This was originally published a few months ago, but the St. Louis Fed re-promoted it tonight on Twitter and I hadn't seen it before. Behold this scatterplot of state spending vs. state politics:
What this shows is that there's very little difference between red states and blue states when it comes to government spending. Wyoming, for example, is very red but spends 25% of its GDP. Massachusetts is very blue but spends only 17% of its GDP. The correlation between spending and political leanings is close to zero.
So if you want to live in a state that's tightfisted with its money, Georgia and Texas should be on your list. But so should New Hampshire and Connecticut and Delaware. Red states may talk a good game, but it turns out that on average they spend just as much as blue states.
How useful is state-level GDP, though? We have a national economy, not 50 state economies.
I think a better measure than % of GDP spent would be raw dollars per capita. MA may spend a smaller proportion of "its GDP" on its budget, but MA's GDP is $597M vs. MT's $40M. That's almost 15x as large, but MA only has a little over 11x the population.
So converting that to very rough raw dollars per capita and it's 25% * 40M = 10M spent on almost 600K people, or about $17 per person for WY. MA is 17% * 597M = 101.5M being spent on 6.8M people, or about $15 per person. Given that there are definitely economies of scale involved here (and that some of the services states have to provide are geographical in nature and that means larger states will have to spend more because intra-state regional services can only cover so much ground).... that seems like it may as well be about the same to me.
Err, I meant billions of dollars not millions. Add 3 zeroes.
Also for reference: GA = 626B, 10.6M people. That works out to about $9K per person.
TX = 1.776T, 29M people. That's almost $10K per person.
So OK, maybe there's something there for some states... but what if you control for population density?
How useful is state-level GDP, though?
I'd say very useful, if we're aiming to measure how much a state is willing to fund state/local government. There's no other way to do it.
One factor that should be noted, of course, is, depending on how this is measured, it could be the case that net-recipients of federal dollars (usually poorer, redder states) will get an "inflated" number that doesn't correspond very tightly to the burden of spending within the state. They're getting subsidized by the rest of the country. (Medicaid would be the biggest program reflective of this dynamic, I think).
I'm just wary of how much, say, large petroleum companies skew the GDP per capita of certain states (and perhaps other industries for other states). It's not really analogous to the GDP of a nation, but using the measuring stick of how much they're willing to fund state/local government . Basically what I'm saying is: we need to be looking at income of actual state residents (and resident corporations, to some degree, although no-income-tax states make this weird), because just using GDP doesn't really allow us to compare apples to apples. Especially since GDP counts for things in a kind of funny way at a state level (I mean, it's also kind of funny at a national level, but not so bad) when so many of the activities are really interstate commerce, but just so happen to be located in one state or another. And then there are the border metropolises like Kansas City, DC, NYC, etc., where there is just tons of bleed-through from state to state.
The article actually starts with per capita "As reported in the table below, we observe weak-to-moderate correlations between state political leanings and per capita state government expenditures. " then says GDP is better, "To control for both population and income, we can measure state government expenditures as a ratio of state GDP in 2018, the typical measure used in the literature to compare government expenditures across entities... When we measure expenditures in the preferred way, the correlations diminish in magnitude across the board (save for the transportation category)." However, I agree with you and other commenters here that there are confounding issues that make the GDP comparison less accurate. The most gaping hole is that a comparison based on % GDP ignores the likely correlation between increased GDP and investing in things like infrastructure and citizen well being. If I invest in a tractor and plow and my neighbor invests in a hoe it may be an equivalent investment as a percentage of our respective farm profits but I will have a much more comfortable life.
We have a world economy, not 200 national economies.
I don't actually believe that, but it makes as much sense as your claim that there are not 50 state economies.
I agree with cmayo on dollars-per-capita; it would be useful to see that charted. And in addition to land area that cmayo mentions, there are factors like the degree of urbanization and climate.
I have to admit I don’t have any idea of the breakdown of states’ spending, even for states I’ve lived in. I suppose road maintenance and education are big items, but how big? Pie chart? I guess I’ll have to look into this …
Answering my own question, taken from the FRB St. Louis page, the main categories of state and local spending are:
Education
Social Security and Income Maintenance
Transportation
Public Safety
Environment and Housing
Several confounds spring to mind: blue states have higher per-capita GDP (as NeilWilson notes in this thread); the geographically-smallest states are mostly blue (New England, Delaware, Maryland, New Jersey); and I believe blue states are generally more urbanized than red. Size seems like a major factor for transportation expenditures, and urbanization may make the delivery of agency-to-person services more efficient. I don’t see how any conclusions can be drawn about policy effects on spending until confounding factors are removed.
how is percentage of state GDP any kind of metric for comparing state spending?
within a single state, that metric may be useful for tracking spending over time, but apart from that it seems pretty useless.
the whole post seems kind of brain dead regarding the conclusions it draws.
how is percentage of state GDP any kind of metric for comparing state spending?
For the same reason we make such comparisons between countries. If Finland has a public sector that consumes 45% of GDP and Australia has one that consumes 34% of GDP, we can be reasonably certain that Finns have a bigger appetite for government than Australians.
I'd be careful throwing around the term "brain-dead" if I were you.
p.s. and that includes the title.
So...this doesn't include federally subsidized outlays....
The other issue is what are they spending money on? Everyone will do roads and bridges. For other expenditures, how much goes to large companies, i.e. their shareholders, and how much stays local?
“Everyone will do roads and bridges.”
Don’t be so sure. Pennsylvania for example seems to be allergic to ever widening any of its existing roads or bridges… or building new highways/arterials/bridges… or even improving intersections and exit ramps that are perennially congested and dangerous. With the exception of a single new highway about 5 miles in length that took 20+ years to plan and build and the addition of new “local traffic only” roads built by private developers to solely service new subdivisions, office parks or shopping centers they built, the county in which my mom’s family lives has the exact same road network (with each road having the same number of lanes) it did 50 years ago. I don’t know what PA spends its money on - perhaps pensions for former DOT workers? - but it certainly isn’t spending as much on actually improving its road network as any other state I’ve lived in.
Being allergic to widening roads is actually a great thing.
There's fairly compelling evidence that some states do not do bridge maintenance, or at least don't do enough bridge maintenance.
So really... our entire concept of fiscal conservatism is basically a joke. Well, this is what democrats are trying to change. Spend more on social programs, spend more on defense, spend more on infrastructure, spend more on tax cuts for parents, spend more in child care, and to heck with any sort of fiscal discipline. Such things are not required.
All this is fine with me... except defense spending.
Red states are poorer than blue states. (It's a fact. So there is no point in discussing it.)
Just as poor people need to spend more of their money on necessities, doesn't it follow that poor states need to spend more money on necessities?
Not sure where I am going with this. MA can afford to spend less of GDP on roads while KY needs to spend more (AS A % OF GDP) just to get acceptable roads.
So I think a comparison of raw $ spent would be helpful. Maybe one taking into account the average wage in the state too.
No trend line?
Also, if one looks at just the big, center cluster - eschewing the five or so outliers - does the conclusion hold?
It's not clear if "government spending" here refers to just the state budget, or all state, county, and local government services. Also, I would add that the biggest difference between red and blue states is not necessarily now much public spending there is, but who pays for it. As Kevin has often pointed out, Texas is a low-tax state...if you're rich. The poor and working class are actually the ones footing the bill for the roads, schools, hospitals, etc., whereas NY and California soak the rich to do that stuff.
I find this finding very interesting. Many of these 'Red' states feature no state income tax. It would be interesting to better understand the Red state municipal revenue model (extraction fees, property taxes etc).
Kevin shows the correlation based on spending normalized by state GDP, which the authors of the study prefer. But if is done simply on a per-capita basis, blue states spend more total per person since their GDP is higher on average. So if you are poor you may be better off in a blue state, because more will be spent on you - although cost of living may be lower in red states.
This effect might be somewhat reduced because the spending is by total population, not poor population. Red states tend to have more poor people and that would dilute the spending in those states (although California, for example, has a lot of poor people). Similar considerations may apply if you are rich - there are probably more rich people in blue states to spread the tax burden.
Yes, red states spend as much as blue states but only because they have a net inflow of federal dollars. Look at AK on the chart. They have long received all kinds of targeted federal money -- much for infrastructure -- that no other state receives. Think of Sarah Palin's Bridge to Nowhere. If you adjusted state spending to remove spending that relies on federal money, you would see that red states spend proportionately less of THEIR money than blue states.
The FRB-STL article reads well from an academic perspective, but the conclusion they draw about willingness to spend is deeply flawed and there is no acknowledgement of it in the article. No wonder they call economics the "dismal" science -- not so much for its content I think, but for the shallowness of its practice. IMHO, at least.
IIRC Alaska also gets a large chunk of change from the oil industry exploitation of the State's resources.
Article ignores the fact that Red states are much poorer than blue states. So when they spend the same per cent of the gdp, they're spending a lot less