This appendix provides information about the nature and sources of the quantitative data that I used throughout the book. Although some issues of measurement and operationalization are discussed extensively earlier in the text (specifically with regard to the dependent variables in the analysis), in some cases I omitted a thorough discussion of the independent and control variables to aid the book's exposition. Below, I discuss those variables in the order that they appear in the text.

Chapter 3

In this chapter, I examined whether changes in executive and legislative power were correlated with regime-compatible democracy assistance. The data on changes in countries’ executive leaders came from the Archigos data set. I relied upon the variable “Exit” to identify leaders who exited from office for reasons other than natural causes. The data on whether incumbents and incumbent parties won national elections came from the National Elections Across Democracy and Autocracy (NELDA) database. Specifically, I used variable NELDA24 (“Did the incumbent's party lose?”).

Later in the chapter, I calculated the proportions of elections that seemed unlikely beforehand to be free and fair and that monitors refused to observe. For those calculations, I again used NELDA; the specific variables used were NELDA11 (“Before elections, are there significant concerns that elections will not be free and fair?”) and NELDA49 (“Did any monitors refuse to go to an election because they believed that it would not be free and fair?”).

Chapter 4

In this chapter, I examined how delegation relationships shape the design and implementation of democracy assistance. Chapter 4 contains information about the nature and sources of the data for the dependent variables.

I used three different indicators of delegation relations in democracy assistance. The first indicator was Donor-country NGOs and applied to aid being channeled via NGOs. This dichotomous variable was coded 1 if a project was implemented by a donor-country NGO and 0 otherwise.