Deck 3: Empirical Tools of Public Finance

Question 1

If two measures,A and B,are correlated,which is definitely TRUE?

A) A causes B.
B) B causes A.
C) There is no causation between A and B.
D) It is possible that there is a causal relationship between A and B.

Answer 1

It is possible that there is a causal relationship between A and B.

Question 2

Suppose that a study were to estimate the effect of public smoking bans by looking at the rates of smoking across states over time when some of those states adopt smoking bans.The study takes which approach?

A) time series analysis
B) cross-sectional regression analysis
C) a randomized trial
D) a quasi-experiment

Answer 2

a quasi-experiment

Question 3

Suppose that to look at the effect of the TANF program on labor supply,researchers were to look at how labor supply by poor households changed over the past 20 years and then compare that change with changes in the income guarantee over time.This would be an example of:

A) time series analysis.
B) cross-sectional regression analysis.
C) a randomized trial.
D) a quasi-experiment.

Answer 3

time series analysis.

Question 4

Two economic variables are _____ if the movement of one causes movement of the other.

A) correlated
B) uncorrelated
C) causally related
D) casually related

Question 5

Suppose you estimate the effect of a treatment using a cross-sectional regression analysis as 10 while the true effect is 15.The difference of 5 represents which of the following?

A) measurement error
B) bias
C) small sample error
D) difference-in-difference estimation

Question 6

Which of the following would be the most useful source of time series variation with which to identify the effect of income guarantees on labor supply?

A) an abrupt increase in the income guarantee over a short time
B) a steady increase in the income guarantee over a long time
C) no change in the income guarantee over time
D) an expansion of the Earned Income Tax Credit

Question 7

You observe in the data that two variables move together.However,since you'd like to know whether a movement in one variable causes movement in the other,you must solve a(n):

A) social efficiency maximization problem.
B) social welfare maximization problem.
C) identification problem.
D) model misspecification problem.

Question 8

Which of the following is a TRUE statement?

A) Control variables used in cross-sectional regression analysis typically remove all bias.
B) Attrition of volunteers in randomized trials rarely occurs.
C) Randomized trials are typically expensive.
D) Randomized trials do not yield reliable results.

Question 9

Which of the following are potential problems with time series analysis?

A) Time series analysis cannot effectively establish causation.
B) Time series analysis trials can be very expensive.
C) Time series analysis can cause ethical problems.
D) When there is a slow-moving trend in one variable,it is very difficult to infer its causal effects on another variable.

Question 10

Which of the following is a potential problem with randomized trials?

A) The treatment and control groups tend to be very different.
B) Randomized trials can raise ethical concerns.
C) Randomized trials lead to biased estimates.
D) Randomized trials cannot be used to answer specific questions.

Question 11

Suppose that job training is randomly assigned among applicants.The people who receive the training form the _______ group; the people who do not receive training form the _______ group.

A) control; treatment
B) treatment; treatment
C) treatment; control
D) control; control

Question 12

Which of the following can be defined as the use of data and statistical methodologies to measure the impact of government policy on individuals and markets?

A) constrained utility maximization
B) empirical public finance
C) welfare economics
D) a balanced budget requirement

Question 13

In considering the results of empirical evidence,policy makers should use the results of studies that show a ______,not just a _____.

A) correlation; causal relationship
B) causal relationship; correlation
C) casual relationship; correlation
D) correlation; casual relationship

Question 14

Randomized trials:

A) cannot effectively establish causation.
B) can be very expensive.
C) avoid ethical problems.
D) are the least expensive way to test a hypothesis.

Question 15

An empirical researcher identifies the effect of drug use on high school dropout rates by examining how the two variables changed over the 1980s and 1990s.This empirical approach can best be characterized as:

A) time series analysis.
B) cross-sectional regression analysis.
C) structural modeling.
D) quasi-experiment.

Question 16

Any source of difference between treatment and control groups that is correlated with the treatment but is not a result of the treatment is called:

A) measurement error.
B) bias.
C) small-sample error.
D) difference-in-difference estimation.

Question 17

Which of the following is NOT true?

A) Randomized trials use observational data.
B) Time series analysis uses observational data.
C) Cross-sectional regression analysis uses observational data.
D) Randomized trials provide an estimate of the causal impact of a particular treatment.

Question 18

Suppose you wish to determine whether exercising increases academic achievement.What would be the best approach to take?

A) Compare the achievements of athletes with those of nonathletes.
B) Compare the achievements of baseball players with those of track and field athletes.
C) Compare treatment and control groups in a randomized trial.
D) Compare yourself with a friend who exercises less.

Question 19

Suppose that in order to determine how non-labor income affects labor supply,you give everyone born on an even-numbered day in August of 1979 $1,000 per year and look at how they change their labor supply compared to people born on odd-numbered days in August of 1979.This approach is which of the following?

A) time series analysis
B) cross-sectional regression analysis
C) a randomized trial
D) a quasi-experiment

Question 20

Two economic variables that move together are considered:

A) correlated.
B) uncorrelated.
C) causal.
D) casual.

Question 21

Suppose 15 states pass a law that limits trial lawyers' compensation.No such law passed in the other 35 states.Suppose that the value of all malpractice suits goes from an average of $2.4 billion to $2.8 billion in those 15 states and goes from $2.3 billion to $2.6 billion in the 35 states,over the time in which the laws were passed.What is the difference-in-difference estimator of the effect of these laws?

A) $0
B) $0.1 billion
C) $0.3 billion
D) $0.4 billion

Question 22

Suppose an empirical economist estimates underlying features of utility functions.Which type of empirical analysis would this researcher be most likely to be engaged in?

A) time series analysis
B) cross-sectional regression analysis
C) structural modeling
D) a quasi-experiment

Question 23

Which of the following statements is TRUE?

A) Control variables used in cross-sectional regression analysis typically remove all bias.
B) Attrition of volunteers in randomized trials is rare.
C) Randomized trials are typically inexpensive.
D) Cross-sectional means comparing many individuals at one point in time.

Question 24

Suppose an empirical researcher is interested in analyzing the impact of restricting HMO membership on insurance rates among workers.The researcher noticed that one state passed a law that restricted HMO membership and an otherwise similar state did not pass such a law.Which empirical approach would be best suited to this analysis?

A) time series analysis
B) cross-sectional regression analysis
C) a quasi-experiment
D) a random trial

Question 25

A paper published in a medical journal found that children who said they dieted gained more weight over time than did children who said they did not diet.The authors concluded that dieting may lead to binge eating and therefore cause weight gain,a conclusion that was widely reported in the press.Can you think of any alternative explanations?

Question 26

Which of the following statements is TRUE?

A) Cross-sectional regression analysis cannot account for factors that vary between the treatment and control groups.
B) Cross-sectional regression analysis makes only qualitative,not quantitative,judgments about relationships between variables.
C) Since tastes and preferences cannot be easily measured,empirical researchers need not concern themselves with these.
D) Cross-sectional regression analysis assesses the relationship between two variables while holding other factors constant.

Question 27

Suppose a state were to cut the welfare guarantee 30% between 2012 and 2014.Furthermore,suppose that hours of work per year among single mothers rose 20%.Use the time series estimate of the effect to come up with an elasticity of total hours worked with respect to benefits.

A) -0.67
B) 0.67
C) -1
D) 1

Question 28

Suppose a study finds that college students who received financial aid have a greater tendency to graduate than do college students who do not receive financial aid.What is the identification problem that the authors must solve? If the authors conclude on the basis of the data that receiving financial aid increases the ability of college students to graduate,are they correct? What would be the best solution to the identification problem? Explain.

Question 29

An empirical researcher decides to identify the effect of parental divorce on children's high school graduation by comparing children whose parents divorced with children whose parents remained married,controlling for other observable characteristics.This empirical approach can best be characterized as:

A) time series analysis.
B) cross-sectional regression analysis.
C) a quasi-experiment.
D) a random trial.

Question 30

The difference-in-difference estimator is used in:

A) time series analysis.
B) cross-sectional regression analysis.
C) structural modeling.
D) a quasi-experiment.

Question 31

An empirical researcher wants to identify the effect of HMO membership on health care costs and does so by comparing costs with membership across the states.This approach is best characterized as:

A) time series analysis.
B) cross-sectional regression analysis.
C) a quasi-experiment.
D) a random trial.

Question 32

Which empirical approach would be best suited to estimating the effect of the job-training program on earnings?

A) time series analysis
B) cross-sectional regression analysis
C) structural modeling
D) a quasi-experiment

Question 33

Suppose that a state were to cut the welfare guarantee from $6,000 to $3,000 between 2012 and 2014.Furthermore,suppose that hours of work per year among single mothers rose from 900 to 1,200 hours/year.Using the time series approach,what is the effect of the reduction for the average single mother?

A) -300 hours/year
B) 300 hours/year
C) 900 hours/year
D) 1,200 hours/year

Question 34

R²is a measure of:

A) the statistical significance of the regression.
B) how well the regression model fits the underlying data.
C) the statistical significance of a given coefficient.
D) how well the economic theory explains the behavior of the underlying data.

Question 35

Suppose one study uses data on prices and cigarette smoking behavior to estimate the overall effect of cigarette-price changes on smoking,while another uses the same data to decompose the effect into income and substitution effects.The first study provides _________ estimates; the second provides __________ estimates.

A) reduced form; reduced form
B) structural; structural
C) quasi-experiment; randomized
D) reduced form; structural

Question 36

A statistical method used for assessing the relationship between two variables while holding other factors constant is called:

A) time series analysis.
B) regression analysis.
C) a randomized trial.
D) a quasi-experiment.

Question 37

What is the fundamental difference between correlation and causality? Why is it important to distinguish between them?

Question 38

To produce structural estimates,which of the following must be TRUE?

A) The researcher must observe more information than a researcher who estimates reduced-form estimates only.
B) The researcher cannot impose assumptions on the functional form of utility.
C) The researcher must use a randomized study.
D) The variables in the study cannot be correlated with each other.

Question 39

Suppose a state were to cut the welfare guarantee from $6,000 to $3,000 between 2012 and 2014.Furthermore,suppose that hours of work per year among single mothers rose from 900 to 1,200.Which of the following must be TRUE if the time series estimate of the effect of the reduction in the guarantee is to be unbiased?

A) Single mothers in that state in 2012 must have been very different from single mothers in 2014.
B) There must not have been a period of substantial economic growth or decline between 2012 and 2014.
C) The welfare guarantee must have been cut only for a portion of single mothers.
D) Other factors besides welfare benefits determine how many hours a single mother works.

Question 40

Randomized trials and quasi-experiments:

A) cannot distinguish causality from correlation without making unrealistic assumptions.
B) cannot find the effect of a particular treatment.
C) often cannot tell us why outcomes change with a given intervention.
D) are designed only to identify the consequences of a given intervention.

Question 41

(a)Suppose more students wanted to attend a charter school than the charter school could accept,and you wanted to identify the effect of attending charter school by comparing the outcomes of students who attended charter school with those of students who wanted to but could not.Would a quasi-experimental approach be acceptable if the charter school accepted students on the basis of their characteristics?
(b)Would your answer to (a)change if students were accepted on the basis of a lottery instead?

Question 42

Name one potential problem that could lead to biased results if you were to identify the effect of attending a charter school by comparing the scores of students from the charter schools with the scores of students from traditional public schools.

Question 43

Suppose a study found that gasoline prices are lower in markets in which certain gas stations (call them CheapGas stations)are present.What possibilities can you think of that would explain this correlation?