How I Actually Prepared for the Economics Comprehensive
The comprehensive exam in economics is not a test of whether you know every theorem. It is a test of whether you can move between frameworks without breaking your reasoning. When you sit down with three hours and a prompt like "discuss the welfare implications of quantitative easing in an open economy with rigid wages," the problem is not missing formulas. The problem is knowing which model applies where and why, and being able to hold microfoundations, aggregate demand, and international finance in your head at the same time. I spent about nine months building my prep around a system I called Ideas For Economics Comprehensive. That name is not official. It was just how I labeled my shared folder after a grad student in my cohort asked what the hell I was doing differently. The system is basically a structured way to force yourself to produce complete, model-grounded answers across all three fields before the exam window arrives.
My Ideas For Economics Comprehensive Setup
Here is what the actual setup looked like. I built a matrix. Rows were the major subfields: macro, micro theory, econometrics, plus international and money and banking as smaller categories. Columns were the standard question types that showed up every year: model derivation, comparative statics, critique, policy evaluation, and empirical identification. I filled every cell with a practice question I had either found from old comps or constructed by combining two topics that never got paired in class. The key move was timing. I did not just read solutions. I wrote full timed answers on actual paper, no notes, no laptop, the same conditions as the real exam. A standard answer took me forty five to fifty minutes under timed conditions. I aimed for about six to eight fully written answers per week once the writing phase started. I kept every single one. You need them for review, and you need to see your own pattern of repeated errors. One thing nobody warns you about is how fast your handwriting degrades when you are under pressure and writing dense equations. I learned this the hard way in my third month. My Krugman-type IS-LM diagrams came out looking like they had been dragged through a woodchipper, and my LaTeX-style algebra on paper turned into something the grader could not parse. I started doing a separate ten minute handwriting drill each day where I copied one page of a graduate macro text at normal speed. It took two weeks and cut my diagram legibility problems down to almost nothing. The graders do notice sloppy graphs. They just do not tell you.
For micro, the standard trap is answer inflation. You write four pages when two pages with a clean figure and three clearly stated propositions would score higher. I learned this after my first practice session where I lost points on a mechanism design question because I derived the full Myerson-Satterthwaite conditions but never actually answered what the exercise asked about incentive compatibility in the specific two-agent case. I started using a rule where the first three sentences of every answer had to state the result before any derivation. That forced me to identify the core claim early and cut the rest down. Econometrics was the section where my approach actually broke down in a useful way. Early on I was treating it like applied math and spending too much time on derivations that never showed up on the exam. The real skill is stating the identification strategy, listing the assumptions, and then saying what would go wrong if those assumptions failed. I restructured my practice questions to always include a second part where I had to defend or refute an empirical claim under a changed assumption set. That mirrors what the exam actually does. There is a practical bottleneck with this method that deserves mention. It requires access to old exams, and those are not always easy to get. Some departments do not publish them. When I hit that wall for my money and banking section, I substituted by taking recent Federal Reserve working papers and turning their abstracts into comp questions. I would ask myself what the model assumptions were, what the policy implication was, and how I would test it empirically. It was not identical, but it covered the same ground and forced me to work at the right level of abstraction.
Get the Full Details
Another limitation of the system is that it assumes you already have a base level of familiarity with the material. If you are starting from zero in three fields at once, this approach will overwhelm you. In that case, you need a different structure first, something more like sequential topic mastery before you attempt mixed integrative writing. I saw two people in my cohort try this method too early and waste six weeks reinventing the basics instead of building fluency. What this method does well is expose the gap between passive understanding and active recall under time pressure. You can read Blanchard and Tirole and feel like you understand everything until you have twenty minutes to write a coherent answer that links labor market search theory to aggregate fluctuations without notes. The matrix forces that translation to happen repeatedly over months, not days. If you want to start this, the first step is simply getting the past exams from your department website or asking your program coordinator. Then map out the question types you see. Build the grid. Pick one cell per week to practice under timed conditions. Review your own work against whatever rubric your department publishes, and if there is none, grade yourself against the standards of a clear proposition, a correct model setup, a valid derivation or argument, and a short interpretation that addresses the prompt directly.
It is boring work. It is also the part that actually moves the score. The system is not elegant, but it is honest about what the exam rewards, which is organized thinking written quickly and clearly across multiple economic frameworks.
