Second Short Paper – More Detailed Description

The idea here is to come up with an original hypothesis related to political economy, do a very
simple test of this hypothesis, and then describe ways that this test may be misleading. It is
essentially asking you to practice your critical thinking skills on a proposition that you develop
yourself. It should only be about 3 pages long.

To make it clearer, let me describe the steps:
 1. Create a hypothesis from the field of political economy. A hypothesis is a cause and effect
relationship. Therefore you need both cause and effect. You should have one paragraph that
describes your hypothesis and why you think it might be true. This means discussing a possible
mechanism that leads from cause to effect – how does it work.

 a. Example: My idea for a simple hypothesis would be some relation between types of welfare states
and public opinion. We discussed 3 types of welfare states (liberal, social democratic, and
conservative). You could hypothesize either that a certain type of attitude leads to a certain type
of welfare state or alternatively that a certain type of welfare state leads to certain attitudes.
A hypothesis could be that social democratic welfare states have high levels of interpersonal
trust, liberal welfare states low levels, and conservative welfare state in between levels.
 b. You could try something more complicated if you wish – eg, the relationship between confidence
in the state and levels of development – but that is not necessary. This would require getting data
on two variables (public opinion data and economic data) whereas the previous example would require
just one variable (public opinion data).

  Obtain some simple data to test this relationship. You will test it by drawing a simple graph.
You will use the interocular trauma test for assessing whether the data fit your hypothesis – that
is, does it hit you between the eyes. You don’t need anything more than that. You should write one
paragraph saying what you did and showing the results in a simple graph.
 a. I recommended using the World Values Survey and the online analysis section. You can scroll
through lots of different questions and pick whatever countries are relevant. Then the tool spits
out the country averages. Now you just have to graph these values. If you were comparing types of
welfare states, you would get data for the liberal type (US, UK, Can, Australia, NZ, Ire), the
social democratic type (Swe, Den, Fin, Nor), and the conservative type (Ger, Austria, Bel, Neth,
France, Switz, Italy).

  Now the key part of the paper is critiquing what you just found. If you found a strong
relationship, do you think it is a genuine one – ie, that one variable causes the other? If you
didn’t find a strong relationship, can you think of reasons why? Some of the issues that you should
consider are the following:
 a. Missing variables. I only asked you to consider the relationship between two variables. Could
certain missing variables explain why you didn’t find a relationship or why the relationship you
found isn’t genuine? Explain what these missing variables might be and how they would affect the
relationship.
 b. Endogeneity or reverse causality. Could the direction of causality be the other way? How would
this work? Discuss the alternative mechanism. Which direction do you think is more plausible: from
X to Y or from Y to X?
 c. Measurement. Do you worry about the way that these variables are measured? Could that explain
the relationship? Is there a better to measure them?
 d. Case selection. Are there any problems with the selection of cases here that might bias your
results? How would you change the selection to get better results?

  Your conclusion should tell me the degree of confidence you have in your hypothesis. Are you more
or less confident in it than when you started? Would you embrace an alternative hypothesis? Do you
have any clever ideas for testing your hypothesis in a better way?