Agendas, Alternatives, and Public Policies, 1976, 1977, 1978, 1979 [United States] (ICPSR 28024)
Principal Investigator(s): Kingdon, John W., University of Michigan
Summary: This data collection was created to study agenda-setting and alternative specification in the federal government. It concentrates on two federal policy areas, health and transportation, but the theories generated in the research may be quite widely applicable beyond those two areas. The aim of the work was not to study how issues are decided in some authoritative process like a congressional vote, but instead to study how issues get to be issues in the first place, how items rise and fall on t... (more info)
One or more files in this study are not available for download due to special restrictions; consult the restrictions note to learn more. You can apply online for access to the data. Authentication is required to apply for access.
This data collection includes both quantitative and qualitative data. The quantitative data files are available to all users. The qualitative data, including scanned typescripts of interview write-ups and a redacted list of respondents (e.g., "civil servant in HEW", or "congressional staffer in transportation"), are restricted from general dissemination. Users interested in obtaining these data must complete an Agreement for the Use of Confidential Data, specify the reasons for the request, and obtain IRB approval or notice of exemption for their research. Apply for access to these data through the ICPSR Restricted Data Contract Portal, which can be accessed via the study home page.
The full list of interview respondents, with names and affiliations, is available only through the ICPSR Data Enclave; the full list of respondents will be made available under restricted access status 30 years after the data are first released by ICPSR.
Kingdon, John W. Agendas, Alternatives, and Public Policies, 1976, 1977, 1978, 1979 [United States]. ICPSR28024-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2011-03-18. doi:10.3886/ICPSR28024.v1
Persistent URL: http://dx.doi.org/10.3886/ICPSR28024.v1
This survey was funded by:
- National Science Foundation (7805153)
- University of Michigan. Horace H. Rackham School of Graduate Studies
- University of Michigan. Institute of Public Policy Studies
- John Simon Guggenheim Memorial Foundation
Scope of Study
This data collection was created to study agenda-setting and alternative specification in the federal government. It concentrates on two federal policy areas, health and transportation, but the theories generated in the research may be quite widely applicable beyond those two areas. The aim of the work was not to study how issues are decided in some authoritative process like a congressional vote, but instead to study how issues get to be issues in the first place, how items rise and fall on the governmental agenda, and how the alternatives from which choices are made are generated.
The results of the study were published in John W. Kingdon, Agendas, Alternatives, and Public Policies (First Edition, Little Brown, 1984; Second Edition, HarperCollins, 1995; Longman Classics in Political Science Edition, Longman, 2003; Updated Second Edition, with Epilogue on Health Care Reform, Longman, 2011). The study's methods are described in detail in the Appendix to that book, and are included as part of the documentation for this data collection.
The major data source is a set of interviews that John Kingdon conducted in four waves (the summers of 1976, 1977, 1978, and 1979), with well-informed respondents either in the federal government (both congressional and executive) or involved in health or transportation policy around the federal government (e.g., lobbyists, journalists, academics, consultants). "Elite and specialized" interviews, to use Lewis Dexter's terminology (see Elite and Specialized Interviewing, Northwestern University Press, 1969), are conducted differently than standard survey research interviewing. The idea is to have a two-way conversation with a well-informed and highly involved respondent, rather than strict question and response. As such, the list of questions used was not a hard-and-fast interview schedule or questionnaire, but a kind of guide. The questions were not always asked in the same order, and indeed, not all of the questions were always asked. Question wording may have varied slightly from one interview to another. Various ad hoc probes were inserted as they seemed appropriate. Sometimes in this sort of interview, the interviewer makes a statement rather than asking a question. Still, the central questions were usually asked in roughly the same wording. Thus, when the interview write-up says "Q1," that is the first question in the standard list of questions used.
Interviews were not taped or otherwise recorded verbatim, since the principal investigator firmly believed that, with these sorts of respondents, taping dampened their ability and willingness to be candid. The principal investigator did not want respondents to feel that they were on the record, as respondents were accustomed to dealing with reporters, and when a microphone was in their face, they knew the encounter would be on the record. Notes were taken during the interview, and then written up immediately after; hence, the typescripts of the interviews are labeled "write-up" instead of "transcript." All 247 write-ups have a respondent identification number and the date of the interview on the top of the first page.
The principal investigator also coded the interview write-ups into quantitative data files, despite the nonrandom selection of respondents and the fluid conduct of the interviews. He did this to support quantitative judgments (e.g., "this issue was mentioned frequently in 1978 and not frequently in 1979," or "this factor was hardly ever mentioned in the interviews"). Each interview was coded by two coders, and then their judgments were combined. In addition to generic identifying information, there are two general categories of variables. One category, referred to as "global codes" in the codebook, is composed of ratings of the importance of each of several actors (e.g., mass media, president himself, interest groups, congressional staffers). The other category, referred to as "problem codes" is a coding of the problems that respondents discussed in their interviews, and is divided into health and transportation. A full description of coding procedures is contained in the data collection documentation.
Interview data are supplemented by a series of 23 case studies in health and transportation, and by some attention to other sources of data like congressional hearing records and public opinion data. In addition to various nonquantitative uses of the cases in the study, a quantitative dataset of the case studies was created. Two coders worked independently to judge each of a set of hypothesized influences in the case to be very important, somewhat important, of little importance, or not important. For example, after reading all of the materials for a given case study, a coder would rate the importance of congressional staffers as "very, somewhat, of little, or not" important. In contrast to the interviews, differences between the two coders were not resolved by a combination rule. Instead, the principal investigator and the two coders discussed and reached consensus in each instance in which there had been a disagreement. A full description of coding procedures is contained in the data collection documentation.
Date of Collection:
Data Collection Notes:
All interviews were conducted by John Kingdon. The first two waves of interviews (in 1976 and 1977) were all face-to-face, in Washington, DC. In the third wave, in 1978, Dr. Kingdon interviewed those he had interviewed in the previous two waves by long-distance telephone, and went to Washington for face-to-face interviews only for people he had not met before. In the fourth wave in 1979, he interviewed only people he had interviewed before, and therefore conducted all of those interviews by long-distance telephone. Research assistants under his supervision coded the interviews, generated the quantitative data, and produced the case studies and the coding of those cases. A description of the interview process is contained in the data collection documentation.
The principal investigator took great pains to preserve his promise of anonymity to his respondents. At the beginning of each interview, he assured respondents that "none of this is for attribution," which is a standard way to set the ground rules for an academic researcher. That phrase, "not for attribution," meant that he could use quotations from the interviews (as included throughout the published book), but that he would not attribute the quotation to the respondents or identify them except in the most general of terms. The interview write-ups have only an identifying number and date at the beginning of the write-up, not the name or location of the respondent. Users are required to follow these same rules of confidentiality.
Two separate versions of the coded interviews are available in this data collection.
In "Coded Interviews -- Full Categories," subjects are coded into one of seven categories. "Global codes" categories include: (1) Very important, spontaneously mentioned; (2) Very important, discussed in response to question; (3) Somewhat important, spontaneously mentioned; (4) Somewhat important, in response to question; (5) Little importance, spontaneously mentioned; (6) Little importance, in response to question; (9) Not mentioned in the interview at all. "Problem codes" categories include: (1) Very prominent; spontaneously mentioned; (2) Very prominent; in response to a question; (3) Somewhat prominent; spontaneous; (4) Somewhat prominent; response to Q; (5) Little prominence; spontaneous; (6) Little prominence; response to Q; (9) Never mentioned in the interview.
In "Coded Interviews -- Collapsed Categories," subjects are collapsed into one of two categories. For "global codes," categories 1 and 3 are combined into a single category labeled "very or somewhat important," and 2, 4, 5, 6, and 9 are combined into a single category called "prompted, little, or no importance." For "problem codes," categories 1 and 3 are combined into a single category labeled "very or somewhat prominent," and 2, 4, 5, 6, and 9 are combined into a single category called "prompted, little, or no prominence." Almost all of the data analysis in Agendas, Alternatives, and Public Policies is based on this collapsed version of categories.
Sample: This is not a random sample. The selection and proportion of respondents in health and transportation, over the four years of interviewing, and by generic affiliation, is presented in the documentation.
Mode of Data Collection: face-to-face interview, telephone interview
Original ICPSR Release: 2011-03-18
Use any of the notification links to add this study to your RSS feed; you will then receive notification if the study is substantively updated.
- Citations exports are provided above.
Export Study-level metadata (does not include variable-level metadata)
If you're looking for collection-level metadata rather than an individual metadata record, please visit our Metadata Records page.