The PhD in Health Policy, awarded by the Graduate School of Arts and Sciences, is administered by the Committee on Higher Degrees in Health Policy, representing six Harvard University faculties: Faculty of Arts and Sciences, Harvard Kennedy School, Harvard Medical School, Harvard T.H. Chan School of Public Health, Harvard Business School, and Harvard Law School.
The PhD program has the following requirements:
- Two years (generally) of coursework, including a year-long Core course
- Concentration in one academic discipline (decision sciences, economics , management, methods for policy research, or political analysis) and specialization at the dissertation stage in one policy area (global health, healthcare services, mental health, or public health)
- Three one-term courses, chosen from three concentrations outside a student’s field of concentration. The statistics requirement (noted below) may be used to satisfy one of the three requirements, except for students concentrating in methods for policy research
- Two one-term courses in statistics
- A weekly research seminar starting in the third year
- Written general and concentration examinations following the two years of coursework; the general examination contains an oral component
- A dissertation prospectus and oral examination
- A dissertation based on original research and a dissertation defense
Note: If a grade of B- or better is not obtained in a course offering a letter-grade option, the student will not receive program credit for that course.
PhD students in Health Policy choose a concentration and meet specific curriculum requirements in one of six disciplines:
Decision Sciences (Professor Jane Kim, chair). Decision Sciences is the collection of quantitative techniques that are used for decision-making at the individual and collective level. They include decision analysis, risk analysis, cost-benefit and cost-effectiveness analysis, decision modeling, and behavioral decision theory, as well as parts of operations research, microeconomics, statistical inference, management control, cognitive and social psychology, and computer science. Examples of research topics in health decision sciences include:
- cost-effectiveness analysis of medical technologies and pharmaceuticals
- optimal screening policies for cancer and other chronic diseases
- measurement and evaluation of health outcomes, including quality of life
- policy simulation modeling of diseases such as AIDS, tuberculosis, cancer, and asthma
- optimal resource allocation for biomedical research
Economics (Professors Joseph P. Newhouse and Timothy Layton, co-chairs). The concentration in economics focuses on the economic behavior of individuals; providers; insurers; and international, federal, state, and local governments and actors as their actions affect health and medical care. In addition to examining the literature on health economics, the training emphasizes microeconomic theory, econometrics, and interactions with other disciplines, including clinical medicine.
Methods for Policy Research (Professors Mary Beth Landrum and J. Michael McWilliams, co-chairs). Training in this concentration will enable students to study the effects of a wide range of policies and health services (e.g., health insurance, health care quality improvement, clinical decision-making, drug policy, cost-containment, and socioeconomic factors) on behaviors, access, processes and quality of health care, health outcomes, or costs. Students in this concentration will develop proficiency in experimental and quasi-experimental research design, statistics, relevant social sciences, and other methodological approaches (e.g., epidemiology, program evaluation, qualitative methods, and survey design). Previous students in this track have used innovative methodological and statistical approaches to study, for example:
- the impact of acquiring Medicare coverage on the health of previously uninsured adults
- the effects on health behavior of insurance restrictions on maternity lengths of stay
- the effects of drug coverage on access to essential medications in Medicare
- the effects of regulatory changes in legal drinking ages on health and mortality
- the effects of physician experiences with adverse medical events on under-prescribing of essential medicines
- a controlled natural experiment on the effectiveness of direct-to-consumer drug advertising
- the effects of near-universal Medicare coverage on disparities in cardiovascular disease and diabetes control
- methods to estimate racial/ethnic health care disparities and their effects on health
- international differences in outcomes of medical care for acute myocardial infarction
- the effects of unemployment on mortality
Management (Professors Amy Edmondson and Robert S. Huckman, co-chairs). The management concentration prepares students to do research on the managerial, operational, and strategic issues facing a wide range of organizations in the health care industry including: health care providers; pharmaceutical and biotechnology firms; device and technology companies; and private and public insurers. Students in this track examine how theories and concepts from fields such as technology and operations management, organizational behavior, organizational economics, and competitive strategy can be applied to and further developed for understanding health care organizations. Key research themes include:
- learning and process improvement
- organizational structure and performance in health care delivery
- managing R&D organizations
- managing teams in clinical and research settings
- information technology and the management of health care processes
Political Analysis (Professors Sara Bleich and Ben Sommers, co-chairs). This concentration is intended for students who wish to do research on the relationship between politics and health policy. Students will study theories of individual opinion formation, voting behavior, legislative organization, and interest group formation. In addition, students will examine the role of public opinion, interest groups, the media, and institutions in influencing health policy outcomes. The research methodologies most utilized in this track include survey research methods and quantitative statistical methods appropriate for large-scale databases.
There is no language requirement.
Policy on Incompletes
No grade of Incomplete can be used to satisfy any departmental requirements.
All PhD students in Health Policy must receive human subjects training during their first year in the program (before they embark on research during their first summer in the program)—in connection with the Core course. This training must be updated as required by the University.
Once a student has accepted an offer of admission to the PhD program in Health Policy, they are assigned an advisor. After the first year, students have the option to select their own advisors. During the third year of the program, a student forms a dissertation committee, which replaces the student’s faculty advisor. (However, frequently the student invites the second-year advisor to serve on the dissertation committee.) In addition to convening the dissertation committee for the dissertation proposal orals and final defense, a student is encouraged to meet together with their entire committee several other times during the process as well as individually with members of the committee on a regular basis. In the third year and above, all students are required to register for the research seminar and present once per academic year. Students are encouraged to invite their advisors to attend these presentations.
The program office requests an annual progress report that is shared with the advisor, the concentration chair, the program chair, and the director of graduate studies. The program office monitors Incompletes.
Thus, at all times when enrolled in the program, a student has one or more faculty advisors. Generally speaking, students in this program form strong bonds with their advisors.
Master of Arts (AM)
This program does not award an AM. However, it does award a terminal master’s degree in circumstances that warrant it, by vote of the Committee on Higher Degrees in Health Policy. If students leave the program or are withdrawn after passing the general and concentration examinations and completing all their course requirements with a grade of B- or better, they are eligible for a terminal master’s degree.
With the exception of the decision sciences and management tracks of the program, teaching is not a requirement of this program, but at least one Harvard teaching experience is strongly encouraged. In the decision sciences track, students are expected to have at least five HSPH credits (equivalent to one full semester) of experience as teaching fellows in the core decision sciences courses. In the management track, students are required to complete a teaching engagement of one full academic term that includes at least 8 hours of front-of-class teaching experience and at least 16 hours of teaching preparation time.
General and Concentration Examinations
Generally, students take a full load of classes (four semester-long courses, or the equivalent, per term) for each of their first two years in the program. At the end of the two years, general examinations are administered over the course of a week. The week starts with a one-hour in-class examination followed by a two-day take-home examination in which students must demonstrate that they have a master’s level understanding of the concentrations. It is followed the next week by an oral examination. This examination is the same for all students (i.e., is not concentration specific). Students are strongly encouraged to take the general examination at the end of their second year in the program or at the completion of their coursework. If students would like to take it at the end of their first year in the program, they must petition the program. Only students with prior relevant master’s degrees can petition. Once permission is granted, the student must declare by the end of first term that they will indeed take the generals at the end of their first year. If a student fails either part of the generals, the student is given one opportunity the following year to retake the part or parts not passed. Students also take an examination in their concentration, usually at the end of their second year in the program. Students should complete all concentration course requirements prior to taking a concentration examination. Any exceptions must be approved by the student’s concentration chair.
Dissertation Prospectus and Orals
The purpose of the dissertation prospectus and its oral examination is to provide a formal occasion for the student to receive feedback on, and gain approval of, their dissertation topic(s). This should be done early enough to incorporate significant changes in direction based on faculty input, but it should reflect a fairly advanced stage of study design for at least part of the dissertation. Throughout this process, the student is expected to keep in close contact with their dissertation committee.
Generally, dissertations in the PhD in Health Policy Program consist of three papers or three chapters within a monograph. Thus, the prospectus (also known as a proposal; the words are used interchangeably in what follows) should describe the research to be conducted in each paper (expectations are described in more detail below). Sometimes, however, a student may want to obtain results from two of the papers before finalizing the issue to be investigated in the third paper and describing the research that will occur in the third paper. In recognition of this possibility, the PhD in Health Policy program provides for two options for the dissertation proposal and oral exam. One option is to propose all three papers and defend them at the same time; the other option is to propose two papers, defend them, and then within six months propose a third paper and defend it in a second oral examination. A student choosing the second option will receive a “conditional pass” until the proposal for the third paper has been successfully defended; the term “conditional” will not be interpreted as implying inadequacy in any way.
Within one year of passing the qualifying examination, a student is expected to have formed a dissertation committee. Further, each student is strongly encouraged to submit a written proposal for at least two of the dissertation papers (or monograph chapters) and pass an oral examination on their dissertation proposal before September of their fourth year. The PhD program expects that any student choosing to defend a proposal for only two papers initially will successfully defend a proposal for the third paper or chapter no later than the beginning of the spring semester of their fourth year.
All students must successfully defend their dissertation proposals by the end of the fourth year in the program to maintain satisfactory progress. Also, if a student does not pass the dissertation proposal orals by the end of the third year, they must meet with the proposed committee by June 30 of that year to discuss the student’s progress to date. The committee will then be asked to sign a form indicating that the student is making satisfactory progress toward completing the dissertation proposal. If a student has not set a date for the dissertation proposal orals by January of their fourth year, the program must notify the student and the chair of the student’s dissertation committee. There is no limit to the number of times a student may repeat a dissertation proposal oral examination, but ordinarily failure to obtain acceptance of a dissertation proposal by the end of the fourth year would be considered evidence of unsatisfactory progress.
To monitor and spur progress on the dissertation, a weekly research seminar is required of all students, beginning in the third year. At the seminar, the students present their dissertation work at all stages, starting with the proposal stage.
Students are strongly discouraged from leaving the Boston area before they have passed their dissertation proposal oral defense. Those contemplating leaving must discuss it with the executive committee. In addition, the program encourages students to remain in residence throughout the dissertation stage. A student living outside the Boston area must, like all students in the PhD program, meet with their dissertation committee at least twice each academic year (preferably one meeting per term).
Expectations for a Dissertation Prospectus
- The student should identify three parts for the dissertation. Normally these will be three papers, or three chapters within a monograph. In assessing each of the three parts, the committee should judge that each, if successfully carried out, will lead to a publishable paper.
- The target length of a dissertation prospectus for all three papers combined should be between 25 and 35 pages. This expected length is meant simply as a guide; some topics will deserve more detailed or longer descriptions. As a rule, however, students should consider the prospectus a proposal – a compendium of completed papers is not acceptable. The point of the prospectus is to provide a well-documented and detailed proposal of research to be conducted for each of the dissertation papers (or monograph).
- For each of the proposed papers, the student should present:
- Research Questions. What are the research questions that the student is answering? The student will usually need a few paragraphs of introductory material to set up their questions.
- Background and Significance. Why are answers to these questions important? What policy might depend upon the answers? What is the context for the issues?
- Literature Review and Preliminary Findings. At a minimum, this section should review the main literature in the area(s) in which the student is working and summarize its findings (or at least those of relevance to what they are doing). In particular, the literature review should note how and why prior research in the topic area does not answer the questions the student wishes to address. Perhaps the data used by earlier researchers were not satisfactory or detailed enough; perhaps the theoretical model or framework underlying earlier work missed a key point; or perhaps statistical estimation methods have advanced since earlier research was conducted and now it is possible to answer a question that has motivated research on the topic for some time. The literature review should be perceived as part of “setting up the problem” that the paper is intended to address. Since literature reviews can often exceed five pages, students should feel free to provide the extended literature review as an appendix. Also, if more than one of the proposed dissertation papers focuses on different aspects of the same underlying literature, an appendix with a review of the relevant literature for both (or all three) papers is recommended. The review for any one paper in the prospectus should be kept to a few pages.
- Methods and Research Design. Most of what the student writes should be in this section. This is the section where the student indicates how the proposed paper can successfully overcome or address whatever problems have been identified as causing earlier research to have provided insufficient models or evidence on the issue. The student should be addressing the following types of questions in this section: What theory or conceptual framework is the student using or proposing to develop? What data will the student analyze? What do they propose to estimate? (It often helps to write down the equation the student wishes to estimate.) How will the student estimate what they are proposing? Which statistical, econometric, or modeling methods does the student plan to use? Will the data have enough power to enable the student to be reasonably confident about the answer to the questions they are trying to answer? What are the remaining key uncertainties and what does the student see as the main obstacles to carrying out the research? Students should show preliminary results of estimating models or the start on a theoretical model in this section. Especially when using new data or a new estimation technique, it is incumbent on the student to show that they are likely to be able to successfully complete the research for the proposed paper. If the student has not yet obtained the data, the student should at least establish that the proposed study is feasible (i.e., there is adequate power; important constructs are included in data; etc.)
- In the case of the first paper, a draft manuscript, possibly of a preliminary or partially completed version, is often attached as an appendix. The student should give evidence that they have gotten their “hands dirty” and is immersed in the conduct of this study.
- The suggested page length for the dissertation prospectus—25 to 35 pages total if all three papers are being proposed—could be divided as follows. The descriptions for two papers should be particularly well-developed; each should be described in 10 to 12 or more pages. (Frequently, two papers are variations on an underlying topic or may take advantage of the same data set. In this case, the description of one paper could be 15 pages long and the second could be 10, with references to the description of the data in the first paper’s description.) The description of the third paper may be shorter because the student expects to more fully develop an idea as the results from the other two papers become available. Nonetheless, the description should be at least 5 pages and preferably closer to 10 pages in length. As noted above in “Dissertation Prospectus and Orals,” a student also can choose to defend the proposal for the third paper within six months of defending the first two papers. If a student chooses this second option, the proposal for the third paper should be described in some detail—at least 8 to 10 pages. It is acceptable to present alternative possibilities for the third paper if the student is concerned that one idea may not work (perhaps because data may not become available in time). These page lengths are merely a suggestion and should not be viewed as the “rule.”
Overall, the point of the prospectus is to provide the faculty on the dissertation committee with sufficient detail of the proposed papers (or monograph) for them to protect the student from embarking on a research project that will not lead to a finished, publishable paper. Thus, the suggested page length and the sections to be included in the description for each paper are meant as a guide. Students also should be mindful that completed papers for the prospectus stage are strongly discouraged. Providing evidence that data or statistical methods will enable the successful completion of a paper is not to be interpreted as a suggestion that the paper should be all but finished before the prospectus is defended. The prospectus is also intended as an expression of agreement between the committee and the student that if the student carries out the work as specified, the committee is likely to find the dissertation acceptable. It is, however, not a binding contract since research is unpredictable and problems may surface that prevent its successful completion as initially envisioned.
The Following Steps Must Occur Prior to the Dissertation Proposal Orals:
- The student must appoint a dissertation committee, as described under the section on “Dissertation Committee” that follows.
- The student must ascertain from the members of the dissertation committee that they are prepared to schedule the dissertation proposal oral examination.
- The student must arrange with the dissertation committee a mutually agreeable date, time, and location for the dissertation proposal orals, and make this information known to the program director at least two weeks prior to the scheduled dissertation proposal orals. Two hours should be allowed, including time for discussion, evaluation by the dissertation committee, and feedback to the student after the faculty evaluation.
- The student must submit to the dissertation committee, at least two weeks prior to the scheduled dissertation proposal orals, a written dissertation proposal as described above. The written proposal submitted two to three weeks in advance of the dissertation orals should not be the first time the dissertation committee has seen these ideas presented by the student. The entire dissertation committee should be consulted in advance to ensure approval in principle of the topic(s) and to ensure the suitability of the members of the dissertation committee.
The Dissertation Proposal Oral Examination
The program office will maintain copies of proposals, and these may be borrowed by students planning their proposal orals. It is the student’s responsibility to obtain a dissertation proposal form from the program office and to bring it to the chair of the dissertation committee. The student will be responsible for getting the signed form and a copy of the dissertation proposal to the program office after the proposal orals, for inclusion in the student’s folder.
At the dissertation proposal orals, the student will present the proposal in a 20–30 minute oral presentation, leaving most of the time for discussion. Persons invited to the dissertation proposal orals are: 1) the dissertation committee; 2) other faculty members from Harvard and elsewhere invited by the student to provide additional expertise in evaluating the research proposal; 3) members of the Committee on Higher Degrees in Health Policy; 4) the PhD program director; and 5) other students invited by the presenting student, possibly including one who has agreed to take notes. Discussion will be limited, however, to the presenting student, the dissertation committee, and invited faculty in categories 2 and 3 above.
At the conclusion, the dissertation committee meets in closed session to discuss the disposition of the proposal. No grade or ordinal evaluation is given. In evaluating the student’s performance at the orals, the dissertation committee will take into account the quality of the student’s oral presentation, the quality of the student’s responses to questions from the dissertation committee, and the written material prepared prior to the oral date. Possible results are: pass or conditional pass (when two papers are successfully defended) which will be changed to a pass when the third paper is successfully defended within an additional six months. Note that a conditional pass also may be the result if the committee feels that a three-paper prospectus has one paper that is not sufficiently well-developed; again, the conditional pass will be changed to a pass when the third paper is successfully defended within an additional six months. On the rare occasion when the dissertation committee feels that the prospectus is not ready for a defense, the committee will adjourn the oral examination and simply convene a meeting with the student.
Content of the Dissertation
Dissertations may be applications of analytical tools to health policy issues, or they may be primarily theoretical. The dissertation should be written in a scholarly style, including thorough literature reviews, and it must include detailed descriptions of methods, data, and analyses.
The dissertation can take either of two formats: a three-paper format or a monograph format.
The first format consists of three publishable papers relating to health policy. The papers are typically related, either by their substantive content or by methodology, but this is not a requirement. It is recommended that at least two of the three papers be related, either by content or methodology, but this is left to the discretion of the student’s dissertation committee.
While publishability is a necessary condition for the acceptability of the dissertation, the fact that a paper has been published in a peer-reviewed publication does not necessarily make it acceptable for the dissertation. Material such as literature reviews, or detailed description of analytic methods and data, which may be excluded from published versions due to page constraints, must be included in the dissertation, possibly as appendices or as separate background papers.
On occasion, one or more of the papers may have been published prior to submission of the dissertation. However, none of the papers may have been published prior to the student’s matriculation into the program, and the majority of the work on the dissertation must be completed after matriculation.
The dissertation must include an overview summarizing the papers. It must also contain material that describes, in non-technical terms, the implications of the papers’ findings for the real world, as well as directions for future research that are suggested by the papers’ findings and/or limitations. This material may be incorporated into the individual papers (e.g., in discussion sections), or in a separate concluding section of the dissertation.
The second format option is a traditional monograph-style dissertation. Such a dissertation must either (a) contain at least three independently publishable units (which may be chapters) or (b) be suitable for subsequent publication in book form. The rules and recommendations described above for the three-paper format, regarding prior publication, dissertation summary, and concluding sections, apply to a traditional monograph-style dissertation as well.
The dissertation must be innovative, in the sense that an existing method is being applied in a new way or to a new problem area, or in the sense that a methodology is extended or modified in a significant way. Primarily theoretical dissertations must still include a substantial demonstration of their applicability to a real-world, contemporary health policy issue, and this application should be the major focus of one or more papers or chapters.
Coauthored dissertation papers or chapters are permitted. Order of authorship should follow the conventions of the field to which the paper is being submitted. The student should be first author for journals where first authorship indicates primary responsibility for the paper. Faculty members and students are cautioned that a faculty advisor should be a coauthor only if they contribute substantially to the development of the database or analytical methodology for the paper or chapter.
If, however, the faculty member is primarily responsible for both the data and method, then the paper probably does not qualify as independent work by the student. If the faculty member has developed the methodology in a previous research study, then it is expected that the student will apply the methodology independently to the problem under investigation.
The student is responsible for selecting a dissertation committee consisting of a dissertation advisor and at least two additional faculty members. The dissertation committee must include at least two full-time faculty members at Harvard University and at least one member of the Committee on Higher Degrees (CHD) in Health Policy. The dissertation advisor chairs the dissertation committee and must be a full-time faculty member at Harvard University, and preferably will be a member of the CHD in Health Policy. At least two of the three dissertation committee members must be members of the Faculty of Arts and Sciences and/or members of the CHD in Health Policy. If only one of the three members has an FAS or CHD in Health Policy appointment, then the chair of the CHD in Health Policy must also approve and sign the dissertation. The dissertation committee may include a faculty member from another university, an emeritus professor, or an adjunct professor at Harvard or elsewhere. (A member not on university faculties may be included only as a fourth member with the approval of the CHD in Health Policy.) The membership of the dissertation committee must be approved by the executive committee of the CHD in Health Policy before the dissertation proposal oral examination is scheduled.
After passing the proposal oral examination, students are expected to meet with their dissertation committee at least twice each academic year (preferably one meeting per term). This is the case whether or not a student is in residence. One of the meetings must be with all the members of their dissertation committee physically present. The other meeting may be by teleconference if necessary. For documentation of these meetings, the program office will provide forms, which students will be expected to return to the program office.
It is advisable to arrange a tentative date for the defense well in advance to resolve possible scheduling conflicts between dissertation advisors. However, a dissertation defense may not be scheduled until at least drafts of all three papers have been submitted to all members of a student’s dissertation committee. All defenses will have a public presentation component. It is the option of the dissertation committee to have the entire defense public or to close the examination part, followed by a public presentation open to faculty, students, and other interested parties—all of whom may ask questions. In both cases, a student must allow for time after the defense to work on revisions required by the committee. A draft copy of the dissertation must be submitted to each member of the dissertation committee at least two weeks prior to the defense, and the program office must be notified of the time for the defense at least two weeks prior as well. At this point, anyone on the Committee on Higher Degrees in Health Policy has the right to ask for a draft copy of the dissertation. The program office will provide the GSAS Dissertation Acceptance Certificate for the dissertation committee members to sign at the defense.
The rules of the Graduate School of Arts and Sciences regarding the format in which the dissertation will be submitted will apply. The PhD Program in Health Policy requires submission of one bound copy of each dissertation to the program office. In addition, the student will need to submit an electronic copy of the dissertation abstract to the program office for inclusion on the program’s website.
Most students complete the entire PhD in approximately five years. All work for the PhD, including the dissertation, should be completed within eight years. Students whose work is not completed within this period will generally be asked to withdraw from the program, but will thereafter be allowed to apply for readmission so as to re-register for the purpose of receiving the degree, once the completed dissertation has been judged satisfactory by the dissertation committee. Exceptions to this rule will be granted only under extraordinary circumstances.