Guide to Reading Economics
Papers
Jeffrey Parker, Reed College
In
your Reed economics courses, you will often be asked to read papers from professional
journals that were written for an audience of Ph.D. economists. Some of these
will be fairly easy to understand; others will be rather opaque. The outline below
suggests a method of extracting what I think are the essential (for class purposes)
elements from papers.
- Who are
the authors? Do they have an established point of view or position in a controversial
debate that they would want to defend? If so, you should recognize this when interpreting
the results of the paper. This may be difficult for you to know without an extensive
knowledge of the literature and is often not an important issue, but when it is
important it can be very important.
- What
question is the paper addressing? This should be obvious from the title, abstract,
and/or introduction. Identifying the target question as you begin reading should
help you put the component parts of the paper into context.
- What
method does the paper use to answer the question? In other words, how does
the paper reach its conclusions?
- Is it a theoretical
paper that deduces a conclusion from a set of theoretical assumptions?
The internal logic of most theoretical papers can usually be assumed to be validated
by the journal's referees and editors. The central questions about the paper may
then revolve around its applicability: Are the model's assumptions plausible for
the question being addressed? It is important to recognize that no economic
model ever has assumptions that are strictly true. A "good" model captures
some essential core characteristics of the world but ignores less important details.
For you, the essential question is: Is this a good model for this question?
- Is
it an empirical study that attempts to assess whether the evidence
is consistent with a particular hypothesis or to estimate an important economic
parameter? Empirical papers are open to question on myriad grounds because there
are many valid ways to test most hypotheses. This is why there are many empirical
papers addressing any important question. To understand an empirical paper, focus
on its essential parts: the logic of the test, the data used, which are the dependent
and which the independent variables, and the econometric methods used to estimate
the model. As a student, it is most important to understand the logic of the test:
What does it mean for a key estimated parameter to be positive or negative? How
does the analysis distinguish between a state of the world in which the central
hypothesis is true and a state where it is false? Are the estimates of the key
parameters statistically significant? Economically significant?
- Is
it a case study that looks in detail at one or more examples of
the phenomenon under study? Is the case chosen typical of other important applications?
Does the author interpret the case as an illustration of a particular theory?
Are there other theories that might explain the case equally well or better?
- Is
it a "meta-study" that surveys other papers addressing
the question? These studies often have voluminous bibliographies that are more
useful as reference tools than as objects for detailed study. Which of the studies
being discussed are the most central or seminal? On what results do the central
studies agree and where do they disagree? Does the preponderance of evidence support
one conclusion or another?
- What
are the paper's results and conclusions? How strongly are the conclusions
supported by the results of the analysis? Do the authors present any "robustness
checks" or "sensitivity analysis" that suggests how sensitive the
conclusion is to assumptions, data, and methods used?
- How
does the paper fit into the broader literature?
Does this paper contradict others that find different answers to the same question?
If so, what is the source of the difference in outcomes? Which, if either, paper
seems to have established its case more strongly? Can we tentatively form a conclusion
or are we left with considerable uncertainty about the answer to the question?
These
elements can usually be extracted from a careful reading of specific sections
of the paper without understanding all the details of other sections. Introduction,
model specification, data, results, interpretation, and conclusion sections often
contain the essential elements. As you read more economics papers, you will get
better at extracting these elements.
If you are asked to present or comment
on a paper in class, it is critical that you understand the paper's essential
elements. Depending on the paper, there may be other details that are important
as well, but most papers can be summarized concisely in a page (or in a five-minute
presentation) by writing or talking about the answers to the questions listed
above.