ECO 304 Final Exam
Spring 2025
Due 6/11/25 at 11:59pm
Instructions:
Choose one of the following five topics and write an 8-12 page paper (including tables)
addressing the topic and explaining how you use the provided data to answer the question.
These papers should include (at least) the following five sections: (a) Introduction, (b) Data, (c)
Empirical Specification(s), (d) Results, and (e) Conclusion.
The introduction should state your research question and why it is important. It should
then very briefly discuss your empirical approach and results (1-1.5 pages).
The data section should describe your data. This includes your data source and a little bit
of information about this source (use the Internet). It should also include summary
statistics of your data to provide the reader with some context to understand the data, and
summary statistics that help get at your research question that you will address in your
regression analysis. These summary statistics can be presented in tables and/or figures,
but should also be discussed in the text. (2-3 pages).
The empirical specification section should describe the regressions you plan to estimate
including (i) some discussion about what variables you are including in your specification
and why; (ii) which coefficient constitutes your “key” coefficient, and (iii) potential
sources of omitted variable bias that could bias your estimates. Be sure to write out your
regression equations, where equations that include that actual outcome variable and
actual explanatory variable on the key independent variable are better than generic Y and
X. That said, you can use a generic X for a group of other variables (e.g. demographic
variables), but what exactly is included in X should be described in the text.
(2 pages)
The results section should present your empirical results both in words and in
tables/figures. Please limit your discussion to the coefficients on the independent
variables of interest. The same is true for the tables. (2-4 pages)
The conclusion should briefly state your research question and findings. It should then
discuss some combinations of the policy implications of these results, the limitations of
your study, and some ideas for future studies. (1 page)
All submissions MUST include a do file showing how they analyzed the data. Half of the
points are associated with the analysis. Half are associated with your write-up.
The five potential topics include:
- Topic 1: Does Gentrification Increase Apartment Rents for Other Residents?
- Topic 2: How does the Gender Wage Gap change over the life cycle for College-
educated women?
- Topic 3: Birth Order and Parental Investments in Children
- Topic 4: How much of the labor market returns to college completion reflect differences
in underlying ability? Does this effect differ by the level of schooling?
- Topic 5: Forecasting Medicare expenditures and government deficits from Medicare
Each of these questions if described in more detail below. However, I have also added a separate
folder on D2L for each topic, which includes a more detailed discussion about the topic,
including discussion of the associated data.
Good luck!
Topic 1: Gentrification and Apartment Rents
How does neighborhood gentrification affect apartment rents paid by other residents? How does
it impact other housing characteristics, i.e. does gentrification push existing households into
smaller units? Estimate elasticity coefficients and interpret them accordingly. Does this change
in rent/housing characteristics seem to be due to improved economic circumstances of all
individuals in the neighborhood or simply increased demand for local housing from the
gentrification? Are there other benefits that might justify these higher rents? Answer these
questions using American Community Survey data for Chicago over the period 2012-2019.
Note: Gentrification in this set-up is an increase in the number of young college-educated
professionals living in a specific neighborhood. So, it is essential to include neighborhood
dummy variables in your regression models.
Topic 2: How does the Gender Wage Gap change over the life cycle for college-educated
women? How important are occupational choice and labor force participation (hours and
weeks worked) in explaining these penalties?
Use the American Community Survey over the period 2005-2019 to estimate the gender wage
gap over a college-educated woman’s lifecycle, i.e. the gender wage gap as someone ages. What
does this say about when the penalty starts and grows? How much do occupational choice and
labor force participation decisions (i.e. hours and weeks worked) explain this gap? Given what
you find out, can you offer any potential explanations for why the remaining gender wage gap
exists? How much of the gap appears to reflect discrimination versus differences in labor market
choices for men and women?
Topic 3: Birth Order and Parental Investments in Children
There is growing evidence that birth order matters for cognitive development in children where
first-born children tend to outperform their younger siblings in many schooling outcomes. Much
of this effect seems to assigned to the fact that first-born children get a lot more time with their
parents. That said, second and third children may benefit from parents being more financially
stable, which tends to occur as people/parents age. Evaluate whether first-born children indeed
get more time and attention from their parents using data on time investments with children from
the Survey of Income Program Participation (SIPP). Then, examine whether subsequent children
receive greater financial investments such as signing them up for music lessons or sending them
to private school.
Hints on Empirical Approach: Examine the impact of being a first born child on time and
financial investments. Be sure to also condition on whether the first-born is the only born as the
effect might disappear after a second or third child is born. Be sure to condition on other
important factors such as parents age and others. Also, be sure to discuss other potential sources
of omitted variable bias. Note, the interpreting the coefficients becomes a little tricky when
multiple measures of birth order are included in the regression.
Topic 4: How much of the labor market returns to education reflect differences in
underlying ability? Are these effects different for different levels of schooling.
There is extensive empirical evidence that education increases labor market earnings. However,
a common concern across many studies is that it is difficult to identify the effect of education
separate from other factors associated with both college attendance and earnings, such as
underlying intellectual ability. Use data from the 2006 wave of NLSY79, which includes data on
an individual’s AFQT test scores (a standardized test score similar to the SAT) to examine how
much of the returns to education simply reflect differences in underlying cognitive ability. Are
the effects different for different levels of schooling? Using dummy variables, which allow you
to estimate non-linear effects to education, what are the returns to a high school diploma, an
associates degree, and a college degree? How do these different returns change once you
account for underlying ability?
Topic 5: Forecasting Medicare expenditures and government deficits from Medicare
Healthcare expenditures are a large and growing portion of the U.S. federal government’s
budget. Document this growth in healthcare costs and using data on Medicare expenditures,
health insurance provided by the U.S. Federal Government for all people over age 65. How
much of the growth is due to increases in the number of recipients versus increases in the cost of
providing care. Then, forecast Medicare costs and the resulting deficit from Medicare to 2030.
Be sure to justify your forecasting model. Is this growth in healthcare costs sustainable? If not,
propose a policy solution to address the problem. Note that the dollar values are in billions and
the population variables are in millions.
