In late March, Georgia passed a restrictive new voting law that, in effect, permits the Republican state legislature to put partisan operatives in charge of disqualifying ballots in Democratic-leaning precincts. The law is one of at least eight proposals from GOP lawmakers in state legislatures around the country for increasing partisan influence over electoral administration — and one of more than 360 state bills that would curtail voting rights in one way or another.
New political science research suggests this wave of attempts to restrict the franchise is not an anomaly: Republican control over state government is correlated with large and measurable declines in the health of a state’s democracy.
The paper, by University of Washington professor Jake Grumbach, constructs a quantitative measure of democratic health at the state level in the US. He looked at all 50 states between 2000 and 2018 to figure out why some states got more democratic over this period and others less. The conclusions were clear: The GOP is the problem.
While many researchers have attempted to quantify the health of democracy in different countries around the world, Grumbach’s paper is the first effort to develop some kind of ranking system for US states. It’s still in working paper form, which means it has not been peer-reviewed. But Grumbach’s work has been widely praised by other political scientists who had read a draft or seen him present it at a conference.
“This is one of those papers that makes me proud to be an empirical political scientist. It’s important, carefully done, and just plain smart,” writes Vin Arceneaux, a professor at Temple University. “It helps us not only understand American politics but democratic backsliding in general.”
And it’s yet another piece of evidence that the Republican Party has become an anti-democratic political faction.
What the paper found
When I spoke to Grumbach on the phone about his work, he explained that his approach was inspired by V-Dem, the political science gold standard for quantitatively measuring democracy in countries around the world.
V-Dem breaks down democracy into individual component parts, like whether the press is free or the people can assemble peacefully, which can be measured and added together to produce an overall “democracy” score for any one country. You can’t just apply this to American states directly; no place in the United States is violently repressive in the way that China or Russia is, so the measurements might not be precise enough to clearly illustrate differences between states.
Grumbach’s State Democracy Index (SDI) is the first attempt to use a V-Dem-style approach to measure the more subtle ailments afflicting democracy in the United States. Metrics include the extent to which a state is gerrymandered at the federal level, whether it permits same-day voter registration, and whether felons are permitted to vote. He also includes criminal justice indicators, like a state’s Black incarceration rate, that are designed to measure state coercion.
To turn these metrics into an actual score, Grumbach uses a process that’s part subjective and part algorithmic.
The subjective part strives to determine whether an individual practice, like voter ID laws, is helpful or harmful to democracy. Grumbach then uses an algorithm to determine how much each of these practices should count toward a state’s overall score, either negatively or positively. This automated process ended up downplaying the criminal justice metrics, which barely factor into a state’s ultimate score. By contrast, the algorithm gave significant weight to electoral practices like gerrymandering (negative) and same-day registration (positive).
With a system in hand, Grumbach then proceeded to score all 50 states in every year between 2000 and 2018, from -3 (worst) to 2 (best). The following maps show the changes in state scores; the lighter the color, the worse the score is.
Just eyeballing the map, you can pick up on the clear pattern of states with Republican governments scoring lower in 2018 than they did in 2000. That year, the average GOP-controlled state was slightly more small-d democratic than its average Democratic-controlled peer. Over the next two decades, the average Republican score declined dramatically.
There could be all sorts of reasons why this is happening. Maybe Republican states had high levels of demographic change, causing white voters to embrace voter suppression in response. Maybe Republicans won power in states in which the parties were highly polarized, which raised the stakes of political conflict and caused incumbents to try to secure their hold on power.
To test these theories, Grumbach ran a series of regression analyses designed to isolate correlations between a state’s democracy score and variables like percentage of nonwhite voters and measures of state-level polarization. Strikingly, these things either barely mattered or didn’t matter at all.
Only two things really did: whether a state was controlled by Republicans and whether Republicans had gained that control recently. Republican-controlled states in general were far more likely to perform worse in the State Democracy Index over time; Republican states with a recent history of close elections, like Wisconsin and North Carolina, were especially likely to decline from 2000 to 2018.
“Regardless of the particular circumstances or geography, state governments controlled by same party behave similarly when they take power,” Grumbach writes. “The Republican controlled governments of states as distinct as Alabama, Wisconsin, Ohio, and North Carolina have taken similar actions with respect to democratic institutions.”
The GOP’s anti-democratic agenda is real
To be sure, this is just one study and not yet peer-reviewed, to boot, so it shouldn’t be taken as definitive. And skeptics can certainly poke holes in some of Grumbach’s choices.
For instance, one could question just how bad the anti-democratic infractions Grumbach cites really are. Voter ID laws, to take one example, counted against many Republican states — but the evidence on whether they actually reduce voter turnout is surprisingly mixed. A conservative might object that Grumbach is making a mountain out of a molehill: creating a metric that makes Republican states look worse than Democratic ones when, in reality, the differences just aren’t that big.
But methodological quibbles aside, the paper does seem to capture something real. One of its most valuable contributions is the way it treats gerrymandering.
Drawing lines to give your party a leg up disproportionate to its numbers is one of those practices that no one can really defend in democratic terms. Elected authoritarians abroad, like Hungary’s Viktor Orbán, have abused gerrymanders to ensure that they maintain a hammerlock on national power despite winning less than a majority of national votes.
The SDI shows that, in the United States, gerrymandering is not a “both sides” problem. It uses 16 different measurements of gerrymandering to assess how prevalent the tactic is in different states; 10 of these measures are the most heavily weighted factors in a state’s ultimate democracy score. These metrics show that Republican legislators abuse gerrymandering in a way that Democrats simply do not.
Some of this abuse has happened quite recently. Take a look at North Carolina’s SDI score over time — it starts to plunge shortly after Republicans drew new maps in 2011, ones that allowed them to win 77 percent of the state’s House seats in 2018 with just under 50 percent of the state vote:
(It’s worth noting that one of the worst abuses by North Carolina’s Republican legislature — stripping Democratic Gov. Roy Cooper of key powers after his 2016 victory — doesn’t factor into the state’s score, as Grumbach hasn’t decided on a satisfactory way to quantify it. A similar maneuver performed by the Wisconsin state legislature after Democratic Gov. Tony Evers’s victory in 2018 also doesn’t count against the state’s already dismal SDI score.)
Offering empirical ballast for a phenomenon we’ve observed in real time these past few years, Grumbach’s paper is another passage in the dominant political story of our time: the Republican Party’s drift against democracy. And, crucially, it’s a drift that began before Trump and his allegations of fraud in 2020. Republicans didn’t need Trump to enact extreme gerrymandering after the 2010 census; his anti-democratic instincts helped bring out something that was already there.
We have every reason to expect things will get worse, not better.
Georgia’s law, for example, is more worrying than even voter ID laws. It gives Republicans more direct control over election administration, allowing them to bend the rules in their favor: enforcing strict standards for ballot disqualification in Democratic-leaning precincts and lax ones in Republican-leaning ones, for example.
Once 2020 census data is available, states will do another round of redistricting. This time, those who want to tilt the field in their favor will have a freer hand due to a 2019 Supreme Court ruling that partisan gerrymanders can’t be stopped by courts.
There’s a reason that Grumbach calls states, in the paper’s title, “Laboratories of Democratic Backsliding.” Republicans have been engaging in a series of grand experiments in rigging a political system one state at a time — one that is, slowly but surely, undermining the foundations of America’s free political system.