Geographic polarization of U.S. science

Is science as segregated as our politics?

Even before the election I was thinking about how clustered our scientific institutions are. Now, in light of Trump’s surprising victory (which I still haven’t really processed) it’s hard not to wonder if science is following the pattern of ‘coastal elites’ with tensions of science communication driven in part by geography.

The United States invests over 300 billion dollars a year in science to generate the next cure, technology or insight. But where that funding goes has huge implications for what science happens, which institutions remain research powerhouses, and who is exposed to more science.

If we start by looking at the total science funding by each U.S. state we can map it onto their approximate locations for an easy visualization of how the money is spread out.

Total funding includes government, industry, & universities as compiled by Research!America. Squares are only approximately centered on states’ geographic centers to avoid overlapping squares.

The East Coast clearly has the highest density of funding which might be expected due to high density of other things like people and universities. California has the most science funding of all the states but that probably makes sense since it’s the most populous state.

You could imagine these differences being mostly due to the number of people and hence number of scientists, but looking at the map you can already see that a state like Texas (~27 million people) gets less funding than Massachusetts (~7 million people).

So what happens when we look at the amount of science funding per person? Well, we start to see large disparities with some places getting much more science money than their size would suggest.

The shape of this distribution is telling. Lots of states get below average for science money per person, while only a few states get way more funds than the average.

It gets even more pronounced if we include Washington DC in this analysis. DC gets more than two and a half times the science dollars per person that the highest state gets.

So what causes the disparity of science funds among states? Are there common demographic or geographic causes?

Geographically, the Northeast and Pacific Coast do the best while the South struggles to attract science funding even when corrected for population. It’s possible this could be explained by education level, populous centers that bring enough people together for collaboration, or the number of institutions where scientists could work, so I tried to find demographic data to compare to funding levels.

My best attempts at finding a clear relationship (political lean, college enrollment rate, population density, and academic institutions per person) show no particularly strong correlations, but somewhat surprisingly political lean came out on top.

Looking at a linear regression for political lean, college enrollment, population density, and number of academic institutions gave r² values of 0.36, 0.25, 0.034, and 0.015 respectively. Political lean (as defined here by percent win/loss for Obama in 2012) shows the strongest (but still fairly weak) relationship, with larger Obama wins being somewhat correlated with more investment in science.

We can see that Washington DC is again a major outlier, and we see that the linear regression does not show a particularly convincing relationship. The connection between political makeup and science funding might be weak, but any possibility that interactions with science could become more defined by geography or politics should be concerning to a field that depends crucially on public investment.

While we often think about the amount of science funding and the type of research being funded, we need to address what happens when there are huge disparities in WHERE science research is being funded.

Scientists are more likely to collaborate with other nearby scientists, and cities produce a huge amount of our national science. Many ideas and research partnerships are still started by in person meetings. All of these make geography crucial to how science develops.

Even beyond the actual research, the abundance of scientists in a region will affect the number of people who can do scientific outreach or be role models for local students. From my own experience, I can see how much more students in the Boston area interact with scientists than when I grew up in Evansville, Indiana. Now, in Cambridge Massachusetts, it’s hard for me to not bump into a fifty scientists when trying to grab a sandwich in Kendall Square.

While the election has pundits writing plenty about sorting geographically into homogeneous bubbles by our politics, there are many other levels on which we fail to communicate with people who are not like us.

It’s easy for something to be dismissed when it’s not present around you in your community. I fear, science is not immune to that.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.