Chapter 2
A Framework for Analysis
◼ Solutions to Problems
1. Proximate causes are causes that directly affect the variable of interest. Low levels of physical and
human capital, technology, and efficiency are all examples of a proximate cause of low GDP per
capita.
2. Fundamental causes are causes that indirectly affect the variable of interest by systematically
3. To show different levels of factors of production, the figures must not intersect at the same level of
output. To show different levels of productivity, the figures must have different slopes. In the figure
4. In the long run, the two countries would be expected to have the same levels (and thus growth rates)
Chapter 2 A Framework for Analysis 5
5. A correct solution is a variable that systematically rises or falls only because of a rise or fall in
GDP per capita. Take for instance expenditures on unproductive luxury items. It is reasonable to
conclude that as GDP per capita rises, individuals allocate a larger amount of money to purchasing
6. Simply finding a correlation between being overweight and having a heart attack does not imply
causation. The correlation could be due to a missing variable like genetics which may be a factor in
7. a. Although the majority of right-wing voters may live longer, the inference that being a political
8. a. Positive Correlation. It is reasonable to assume that higher (lower) GDP per capita increases
b. Negative Correlation. The higher GDP is per capita, the more likely it is that basic nutrition needs
of the population will be met, and the smaller the number of people suffering from malnutrition,
c. No Correlation or Positive Correlation. There are two things to consider. First, does eyesight
progressively deteriorate with age? Second, does the level of GDP positively affect both one’s
ability to diagnose and correct vision problems and one’s life expectancy through access to better
nutrition, health care, and so on? If one does not assume the above to be true, then there should
6 Weil • Economic Growth, Second Edition
9. To construct a randomized controlled trial, one would advertise the study on the campus, and once a
certain number of subjects (enough to give the RCT statistical power) has signed up, subjects would
be randomized into two groups: treatment and control. The treatment group would receive the drug,
while the control group would not. To assess the outcomes, first randomization has to be checked, by