Chris Blattman

Close this search box.

PhD applicants: Writing your statement of purpose


I’ve read a lot of personal statements for PhD applications. I sat on admissions at UChicago, Columbia, and Yale, mostly in economics, political science, and public policy. Here’s the advice I’ve given my own students and research assistants to craft their statements. I give it because, sadly, I don’t find most statements helpful. This means they are not helping you, the applicant.

As with all my advice posts, it’s important that students outside elite colleges get this information, so here are some personal thoughts.

[Note: You can now subscribe by email to receive posts to your inbox.]

First, let’s clarify your number one job as an applicant: Send the best, clearest signal of your abilities as a future researcher, and minimize the noise around that signal. I explain why in a longer post on whether and how you should apply to PhD programs (including the other elements of an application packet):

the fundamental problems in graduate admissions are “information overload” and “noise”. For every slot in a PhD program, there are probably 30 to 50 applicants. A department that plans to have a class of 20 students may receive 1000 applications.

Meanwhile, most departments delegate admissions to a small committee of two to six faculty. They don’t have time to read 1000 applications in detail. And the committee may change every year. Thus, their experience may be limited. And you never know who will be on the committee or what they care about. This adds further randomness.

These faculty want to admit the most talented and creative young researchers who will push the field ahead. And they also want you to pass all the most technical classes, because they hate kicking students out. So the admissions committee are looking for strong signals of intelligence, creativity, determination, and other proclivities for research.

But this is hard. There are too many applications. Applicants don’t have many good ways to signal quality. All applicants are trying to send the same signals. And there is a ton of uncertainty around each signal. Hence: Information overload and noise.

Yet most schools as for a written statement of some kind. Sometimes they ask for both a biographical statement and a research statement. What do they want and what should you write?

  1. Don’t tell your life story. This statement is not an undergrad entry essay where you describe your life’s trials and tribulations, or your journey to wanting to do a PhD. It’s not that we don’t care. It’s just that it’s probably not relevant to judging your ability as a researcher. If it is, then weave that into the narrative around your research interests and plans. We have hundreds of these things to read and so you only want to focus on the most important information.
  2. Don’t be cliché. Do not start your with your epiphany—the day the scales fell from your eyes and you realized you wanted to be a professor, or were inspired tackle big questions and social issues. Especially if it involves a child in a poor country. This approach is overused and unoriginal, and the information does not help us judge whether you will be a great researcher (see point 1).
  3. Most material is unnecessary and unhelpful; delete it. Be information dense. Every sentence should communicate important ideas and information about your abilities as a researcher. You see, there are so many applications that readers are looking for an excuse to stop reading or skip a paragraph. Busy people will look at your statement for for 20 seconds. If its information dense they will look at it for for 45 or maybe 60 seconds. Every time you give banal information, it is another reason to stop reading. Some examples of things you should avoid:
    • Platitudes about wanting to be a professor or researcher
    • Generic or flattering statements about being excited to join a program, your admiration for the faculty, etc.
    • Unspecific interests in a research subject or field
    • Routine information such as “I am graduating in May…”
    • Filler sentences like “Please find enclosed…”
  4. The reader should immediately understand what kind of scholar you want to be. I recommend that he first 1-2 paragraphs of your statement do the following:
    • Start with your broad fields of interest (e.g. “I am principally interested in labor and development economics” or “I want to work at the intersection of comparative politics and international relations”)
    • Then give 2-3 examples of broad topics and questions that interest you. (“I’m interested in studying inefficiencies in labor markets, especially market power and monopsony. I’m also interested in…”)
    • You can also describe who you would like to work with in the department and why this is a good fit. Sometimes I suggest putting this at the end, after the specific research proposal. Wherever you put this, make sure that most of the faculty you mention:
      • Are tenure or tenure-track faculty
      • Have their primary appointment in the department you are applying to
      • Are actually there and take students (i.e. they didn’t retire last year, etc.)
  5. Then develop 1-2 of these ideas as specifically as possible. This is the core of your statement. The idea is not to say “this is what I will do for my dissertation”. No applicant knows that. The goal is to show that you know how to ask an answer an interesting and innovative research question. This is hard to do (because you don’t yet have a PhD) but doing it well is a good signal of your creativity, knowledge of the field, and potential as a researcher.
    • You could discuss two ideas in moderate depth, or one idea in greater depth. Either way, I recommend this research discussion be 40-60% of your entire statement.
    • Ideally this is a question or topic of current interest in the field. One thing I often see is that students are focussed on the research frontier 10 years ago (because those are the papers they read in their classes) and are not clued in to some of the current puzzled and priorities. This is hard to avoid, but some reading and your advisors should be able to help you avoid this.
    • The best discussions will (if empirical) identify interesting data and discuss plausible empirical strategies. This is difficult, which is why it is a good signal if you do it well.
    • It’s important to locate your question in the literature without overdoing that discussion. Try to motivate the question with reference to recent and recognizable research papers and agendas. If you are mainly citing articles with few citations, in lower-ranked journals, this is a sign that you need to link your idea to bigger debates in the field, or perhaps rethink the question you are proposing.
    • This is (in my experience) the most crucial section for most social science departments. Except possibly economics. It’s not clear how seriously many departments take your statement in economics, and some of my colleagues profess to never look at the statement. That may be true, but some will look, and you have to have a statement, so I suggest following this advice to make it a research proposal.
  6. Only if necessary, give information that might help us understand any apparent weaknesses or puzzles in your application. Some examples:
    • Why you studied physics but now are doing political science
    • What happened in that single bad semester on your transcript
    • How to interpret your foreign GPA, and where you ranked in your class
    • Clarify your classes if they have off names (e.g. “My class called XX was a Real Analysis class using textbook X, and so I have all the mathematical requirements for entry.”
  7. Get help. Your letter writers, professors you work for, or PhD student you know can read and give feedback on your statements. Ask them for their advice. Do this early–a couple months before the application, ideally. they can help you frame your question in a more interesting way, decide what papers to mention, or what is or is not frontier.
  8. Don’t be repetitive. This is not the place to restate your CV (“First I worked for Professor… and then I worked for…”). They have your CV. Use this document to do something no other in your application can do. Only mention work or other experience if you can add essential, high-density information the reader cannot get elsewhere in the application packet. Maybe you picked up specific technical skills working on a project that relate to the research proposal you just described? If not, you don’t have to say anything at all about your past. Just let the research proposal speak for itself.
  9. Delete useless words and sentences! After you have deleted all the plartitudes and routine sentences (see point 3) keep deleting! Every extraneous word or sentence lowers the average quality of the document. Look for the least useful paragraphs. Delete them, or at least cut most of that material. Try to make a 6-line paragraph 4 lines. Try to make a 15-word sentence 10 words.
    1. I recommend using the Hemingway Editor as a tool to write more clearly. Some long and complex sentences are ok, but sparingly. And they can often be improved. Aim for a grade 10 reading level.
  10. Make it easy to skim and read quickly. In particular:
    • Use active voice
    • Omit needless material and words (see points 3 and 9)
    • Limit jargon
    • Each paragraph should be a distinct idea
    • Paragraphs should have a hierarchical structure, with the big idea or general point as the first topic sentence, and the rest of the paragraph elaborates. Someone should be able to get an “executive summary” but simply reading the first line in every paragraph. they should make sense as a story/summary.
    • Use subheadings if possible, to delineate sections such as your broad fields of interest (point 4), your research proposal (point 5), and other key information (point 6)

This is just my view. Other professors will have different preferences and advice here. So ask them. Get more opinions. Or put your advice in the comments below.