About Election Interference
Spoilers:
- Election interference is very common. The US and the USSR interfered in one out of nine elections between 1946 and 2000.
- There's evidence to suggest that it is really effective, and that overt interference tends to work more effectively than covert interference.
The full movie:
Folks got a lot to say about election interference these days: the idea that Russia may have interfered in the US elections is one of the thousand scandals we’re juggling here in the States and a major impetus behind the current investigation of the sitting President. This whole debate is rich, because the US and the artist formerly known as the Soviet Union are well known habitual line steppers when it comes to international electoral politics.
Most importantly for us, their line stepping is not just the stuff of Twitter hot takes and the occasional zinger in an op-ed: it is also a fairly well studied phenomenon, owing a bit to the availability of information about election intervention in the States (check out this and this). So, I figure, this would be a good time to start up the social science section of this site.
Meet the our new friend, the paper “When the Great Power Gets a Vote: The Effects of Great Power Electoral Interventions on Election Results” by Dov Levin in International Studies Quarterly. The picture this paper paints is pretty grim: “Between 1946 and 2000, the United States and the USSR/Russia intervened in this manner 117 times, or, put another way, in about one of every nine competitive national-level executive elections during this period. Their methods ranged from providing funding for their preferred side’s campaign (a tactic employed by the Soviet Union in the 1958 Venezuelan elections…to public threats to cut off foreign aid in the event of victory by the disfavored side…” Levin finds that 69% of those interferences were done by the US.
- Election interference is very common. The US and the USSR interfered in one out of nine elections between 1946 and 2000.
- There's evidence to suggest that it is really effective, and that overt interference tends to work more effectively than covert interference.
The full movie:
Folks got a lot to say about election interference these days: the idea that Russia may have interfered in the US elections is one of the thousand scandals we’re juggling here in the States and a major impetus behind the current investigation of the sitting President. This whole debate is rich, because the US and the artist formerly known as the Soviet Union are well known habitual line steppers when it comes to international electoral politics.
Most importantly for us, their line stepping is not just the stuff of Twitter hot takes and the occasional zinger in an op-ed: it is also a fairly well studied phenomenon, owing a bit to the availability of information about election intervention in the States (check out this and this). So, I figure, this would be a good time to start up the social science section of this site.
Meet the our new friend, the paper “When the Great Power Gets a Vote: The Effects of Great Power Electoral Interventions on Election Results” by Dov Levin in International Studies Quarterly. The picture this paper paints is pretty grim: “Between 1946 and 2000, the United States and the USSR/Russia intervened in this manner 117 times, or, put another way, in about one of every nine competitive national-level executive elections during this period. Their methods ranged from providing funding for their preferred side’s campaign (a tactic employed by the Soviet Union in the 1958 Venezuelan elections…to public threats to cut off foreign aid in the event of victory by the disfavored side…” Levin finds that 69% of those interferences were done by the US.
What?
First up is the what – what does this paper set out to prove? It’s got two hypotheses. First, Hypothesis 1: “An electoral intervention for a particular candidate or party will increase its electoral chances.” Makes sense to me – it would be weird if anyone did something 117 times that didn’t work (but, then again, we have seen human organizations fuck up with that level of consistency, haven’t we?). Second, Hypothesis 2: “Overt electoral interventions are more likely to benefit the aided candidate or party than covert electoral interventions.”
The overt/covert distinction is more or less what you’re thinking. “overt” electoral interventions are the public kind, like getting on the country’s equivalent of CNN and yelling “if yall elect this motherfucker who wants to redistribute land instead of the one taking care of our money we’re going to stop sending food aid to your children”. Everybody knows what the deal is, and as far as the great power is concerned, people can feel some type of way about it if they want to. Covert electoral interventions are the kind that happen on the low, like sliding into the DMs of the electoral commission of the country you’re targeting and whispering “if yall elect this motherfucker who wants to redistribute land instead of the one taking care of our money we’re going to stop sending food aid to your children. Oh, and some of yall niggas is getting shot.”
First up is the what – what does this paper set out to prove? It’s got two hypotheses. First, Hypothesis 1: “An electoral intervention for a particular candidate or party will increase its electoral chances.” Makes sense to me – it would be weird if anyone did something 117 times that didn’t work (but, then again, we have seen human organizations fuck up with that level of consistency, haven’t we?). Second, Hypothesis 2: “Overt electoral interventions are more likely to benefit the aided candidate or party than covert electoral interventions.”
The overt/covert distinction is more or less what you’re thinking. “overt” electoral interventions are the public kind, like getting on the country’s equivalent of CNN and yelling “if yall elect this motherfucker who wants to redistribute land instead of the one taking care of our money we’re going to stop sending food aid to your children”. Everybody knows what the deal is, and as far as the great power is concerned, people can feel some type of way about it if they want to. Covert electoral interventions are the kind that happen on the low, like sliding into the DMs of the electoral commission of the country you’re targeting and whispering “if yall elect this motherfucker who wants to redistribute land instead of the one taking care of our money we’re going to stop sending food aid to your children. Oh, and some of yall niggas is getting shot.”
To all you murderous dictators out there out there, who don't wanna be on a record label where the executive producer's all up in the videos, come to White House Records.
How?
How does Levin investigate these hypotheses? Levin did a quantitative study using a dataset of national elections developed by Dieter Nohlen and a team of researchers. They spent two decades tracking down data for all of the thousands of elections since 1946, and coding them in a way that made it possible to run statistical analyses that test out ideas like the hypotheses in the last section. Using that, Levin constructs their own dataset: the same data about national elections, whittled down to the elections where either the US or USSR interfered.
I have to note, since I’m on my Third World shit today: that is to say that all this was possible because of a vast investment of time and resources – not only by these researchers themselves but by the institutions that housed and financed them and the international community of researchers which made this worth doing by having the right kinds of interests and research questions. Imagine what hypotheses we could test, what datasets we would have if those institutions and researchers weren’t so skewed heavily towards the First World?
But I’m getting off track. Okay, here’s how Levin tests the hypotheses. They create sets of statistical models: mathematical representations of what seems to affect election outcomes, basically lists of things that seem to matter for predicting the data, weighted by numbers that represent how much they effect the results – a country’s GDP, its wealth, the international policy context (i.e. is this election happening during the Cold War or nah?). Those things are “independent variables”, things being analyzed for how they affect something else. The dependent variable in all of the models – the thing being affected by the list of (potentially) causal stuff in the model – was the incumbent candidate or party’s “vote share” (the percentage of the total vote that they got).
They include in these models a variable for covert election interference and another one for overt interference (basically they just put them on the lists of things the computer should use to analyze what explains election outcomes). Then they test these models on “intervenable” elections – elections that were competitive enough that its worth thinking about how intervention might have affected them. Then, these statistical models should be able to give us an educated guess about how much interventions effect vote share of incumbent parties – that is, whether or not they swing voting behavior elections, and by how much.
Finally, Levin uses 10 separate models – i.e., tests the hypothesis again 10 different weighted lists of what might matters for predicting an incumbent candidate’s vote share. These help make sure that their findings are “robust” – basically, that they aren’t just getting the results they are looking for because of particularities about how they set up the model. Their results hold up, more or less, in all ten models.
If either independent variable (the ones being analyzed for their effect on the dependent variable) for intervention is positive and statistically significant, that’s evidence that hypothesis 1 is right. If the effect of the variable for overt interference is larger than the effect of the variable for covert interference, that’s evidence that hypothesis 2 is right.
What happened?
Hypothesis 1: “On average, an electoral intervention in favor of one side contesting the election will increase its vote share by about 3 percent—quite a significant effect. For example, such a swing in the vote share from the winner to the loser in the fourteen US presidential elections occurring since 1960 would have been sufficient to change the identity of the winner in seven of these elections.”
Hypothesis 2: “As can be seen from Model 1, the results support Hypothesis 2, with overt interventions on average increasing vote share by 3 percent more than covert interventions. This is a major substantive difference in the effects of each subtype.”
How does Levin investigate these hypotheses? Levin did a quantitative study using a dataset of national elections developed by Dieter Nohlen and a team of researchers. They spent two decades tracking down data for all of the thousands of elections since 1946, and coding them in a way that made it possible to run statistical analyses that test out ideas like the hypotheses in the last section. Using that, Levin constructs their own dataset: the same data about national elections, whittled down to the elections where either the US or USSR interfered.
I have to note, since I’m on my Third World shit today: that is to say that all this was possible because of a vast investment of time and resources – not only by these researchers themselves but by the institutions that housed and financed them and the international community of researchers which made this worth doing by having the right kinds of interests and research questions. Imagine what hypotheses we could test, what datasets we would have if those institutions and researchers weren’t so skewed heavily towards the First World?
But I’m getting off track. Okay, here’s how Levin tests the hypotheses. They create sets of statistical models: mathematical representations of what seems to affect election outcomes, basically lists of things that seem to matter for predicting the data, weighted by numbers that represent how much they effect the results – a country’s GDP, its wealth, the international policy context (i.e. is this election happening during the Cold War or nah?). Those things are “independent variables”, things being analyzed for how they affect something else. The dependent variable in all of the models – the thing being affected by the list of (potentially) causal stuff in the model – was the incumbent candidate or party’s “vote share” (the percentage of the total vote that they got).
They include in these models a variable for covert election interference and another one for overt interference (basically they just put them on the lists of things the computer should use to analyze what explains election outcomes). Then they test these models on “intervenable” elections – elections that were competitive enough that its worth thinking about how intervention might have affected them. Then, these statistical models should be able to give us an educated guess about how much interventions effect vote share of incumbent parties – that is, whether or not they swing voting behavior elections, and by how much.
Finally, Levin uses 10 separate models – i.e., tests the hypothesis again 10 different weighted lists of what might matters for predicting an incumbent candidate’s vote share. These help make sure that their findings are “robust” – basically, that they aren’t just getting the results they are looking for because of particularities about how they set up the model. Their results hold up, more or less, in all ten models.
If either independent variable (the ones being analyzed for their effect on the dependent variable) for intervention is positive and statistically significant, that’s evidence that hypothesis 1 is right. If the effect of the variable for overt interference is larger than the effect of the variable for covert interference, that’s evidence that hypothesis 2 is right.
What happened?
Hypothesis 1: “On average, an electoral intervention in favor of one side contesting the election will increase its vote share by about 3 percent—quite a significant effect. For example, such a swing in the vote share from the winner to the loser in the fourteen US presidential elections occurring since 1960 would have been sufficient to change the identity of the winner in seven of these elections.”
Hypothesis 2: “As can be seen from Model 1, the results support Hypothesis 2, with overt interventions on average increasing vote share by 3 percent more than covert interventions. This is a major substantive difference in the effects of each subtype.”
“Its effects also stack up quite well in comparison to another well-known major effect on election results: the state of the economy. As can be seen in Model 3 (where the interactions of the economic growth variable with two other variables are excluded), a 1 percent increase in the real GDP per capita would increase incumbent vote share by about 0.4 percent. To illustrate its effects, had Carter in 1980 run for reelection with the economy that Reagan had in 1984 (+6.6 percent) rather than the economy he actually had in 1980 (−1.9 percent), my model would have predicted Carter’s vote share to have increased by about 3.4 percent.”
