*******************************************************************************
*Project: Devins_Lewis 2022
*Author: David Lewis
*Date: February 6, 2022
*Data: plum_combined_000816_021921.xlsx
*Notes: This do file is to conduct the analysis on nominations for Devins Lewis 2022. 
*Data and codebook for all 3 presidencies available.
*******************************************************************************
import excel "C:\Users\lewisde\Dropbox\Pace of Nomination\Combined Plum Book Data\plum_combined_000816_021921.xlsx", sheet("Sheet1") firstrow

**We drop dual-hatted positions so that we do not double count. Some positions are filled
**by persons in other PAS positions, either automatically or by custom. For example, there are a number of ambassadorial positions where an ambassador confirmed *to one position generally becomes ambassador to another country at the same time. For example, a recent nomination for Ambassador to Barbados reads:
*Leandro Rizzuto, of New Jersey, to be Ambassador Extraordinary and Plenipotentiary of the United States of America to Barbados, and to serve concurrently and *without additional compensation as Ambassador Extraordinary and Plenipotentiary of the United States of America to the Federation of Saint Kitts and Nevis, *Saint Lucia, Antigua and Barbuda, the Commonwealth of Dominica, Grenada, and Saint Vincent and the Grenadines.

drop if concurrent==1
summarize

**Now I need to create an variable to separate out the major independents
**Here is an indicator for the big independent regulatory commissions.

generate irc=0
replace irc=1 if dep=="CFTC"|dep=="CPSC"|dep=="FCC"|dep=="FTC"|dep=="Fed"| dep=="EEOC"|dep=="NLRB"|dep=="NRC"|dep=="SEC"|dep=="FEC"|dep=="NTSB"|dep=="FERC"|dep=="FED"


**First, we look at average days vacant over the first two years. Positions not filled assumed to have maximum possible days (e.g., 730 days over two years).
bysort year: ttest days if vacant==1, by(indcom)
bysort year: ttest days if vacant==1&indcom==1, by(irc)


**Second, we look at the proportion of vacancies in each category receiving a nominee.
bysort year: ttest nominee if vacant==1, by(indcom)
bysort year: ttest nominee if vacant==1&indcom==1, by(irc)