Assignment #3

PPol 604

Due: Thursday, 31 January 2013

Type up your answers. Give proper credit to those you work with and/or the text(s).

Solve the following problems. Show all of your work, but keep your answers concise. Highlight your (final) answer to distinguish it from your other numbers and text. Include a copy of your input (e.g. do file) or output (e.g. log file), when it is an appropriate way to show your work. However, do not include unnecessary output (i.e. no data dumps), and format any output so that it is easily readable. An appropriate time to include output is when you put your results in a table--if your results are wrong, then the grader has no idea how you came to your conclusions (i.e. give partial credit) unless you provide some output. Explanation includes statistical and substantive explanation (explain so that a statistical layperson can understand it, and so that a statistical analyst will see your erudition).

1. {5} Do Problem 12.9 in Stock and Watson.
2. {30} [from Bartels 1991] Use the Stata dataset here (spending86.dta) to solve this problem. This dataset contains information on 250 non-freshmen incumbents in the 1986 election who had major-party challengers in the 1986 and 1984 elections (excluding a couple of outliers). See the data set for variable descriptions. Note that the spending variables are logged to reflect diminishing returns to spending. Investigate how incumbent and challenger spending affect the percentage of the vote that the incumbent receives. Put your results in a table with four columns for parts b, e, f, and h.
   a. Before running a regression, hypothesize how party (democrat), challenger quality (hq), previous vote (vote_1), incumbent spending (lispend), and challenger spending (lcspend) will affect the two-party vote percentage, vote, and explain why.
   b. Run an OLS regression using the variables in part a. (do not worry about endogeneity yet). Compare your hypotheses with the results. If there are any differences, why?
   c. Give two reasons why incumbent spending might be endogenous.
   d. The variable incumbent spending in the previous election cycle (lispend1) is proposed as an instrument for incumbent spending in the current election cycle. Which reason(s) of endogeneity might this fix? Why?
   e. Conduct TSLS (sometimes called 2SLS) "by hand," i.e. you run the two stages yourself, without using the Stata command ivregress (or commands like it).
   f. Conduct TSLS using the Stata command ivregress 2sls. Compare the results to the previous two parts. Interpret and explain (statistically and substantively) your findings. Compare your hypotheses with the results. If there are any differences, why?
   g. Conduct a test whether incumbent spending is endogenous. Use whatever tests you can for instrument validity. If you cannot include a test, explain why.
   h. There is a theory that challenger quality does not directly affect the vote, but only affects it through incumbent spending. Thus, challenger quality could be used as an instrument and not included as a control variable. Using this specification, test whether incumbent spending is endogenous. Use whatever tests you can for instrument validity. If you cannot include a test, explain why. What do you conclude about this theory?
   i. What other instruments could you suggest for incumbent spending? Would these overcome both types of endogeneity discussed in part c.? Why might challenger spending be endogenous? What instruments could you suggest for challenger spending?
   j. Justify or criticize the races that are excluded from the data set: open seats (where there is no incumbent), unopposed incumbents (in either 1984 or 1986), first-year incumbents, and the outliers (who were financing national campaigns).
3. {40} [from Mullahy and Sindelar 1996 via Stock 2007] Case study: Alcohol and jobs.
   Conventional wisdom holds that heavy drinking can result in difficulties in the workplace and problems holding a job. Job loss associated with heavy drinking could exacerbate other problems within the drinker’s family and increase reliance on publicly provided social services. But, as a quantitative matter, what is the effect of heavy drinking on the ability of a worker to find and to hold a job? Looked at from a different perspective, suppose there were an effective intervention that could reduce heavy drinking (a drug treatment, support groups, counseling, etc.); would this reduction in drinking improve the job market prospects for the individual? In this problem, you will analyze data on job market status and alcohol consumption. The data are 9822 observations on individual-level data from the 1988 Alcohol Supplement of the National Health Interview Survey. The data (alcojobs.dta) is available here. The variables are described within the data set. (Use the command describe after opening the file.)
   As an analyst for the Congressional Research Service, you have been tasked by a member of Congress to address this issue to determine whether an intervention would be worthwhile. Write a professional report that provides an estimate of the effect of heavy drinking on holding a job. [Remember that the main text of such a report is one page, followed by a (statistical) appendix that often includes professional tables (i.e. not Stata output) and other technical comments, followed by something that shows your work for the grader, e.g. do-files, formattted Stata output.] Is this effect substantively large? How confident are you in the validity of your estimated effect, and why? What are the limitations of your analysis? If you could be given more time and money, what other variables would you include in your analysis? The report should both provide your best estimate (or estimates) of this effect, based on the data provided, and provide a careful assessment of the validity of your estimate(s). Provide the reader with a sense of the statistical uncertainty surrounding your estimates.
   Notes on the data set: The data are from the 1988 Alcohol Supplement to the National Health Interview Survey of the adult (age 18 and older) U.S. population. The data are for men only. The data set contains information on family characteristics, employment status, socioeconomic indicators. These data were linked to state-level data on alcohol and cigarette taxes and to other pertinent labor market and health data. The survey asks, among other things, about alcohol consumption in the two weeks prior to the survey. The measure of heavy drinking used here is that the respondent’s alcohol consumption falls at or above the 90th percentile of alcohol consumption, which is 18.0 ounces of ethanol per two-week period (approximately 36 drinks in two weeks). The survey asks about other health indicators and includes an overall index of health status (excellent, very good, good, fair, poor), constructed using the self-reported health status indicators. Three categories of labor market status are reported: employed, which means the respondent has a paying job; unemployed, which means the respondent does not have a paying job but has engaged in some job search activity over the past two weeks; and out of the labor market, which means the respondent which means the respondent does not have a paying job and is not looking for one.
4. {25} Research Project Proposal:
   Turn in a 1-2 page, double-spaced proposal (standard font and margins) outlining the research question you plan to address, explains a potential causal connection linking an independent and dependent variable, suggests which econometric model(s) will be used, offers at least 4 relevant citations, and discusses possible data sources to be used. If you do not have a ready idea that will use the tools taught this semester, you may wish to look at a panel data set, or some other hierarchical data set.

Back to Assignments page