Do you prefer macro or micro? Does development economics excite you in a way that economic history does not? Whatever you choose to do your dissertation on, you are going to have to live with it in the back of your mind for at least eight months. If you pick a topic that you don’t find interesting it is likely that this will come across to the examiner.
Can you do it?
A dissertation is a way in which you can show off your economic skills to an examiner. These can be broadly divided into three types (with the corresponding ‘dissertation section’ in brackets):
- Essay-based (Literature Review)
- Theory/mathematics (Model building/ manipulation)
- Empirics (Econometrics/ statistics)
Of course, any one topic can be tackled in different ways. But if the literature is heavily empirical, and you really want to avoid doing any econometrics (for instance), you might want to think carefully about your topic choice.
For example, the literature on the consumption function is heavily empirical so any investigation about consumption is likely to have an empirical slant as well. This is not a hard and fast rule: if you have a new theory or mathematical model to derive then this might take the place of some econometrics. But, in general, if the literature is heavily empirical then it is likely that your dissertation will have an empirical slant as well.
Some of the most exciting and seminal contributions in economics have been when economists have broached an entirely new field. Gary Becker won the Nobel Prize for introducing the economic rational-agent model to analyse household fertility decisions, for example. But an undergraduate dissertation is unlikely to do anything groundbreaking (and if you think you can, you should definitely discuss it with your supervisor first!)
Choosing a topic with a large literature is normally good: it means that you have something to discuss in a literature review (should you decide to have one). It also gives you ideas if you decide to build a model or different specifications to critique if you do some econometrics. A large literature should not be seen as ‘using up’ a topic: the more work that already exists, the more different papers you can use for inspiration.
There is nothing more frustrating than trying to track down data (except for possibly cleaning it) if you want to do some econometrics. See the empirical section for a list of data sources. Normally, macro data is more readily available than micro data. It’s best to ask your supervisor early on, and they can let you know before you commit to anything.
One of the major potential advantages of a dissertation is that the extra knowledge that you will gain can look very impressive if included in an exam answer. What’s more, lecture material can usually provide a useful basis for thinking about a topic before you have started researching the literature yourself. Lecture handouts often contain ‘further reading’ sections at the end; and even if they do not, your lecturer should be able to point you in the direction of the key papers.
Having said this, there is a danger in becoming too specialised: in particular, you probably want to make sure that you have enough variety left so that you find your dissertation interesting enough to put effort into. Happily, in practice, there is sufficient variety within each examined module that this is not usually a problem.
This is less important than it first appears, but probably should be included as a matter of completeness. An undergraduate dissertation can normally be successfully supervised by most members of an economics faculty, regardless of specialism.
Having said that getting a specialist to supervise you is always desirable, but it is normally not crucial. The only exception to this is if you are doing some advanced econometrics or computer modelling and would like some advice: some economists received their training long before the latest econometric techniques were taught (or even discovered) so it might be worth vetting potential supervisors before you approach them.