---------------------------------------------------------------------------
---------------------------------------------------------------------------
Code and data to construct tables in:
Francis, Jackson, Owyang (2017) JMCB
"Countercyclical Policy and the Speed of Recovery After Recessions"

---------------------------------------------------------------------------
Data:
FJO2017_Data.xlsx: State-recession specific data for employment and unemployment recessions
StateEMP_Data_1939to2012.xlsx: State employment data from 1939-2012 for constructing 
	recession/recoveries for Table 1
StateEMP_Data.xlsx: State employment data from 1979-2012 for constructing recession/
	recoveries for model estimation
StateUR_Data.xlsx: State unemployment rate data from 1979-2012 for constructing recession/
	recoveries for robustness test model estimation

***Data sources described in the paper***

---------------------------------------------------------------------------
Code for building Table 1: ConstructRecessRecovEmpLoss_1939to2012.m
Constructs recession/recovery experiences for 48 contiguous U.S. states using
employment data from 1939-2012

Building recession/recovery experiences for 1979-2012 to be used for model estimation:
 - ConstructRecessRecov_EMP.m (Baseline model with employment data)
 - ConstructRecessRecov_UR.m  (Robustness test with unemployment rate data)

---------------------------------------------------------------------------
Main code for estimation: MainModelEstimation.m
**Calls subfunctions to conduct estimation

1. Reads in data and constructs summary statistics (mean & standard deviation) for Table 2


2. Estimates baseline AFT model with employment recessions
	- baseline set of potential covariates
	- state-level fixed effects
	- 100% inclusion of policy variables

3. Estimates AFT model with unemployment recessions

4. Estimates AFT model with employment recessions and a variety of policy robustness tests
	- Pre-Great Recession observations only
	- Romer narrative tax shocks (Pre-2008)
	- Wu-Xia shadow rate for FFR
	- Nakamura-Steinnson military spending (Ends 2006)
	- State ARRA fiscal spending

5. Computes Marginal Effects
	*Construct expected recovery times under different scenarios
	(1) All covariates at the mean
	(2) Long recession / Other covariates at the mean
	(3) Short recession / Other covariates at the mean
	(4) Long recession / Accomodative Policy / Other covariates at the mean
	(5) Short recession / Accomodative Policy / Other covariates at the mean
	(6) Long recession / Tight Policy / Other covariates at the mean
	(7) Short recession / Tight Policy / Other covariates at the mean
