Good day to all! The purpose of my research is to determine the influence of various factors on the percentage of votes in the election. The problem is that the dependent variable is not only the percentage of votes, but also the relationship between the votes and the years before the election.
Here is what I mean:
What formula for panel regression can I use? I just used fixed effects only on the regions, I would like to leave them. I would like to apply interaction effects to other predictors. And can this be added to this indicator by the number of years before the election? And is it possible to create a model where all these effects will be together?
Here is what I mean:
| year | Percentage of votes | Years before the election | Other predictors | |
| Belgorod region | 2016 | 62,36 | 4 | ||
| 2017 | 62,36 | 3 | ||
| 2018 | 62,36 | 2 | ||
| 2019 | 62,36 | 1 | ||
| 2016 | 71,9 | 3 | ||
| 2017 | 71,9 | 2 | ||
| 2018 | 71,9 | 1 | ||
| 2019 | 63,71 | 0 | ||
| Vladimir region | 2016 |
| 2 |
What formula for panel regression can I use? I just used fixed effects only on the regions, I would like to leave them. I would like to apply interaction effects to other predictors. And can this be added to this indicator by the number of years before the election? And is it possible to create a model where all these effects will be together?