Krugman Messes with Texas Again: State Spending and Education Outcomes

Absent a Republican president, Paul Krugman's favorite target for disdain is the Texas statehouse.  Professor Krugman's column today equates the likely reductions in spending on education and health care in Texas (and elsewhere) with an attack on children. 

While low spending may sound good in the abstract, what it amounts to in practice is low spending on children, who account directly or indirectly for a large part of government outlays at the state and local level.

 And in low-tax, low-spending Texas, the kids are not all right. The high school graduation rate, at just 61.3 percent, puts Texas 43rd out of 50 in state rankings. Nationally, the state ranks fifth in child poverty; it leads in the percentage of children without health insurance. And only 78 percent of Texas children are in excellent or very good health, significantly below the national average.

Krugman's claim is pretty straightforward: Texas is bad for kids because of the state's low-spending policy choices.  Just one problem, though, at least in terms of public education, Texas is not exactly miserly.

Based on data  from the National Information Center for Higher Education Policymaking and Analysis, Texas had a high school graduation rate of 65.29% in 2008 (the most recent year for which data are reported).  That ranked Texas 43rd in high school graduation.  However, Texas ranked 26th in the country in terms of per capita spending on education ($293) and 15th in state and local education spending per pupil ($7,934).  Despite being at or better than the median in terms of spending on education among states, Texas has one of the worst high school graduation rates.  As incongruent as this may seem at first, Texas's situation is indicative of a larger trend.  Education spending at the state level is not strongly related to state-level educational outcomes.

The (limited) correspondence between spending and outcomes is evident in the following scatterplots.  (Data are from 2008 reported by National Information Center for Higher Education Policymaking and Analysis).  The first plots state education spending per capita against graduation rates.  The second plots the average combined state and local expenditures per pupil (adjusted for cost of living and state enrollment mix) against graduation rates. 

Both scatterplots tell the same story: money alone does not determine educational outcomes.  In fact, both figures show a negative relationship between educational expenditures and educational outcomes.  The correlation between state per capita spending and graduation rates is -0.25 (p=0.08).  The correlation between state and local spending per pupil and graduation rates is -0.40 (p<0.01). 

Obviously, spending does not cause poor educational outcomes at the state level.  In fact, my hunch would be that poor educational outcomes attract remedial spending.  In any event, looking over the list of high performing states, the secrets of their success are pretty evident.  These are the 17 states with the highest graduation rates in 2008---roughly the top third of all states.

Vermont	86.63
Wisconsin	85.57
Minnesota	85.38
New Jersey	85.21
Iowa	83.92
South Dakota	82.70
North Dakota	81.89
New Hampshire	80.62
Pennsylvania	79.52
Nebraska	79.46
Montana	78.76
Maine	78.67
Missouri	78.37
Kansas	78.23
Connecticut	78.12
Idaho	77.62

Basically, the highest performing states in terms of high school graduation are the upper Midwest and New England plus New Jersey (25th in per pupil expenditures) and Pennsylvania (41st in per pupil expenditures).  As a group, these first two sets of states states are richer, more racially and ethnically homogeneous, more sparely populated, less urban, and (not trivially) colder than the rest of the country.  As a descriptive datapoint, Vermont has the highest graduation rate in the country and the lowest adjusted per pupil education expenditures ($2,383---about 30% of per pupil spending in Texas).  Students in these states are (probably) much more likely to come from families with multigenerational histories of relatively high educational attainment and to develop personal expectations of scholastic achievement.

Now, look at the bottom 17 states in terms of high school graduation.

Washington	68.62
Hawaii	68.43
California	68.17
New York	67.13
Arizona	67.13
North Carolina	65.90
Alaska	65.82
Delaware	65.68
Texas	65.29
Alabama	64.03
Mississippi	61.69
New Mexico	60.61
Florida	59.56
Georgia	58.77
Louisiana	58.13
South Carolina	53.85
Nevada	47.56

There are dastardly  red states (Texas, Alabama, Mississippi) and sainted blue states (Hawaii, California, New York), eastern states (North Carolina and Florida) and western states (New Mexico and Nevada), northern states (Delaware) and southern states (Georgia and Louisiana).  In fact, the only link that spring to mind among the states are high levels of racial and ethnic diversity.

If I had to guess, high school graduation rates among middle class and wealthy white students are comparable in these states and in the Yankee states and frozen tundra states in the high achieving cohort.  The aggregate difference between states is due to the difficulty in delivering educational services to students in vulnerable and disadvantaged populations---namely the poor black, Latino, and indigenous populations that are scattered in high-density pockets throughout the country---regardless of how much money is spent on average per pupil or per person in the state.

I know much less about health outcomes the education outcomes---and don't have any data handy to support this intuition---but, I'd bet dollars to doughnuts that health outcomes among young people map pretty closely onto educational outcomes.  States with lots of vulnerable and disadvantaged young people probably have relatively poor aggregate health statistics for young populations regardless of the resources that are available for public health.  Aggregate health and educational outcomes at the state level are probably much more a product of economic and demographic characteristics than public policy choices.

All of which is to say that the state policy (i.e. spending) choices don't *cause* educational and health outcomes as Professor Krugman intimates.  All of which *is not to say* that Texas (or other states) should be as eager to cut spending on education and health care as it (which is to say, its elected officials) appear to be.  I would prefer a mix of tax increases (and spending from the rainy day fund) and spending cuts to either the current cut only proposals in the Texas legislature or the (I presume) tax only (or maybe tax and borrow) ideas percolating in Professor Krugman's mind.  Though, I think the preponderance of the state's fiscal imbalance should be remedied by reduced outlays.

Regardless of personal policy preferences, a newspaper editorial (or blog post) need not be a completely thorough analysis of any given policy problem.  However, Professor Krugman's column today seems to cherry picks data to make Texas look bad and ignores a variety of important factors aside from spending that bear on states' educational and health outcomes.  At a minimum, the implicit empirical claims in today's column are sloppy and unworthy of either Professor Krugman or The New York Times.  At worst, they willfully disingenuous.  Either is a shame.  Professor Krugman's academic position and Nobel Prize give his popular writing a veneer of dispassionate scholarship.  They deserve greater care.