Economic and Policy Research Flashcards

Question 1

What does the term opportunity cost mean in economic reasoning?

Accepted Answer

Opportunity cost is the value of the next best alternative given up when a choice is made. Because resources such as time, money, and land are limited, every decision carries this price. Economists use it to compare options, since the true cost of one program includes the benefits other spending could have produced.

Question 2

What is the difference between positive and normative economics?

Accepted Answer

Positive economics describes how the economy works and makes claims that can be tested against evidence, such as how a tax change affects employment. Normative economics involves value judgments about what outcomes are desirable. Research can inform normative debates but cannot settle which goals society should pursue.

Question 3

What is a ceteris paribus assumption and why do economists use it?

Accepted Answer

A ceteris paribus assumption holds all other relevant factors constant while one variable is changed, so its isolated effect can be studied. Models use it because real economies change many things at once. Its conclusions carry over to real settings only when other factors are stable or controlled for statistically.

Question 4

What is a model in economics and what role do assumptions play in it?

Accepted Answer

An economic model is a simplified representation of a system. Assumptions are deliberate simplifications that make the model tractable and highlight the mechanisms of interest. A model is judged by whether its predictions match evidence and whether its key assumptions are reasonable for the question being asked.

Question 5

What is the difference between correlation and causation?

Accepted Answer

Correlation means two variables move together, while causation means that changing one variable actually brings about a change in the other. Two variables can be correlated because of a third factor, reverse causality, or pure chance. Economists therefore need research designs beyond simple comparisons to support a claim that one factor causes another.

Question 6

What is an endogenous variable and why does it complicate analysis?

Accepted Answer

An endogenous variable is one whose value is determined within the model or system being studied. This creates simultaneity or feedback, so ordinary comparisons can confuse cause with effect. Researchers address it with instrumental variables or structural models that represent both directions of influence.

Question 7

What is a confounding variable and how can it bias a result?

Accepted Answer

A confounding variable is an external factor that affects both the explanatory variable and the outcome, creating a misleading association between them. If it is ignored, the estimated effect may wholly reflect the confounder rather than the factor of interest. Researchers reduce this risk through randomization, matching, stratification, or statistical controls for measured confounders.

Question 8

What is the difference between a stock and a flow in economic measurement?

Accepted Answer

A stock is a quantity measured at one point in time, such as the wealth a household holds on a given day. A flow is a quantity measured over an interval, such as income earned during a month. Over time, flows add to or subtract from stocks, so the two must not be confused.

Question 9

What is a randomized controlled trial and why is it considered a strong design?

Accepted Answer

A randomized controlled trial assigns participants by chance to a treatment group or a control group, then compares outcomes between them. Random assignment tends to balance observed and unobserved characteristics across groups, so differences in outcomes can be attributed to the treatment. Ethical and practical limits sometimes restrict their use.

Question 10

What is a natural experiment and how does it differ from a randomized trial?

Accepted Answer

A natural experiment occurs when an outside event or rule, rather than the researcher, creates treatment and comparison groups. One example is a policy that applies to people just above an eligibility cutoff. Unlike a randomized trial, the assignment was not designed for research, so analysts must argue carefully that the groups are comparable.

Question 11

How does a difference-in-differences design estimate the effect of a policy change?

Accepted Answer

A difference-in-differences design compares changes over time between a group affected by a policy and a comparable unaffected group. Subtracting the control group's change removes shared trends. It depends on the parallel trends assumption, which holds that both groups would have followed the same path without the policy.

Question 12

What is an instrumental variable and what two conditions must it satisfy?

Accepted Answer

An instrumental variable is a factor that affects the outcome only through the explanatory variable of interest. It must be relevant, meaning it strongly predicts the explanatory variable, and it must be valid, meaning it is uncorrelated with the error term. Both conditions are difficult to verify, so instruments need strong justification.

Question 13

What is external validity and how does it differ from internal validity?

Accepted Answer

External validity is the extent to which a study's findings apply to other populations and settings. Internal validity is the extent to which a study correctly identifies cause and effect in its own sample. A study can be internally valid yet fail to generalize, such as when a pilot works only in one city.

Question 14

What is selection bias and how does it affect comparisons between groups?

Accepted Answer

Selection bias occurs when compared groups differ because of how individuals entered them, not because of the treatment. For example, people who choose to enroll in a job training program may already be more motivated than those who do not. Such differences can make a program look more effective than it is.

Question 15

What is regression discontinuity and what is its main assumption?

Accepted Answer

Regression discontinuity compares outcomes for units just above and just below a cutoff that determines treatment. The main assumption is that units near the cutoff are otherwise similar, so no other factor jumps at the threshold. Analysts check this by testing whether people can manipulate their position around the cutoff.

Question 16

What is a placebo test in policy evaluation?

Accepted Answer

A placebo test applies the same analysis to a period or group where no effect should exist, to check for spurious effects. If the placebo shows an effect, the original result may reflect pre-existing differences or a flawed design rather than the policy. Passing does not guarantee validity, but failing is a warning sign.

Question 17

What is the difference between a sample and a population in economic research?

Accepted Answer

A population is the full group a researcher wants to understand, such as all workers in a country. A sample is the subset actually observed, such as a survey of several thousand workers drawn from that country. Sampling designs, including random sampling and weighting, determine how well sample results reflect the population.

Question 18

What is measurement error and how can it affect regression estimates?

Accepted Answer

Measurement error occurs when a recorded value differs from the true concept it is meant to capture, such as underreported income. Random error in an explanatory variable typically biases the estimated coefficient toward zero, while error in the outcome mostly widens standard errors. Systematic error can bias results in either direction.

Question 19

What is an index number and how is it used to summarize economic data?

Accepted Answer

An index number expresses a variable relative to a base value, usually set at 100. Index numbers let analysts compare changes over time or across groups in one figure. A price index, for example, tracks the cost of a fixed basket of goods over time.

Question 20

What is real GDP and how does it differ from nominal GDP?

Accepted Answer

Real GDP measures the total value of goods and services produced in an economy after adjusting for changes in price levels. Nominal GDP values the same output at current prices, so it can rise even when physical output is flat. Comparing real figures across periods isolates growth in actual production from inflation.

Question 21

Why do economists use log transformations on variables such as income?

Accepted Answer

Logarithmic transformations compress large values and make skewed distributions more symmetric, which reduces the influence of extreme observations. In a log-linear model, coefficients can be read approximately as percentage changes in the outcome for a one-unit change in the regressor. Analysts must interpret the results carefully, since the log of zero is undefined.

Question 22

What is a panel dataset and what advantage does it offer?

Accepted Answer

A panel dataset follows the same units, such as households, firms, or countries, across multiple time periods. Its main advantage is that researchers can control for unobserved characteristics that stay constant within each unit, reducing one source of omitted variable bias. Panels show how individual outcomes change over time, which single snapshots cannot.

Question 23

What is missing data and why can it threaten the validity of findings?

Accepted Answer

Missing data occurs when some observations lack values for variables needed in the analysis, such as nonresponse in a household survey. It threatens validity when the missingness is related to the outcome or treatment, because the observed sample then differs systematically from the full population. Researchers address this with weighting, imputation, or sensitivity analysis.

Question 24

What is an administrative dataset and what are its strengths and limits?

Accepted Answer

An administrative dataset consists of records collected by government agencies or organizations during routine operations, such as tax filings, benefit claims, or school enrollment. Its strengths include large coverage and detail that surveys often lack. Its limits include variables gathered for administrative rather than research purposes, and records that may miss people outside the system.

Question 25

What is cost-benefit analysis and what does it aim to determine?

Accepted Answer

Cost-benefit analysis compares the monetized benefits and costs of a policy, including effects that do not pass through markets, to judge whether benefits exceed costs. It also reveals how net value compares with alternatives. Results depend heavily on the values placed on lives, time, and environmental effects, so those assumptions must be transparent.

Question 26

What is a discount rate and why does it matter in policy evaluation?

Accepted Answer

A discount rate is the rate used to convert future costs and benefits into present values, reflecting time preference. Small changes in the rate can dramatically change which long-term policies appear worthwhile. Analysts often report results under several discount rates, and the choice itself is a value judgment that should be explained.

Question 27

What is a counterfactual and why is it central to policy evaluation?

Accepted Answer

A counterfactual is what would have happened to the same people or places in the absence of the policy. It is central because the effect of a policy is defined as the difference between observed outcomes and this unobservable comparison. Research designs are essentially strategies for constructing a credible stand-in for the counterfactual.

Question 28

What is a cost-effectiveness analysis and how does it differ from cost-benefit analysis?

Accepted Answer

Cost-effectiveness analysis compares the cost of achieving a single, non-monetized outcome across policy options, such as cost per case averted. Unlike cost-benefit analysis, it does not place a dollar value on the outcome itself. It is useful when the goal is fixed and the question is which option achieves it most cheaply.

Question 29

What is a sensitivity analysis and why do policy analysts perform one?

Accepted Answer

A sensitivity analysis tests how results change when key assumptions, parameters, or inputs are varied across plausible ranges. Analysts perform one because forecasts and cost estimates often depend on uncertain values, such as future growth rates or take-up. If conclusions reverse under modest changes, decision makers should treat the recommendation with caution.

Question 30

What is a logic model or theory of change in program evaluation?

Accepted Answer

A logic model or theory of change is a diagram or narrative linking a program's inputs and activities to its outputs, short-term outcomes, and long-term goals. It states the causal pathway the program expects to follow and the assumptions behind each link. Evaluators use it to decide what to measure and where failure might occur.

Question 31

What are general equilibrium effects and why might a partial analysis miss them?

Accepted Answer

General equilibrium effects are feedbacks that occur when a policy changes prices or behavior in other markets, which then alter the original outcome. A partial analysis examines only the directly affected market and holds everything else fixed. Ignoring these spillovers can misstate a policy's full impact, especially for large reforms.

Question 32

What is a microsimulation model and what kinds of policy questions does it address?

Accepted Answer

A microsimulation model applies policy rules to a representative or actual sample of individuals or households to estimate who gains, who loses, and by how much. It is widely used for tax and transfer reforms, where distributional effects matter. Results depend on behavioral assumptions, so analysts often test alternatives.

Question 33

What does replication mean in economic research and why does it matter?

Accepted Answer

Replication means reproducing a published result using the original data and code, while replicability refers to obtaining similar results from new data. Successful replication confirms that the reported analysis was carried out as described and was not the product of an undocumented error. Sharing data and code lets others verify findings.

Question 34

What is publication bias and how does it distort the published literature?

Accepted Answer

Publication bias is the tendency for studies with statistically significant or striking results to be published more often than null or unexpected findings. As a result, the published literature can overstate the size of an effect. Registering studies in advance and reporting all results helps counter this.

Question 35

What does a p-value tell a reader, and what does it not tell them?

Accepted Answer

A p-value is the probability of observing data at least this extreme, assuming the null hypothesis of no effect is true. It does not give the probability that the null hypothesis is true, nor the size or importance of the effect. A small p-value signals data unlikely under the null.

Question 36

What is a confidence interval and how should readers interpret it?

Accepted Answer

A confidence interval is a range of values that, under repeated sampling and correct model assumptions, would contain the true parameter at a stated rate. It conveys both the estimated effect and the precision of that estimate. Readers should examine the full range, since a wide interval allows very different conclusions.

Question 37

Why do economists use robustness checks in empirical research?

Accepted Answer

Robustness checks test whether a main result holds when the analysis is changed in reasonable ways, such as using different variable definitions, samples, or model specifications. If the estimated effect stays similar across these variations, readers can have more confidence that it is not an artifact of one arbitrary choice.

Question 38

What is pre-registration and what problem does it help address?

Accepted Answer

Pre-registration is a public, time-stamped record of a study's hypotheses, outcomes, and analysis plan, filed before data are analyzed. It helps address the problem of researchers choosing analyses after seeing results, which can inflate false positive findings. Deviations from the plan can still be made when justified, but they should be reported openly.

Question 39

Why should readers consider funding sources and conflicts of interest when evaluating policy research?

Accepted Answer

Funding sources and conflicts of interest can shape which questions are studied, how analyses are framed, and which results receive attention. Disclosure does not make research wrong, but it allows readers to weigh possible incentives. Strong work discloses funders, describes any independence safeguards, and invites scrutiny of its methods.

Question 40

How can an economist communicate uncertainty to policymakers without losing their attention?

Accepted Answer

Lead with the main finding in plain language, then state the range of plausible effects and the key assumptions behind it. Use one clear chart or table rather than many dense ones, and explain what would change the conclusion. A clear bottom line paired with honest limits serves policymakers best.

Economic and Policy Research Flashcards

1Foundations of Economic Reasoning

2Research Design and Causal Inference

3Data Sources and Measurement

4Policy Analysis Methods

5Evaluating and Communicating Research

Learning Resources