Estimate (using Stata) the three regression models given below and perform appropriate
hypothesis tests and diagnostic tests
The data to be used for the estimation is in the file projectdatas24.dta and includes the
period January 2007 – January 2024. All the variables are from the FRED database
https://fred.stlouisfed.org/
Model 1
realcei = β0 + β1 realpdii + β2 homepii + β3 csenti + ui (1)
Model 2
Add an independent variable to the model given by equation (1) in order to determine if the
average level of realce (US real personal consumption expenditures) during a recession is lower
than the average level of realce during a non-recessionary period.
Model 3
Add an independent variable to the model given by equation (1) in order to determine if the
change in realce resulting from an addition $1 billion of realpdi during a recession is smaller
than the change in realce resulting from an addition $1 billion of realpdi during a non-
recessionary period.
(Hint: this variable can be generated on the basis of variables in the project data file
projectdatas24.dta posted on Canvas).
where:
realce = US real personal consumption expenditures (billions of chained 2017 dollars)
realpdi = US real disposable personal Income (billions of chained 2017 dollars)
homepi = S&P/Case-Shiller US Home Price Index, seasonally adjusted (index Jan 2000=100)
csent = University of Michigan Consumer Sentiment Index (index 1966:Q1=100)
rec = 1 if the U.S. economy was in a recession, 0 otherwise
Data sources:
realce : FRED database (series PCEC96)
realpdi : FRED database (series DSPIC96)
homepi: FRED database (series CSUSHPISA)
csent: FRED database (series UMCSENT)
rec: FRED database (series USREC)
2. Write a 3-6 page project report (excluding the title page). In the report students are required to:
a. Describe a model of consumption expenditures and the theoretical foundation for the model
b. Specify the econometric model — what is the dependent variable, what are the independent
variables in the regression model, alternative specifications adopted (e.g. Models 1, 2, 3 above).
c. Describe the data sources and the variables used in the research
d. Report the results of each estimated model (the estimated equations or a table that contains
signs and magnitudes of the estimated coefficients along with their std. errors, F-statistics, R2 )
and interpret these results in detail.
e. Perform the diagnostic tests for Model 2 or Model 3 (for multicollinearity, specification errors,
and autocorrelation).
f. Analyze the empirical findings based on the econometric estimation and the hypothesis tests.
Based on the results of the diagnostic tests, provide suggestions (if necessary) for improving
your model of U.S. consumption expenditures.
1. Submit the following files:
a. Stata log file (.smcl format) containing the commands and estimation results
Category: Econmetrics
Estimate (using Stata) the three regression models given below and perform appropriate
hypothesis tests and diagnostic tests
The data to be used for the estimation is in the file projectdatas24.dta and includes the
period January 2007 – January 2024. All the variables are from the FRED database
https://fred.stlouisfed.org/
Model 1
realcei = β0 + β1 realpdii + β2 homepii + β3 csenti + ui (1)
Model 2
Add an independent variable to the model given by equation (1) in order to determine if the
average level of realce (US real personal consumption expenditures) during a recession is lower
than the average level of realce during a non-recessionary period.
Model 3
Add an independent variable to the model given by equation (1) in order to determine if the
change in realce resulting from an addition $1 billion of realpdi during a recession is smaller
than the change in realce resulting from an addition $1 billion of realpdi during a non-
recessionary period.
(Hint: this variable can be generated on the basis of variables in the project data file
projectdatas24.dta posted on Canvas).
where:
realce = US real personal consumption expenditures (billions of chained 2017 dollars)
realpdi = US real disposable personal Income (billions of chained 2017 dollars)
homepi = S&P/Case-Shiller US Home Price Index, seasonally adjusted (index Jan 2000=100)
csent = University of Michigan Consumer Sentiment Index (index 1966:Q1=100)
rec = 1 if the U.S. economy was in a recession, 0 otherwise
Data sources:
realce : FRED database (series PCEC96)
realpdi : FRED database (series DSPIC96)
homepi: FRED database (series CSUSHPISA)
csent: FRED database (series UMCSENT)
rec: FRED database (series USREC)
2. Write a 3-6 page project report (excluding the title page). In the report students are required to:
a. Describe a model of consumption expenditures and the theoretical foundation for the model
b. Specify the econometric model — what is the dependent variable, what are the independent
variables in the regression model, alternative specifications adopted (e.g. Models 1, 2, 3 above).
c. Describe the data sources and the variables used in the research
d. Report the results of each estimated model (the estimated equations or a table that contains
signs and magnitudes of the estimated coefficients along with their std. errors, F-statistics, R2 )
and interpret these results in detail.
e. Perform the diagnostic tests for Model 2 or Model 3 (for multicollinearity, specification errors,
and autocorrelation).
f. Analyze the empirical findings based on the econometric estimation and the hypothesis tests.
Based on the results of the diagnostic tests, provide suggestions (if necessary) for improving
your model of U.S. consumption expenditures.
1. Submit the following files:
a. Stata log file (.smcl format) containing the commands and estimation results