You want to estimate the effect of the construction of a garbage incinerator on

Question

You want to estimate the effect of the construction of a garbage incinerator on house prices in a neighborhood. Suppose that there are rumors that the incinerator would be built starting in 2011, but the actual construction of the incinerator began in 2013, and finished that same year. Your hypothesis is that the price of houses near the incinerator would fall, relative to houses further away from the incinerator.

a) Suppose you run a simple regression of house prices in 2014 (dependent variable) on distance to the incinerator (main independent variable) and other covariates. You find that distance to the incinerator is a major predictor of house prices, but your colleagues argue that you are overestimating the effect of the distance to the incinerator. Explain why they believe you are overestimating the effect of distance to the incinerator, by discussing omitted variable bias, and explaining how you can establish the sign of the bias.

(b) Suppose your colleagues suggest that you can address the omitted variable problem if you used difference-in-differences estimation. They argue that you would need at least two years of data. What type of data would you need to collect (panel/cross sectional/either), and for which years (suppose that data collection is expensive, so you need to pick only two years of data)? Explain how you would use the difference-in-differences method, and how it addresses omitted variable bias. What are the treatment and control groups in your difference-in-differences quasi-experimental estimation?

(c) Which regression would you run to obtain the difference-in-differences estimate that you described in part b? Specify all the dependent and independent variables.

Explanation / Answer

You want to estimate the effect of the construction of a garbage incinerator on house prices in a neighborhood for this purpose you set the hypothesis is that the price of houses near the incinerator would fall, relative to houses further away from the incinerator.

you run a simple regression of house prices in 2014 (dependent variable) on distance to the incinerator (main independent variable) and other covariates. You find that distance to the incinerator is a major predictor of house prices, but your colleagues argue that you are over estimating the effect of the distance to the incinerator because of the presence of the linear relation ship between independent varialbes there is presence of bias in the estimatyes of independent variables the bias may be negative or positive we diagnos it and then refit the models for better accuracy using R2 adjusted method. also there is presence of multicolineartiy between regressors due to linear relation ship between independent variables. in this situation we need to independent variable which is highlly correlated.  

colleagues suggest that to address the omitted variable problem if you used difference-in-differences estimation. They argue that you would need at least two years of data. In contrast to a time series estimate of the treatment effect on subjects which analyze differences over time or cross section estimates of the treatment effect which measure difference between the treatment and control groups , difference in difference uses panel data to measure the differences.

Related Questions

Navigate

• Browse (All)
• Browse Y
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.