2nd Year Workshop: Tips for generating research ideas

Ian McCarthy, Emory University and NBER

Economics PhD Professionalism Workshop

Where to get ideas?

If you’re feeling lost right now, that’s completely normal. You just spent a year or two learning tools. Now you’re supposed to use them, and nobody gave you a manual for that part.

The good news: ideas are everywhere. The hard part is learning to see them.

  1. Read non-academic policy papers/briefs
    • Examples from health econ: Commonwealth Fund, Petersen-KFF, Health Affairs or NEJM opinion pieces
    • Every field has equivalents — find the policy outlets and trade publications in your area
    • Newspapers and podcasts count too
  1. Academic papers and survey articles
    • Read the abstract and skim the rest (you’ll read more carefully later)
    • JEP and JEL are good for survey articles
    • Get TOC notifications from relevant journals
    • NBER working papers
    • Use AI tools to help digest papers (Elicit, Semantic Scholar, Scholarcy)
  1. Seminars
    • Attend and read papers in advance
    • Meet with speakers and ask questions
    • Pay attention to what the audience asks — that’s where the gaps are
  1. Talk to faculty
    • You don’t need a polished idea to schedule a meeting
    • Bring 2-3 half-baked topics and say “I’ve been thinking about X, does this seem like there’s something here?”
    • The goal is to leave with a sharper version of what you came in with, not to impress anyone
  1. Talk to classmates
    • Your cohort sees the world differently than you do — that’s an asset
    • Throw out ideas over coffee. Most will be bad. That’s the point.
  1. Diversify
    • Read outside of your narrow subfield
    • You are an economist first
    • Be curious and engaged

From topic to idea

A topic is not an idea. An idea has a question with an answer you can find in data.

Topic Research question
“I’m interested in hospitals” “Do hospitals change their service mix when a nearby competitor closes?”
“Education policy seems important” “Does reducing class size in early grades affect long-run earnings?”
“I want to study trade” “How do firms adjust employment when tariffs are unexpectedly reversed?”

Most of your early ideas will be topics, not questions. That’s fine. The work is in narrowing.

Kill bad ideas quickly. If you can’t find data, or the answer is obvious, or nobody would care about the answer — move on. Don’t spend a semester on something you could have ruled out in a week.

Keep track of your ideas

Find a system and stick with it. It barely matters which one — what matters is that you write things down when you think of them, not later.

  • Notion, Google Keep, a paper journal, a text file on your desktop
  • Zotero or Research Rabbit for tracking papers alongside your notes

Three general categories:

  1. Brainstorming (no idea is too crazy)
    • Talk about these with peers
    • Look into the literature and potential data
    • Craft an actual question
  2. Research proposals (more developed)
    • Slightly more formalized and developed
    • Talk about these with classmates and faculty
  3. Research plan (ready to go)
    • Vetted by peers and faculty
    • Ready to start the actual research process