools with 50 or fewer students and small orange/brown dots have 50 to 100.  In some areas of the state, you can find small red dots within a few miles of each other (many of which are small town schools – in separate towns – actually serving the same grade levels/ranges).


Okay, so if this story line doesn’t even fit for Vermont, then who? Perhaps no-one! Arguments built on assumptions of “national trends” regarding financing, class size and/or most features of our public education systems – our 50 and then some systems – are generally unhelpful (if not entirely misguided).

Arguments that encourage state legislators across the nation – including those in states like Utah, Arizona and California that now is the time to cut, cut, cut, because we just went through decades of spend, spend, spend, are downright ignorant and irresponsible.

And please check out this post: where I explain that within states – especially inequitable states like NY or IL – if and when anyone did benefit from reduced pupil to teacher ratios and more importantly smaller class sizes, it was often those in the most affluent communities.