I looked at how Brown, Klobuchar, Warren, Sen. Kirsten Gillibrand of New York, Texas' Beto O'Rourke and Sen. Bernie Sanders of Vermont did in their 2018 Senate races. Then I examined how Democratic US House candidates in those same states performed.
In other words, Enten took the percentage of votes that Democratic Congressional candidates won in the state as a baseline for candidate performance in the state and compared the percentage each Senator won to it. The implicit inference here is that if the Senator did better the baseline Democratic Congressional candidates in their state, then they are demonstrating a superior ability to win votes and hence are more electable.
While this inference would get more discussion in a scholarly paper, Enten is doing the kind of thing that social scientists do when trying to gain insight on future events (notice how I avoided using the word "predict" there), That is, they look to past events as an indicator of future ones and they identify a research question that can be answered through a systematic analysis of available data. This analysis an objective empirical finding, or what I like to call a " hard kernel of empirical truth," In the case of Enten's article, it is the findings reported in the following table:.
Most 2020 Democratic contenders outperformed the House baseline in their 2018 Senate races
2020 contenders Sherrod Brown and Amy Klobuchar strongly outperformed the baseline in 2018.
Democratic Senate candidate
|
Senate result*
|
House result*,**
|
Senate overperformance
|
|---|---|---|---|
| Amy Klobuchar, Minnesota | D +24 | D +11 | +13 |
| Sherrod Brown, Ohio | D +7 | R +5 | +12 |
| Kirsten Gillibrand, New York | D +34 | D +29 | +5 |
| Beto O'Rourke, Texas | R +3 | R +6 | +3 |
| Bernie Sanders***, Vermont | I +40 | D +43 | -3 |
| Elizabeth Warren, Massachusetts | D +24 | D +36 | -12 |
Note that, by themselves, these empirical findings do not address the central question of the article, i.e., how electable will these Senators be in the 2020 Presidential election. Instead, they provide the basis for making inferences about how these Senators might perform, especially in comparison to one another. Thus, the hard kernel of empirical truth provides a springboard for making inferential leaps to a conclusion about the primary topic of interest. This is just the sort of things social scientists do all he tine.
