The objectives of the study were to:



  • Gain an understanding as to the reasons for and benefits of interning on Capitol Hill from interns, themselves;

  • Determine the extent to which “hook-ups,” “flirtation,” “dating” and “sex” were discussed or witnessed within the Capitol Hill offices;

  • Distinguish the concerns, challenges, and priorities of Capitol Hill interns relevant to their personal and professional lives;

  • Understand the differences, if any, between drinking alcohol beverages while at college, and while an intern; and,

  • Discern their opinions and projections with respect to public policy issues and long term professional goals.

In order to qualify for participation, interns must have been age 18 or older, interning in one of the Senate or House office buildings, and having finished at least one year at a college or university.


Respondents were personally contacted by tpc staff members, each trained in data collection. All respondents completed the survey in the direct vicinity of Capitol Hill, and were asked to show their “Capitol Complex” intern identification and otherwise verify their current status and qualifications before participating.


Participants were recruited in several locations, including House and Senate office building cafeterias and hallways, in the offices of individual members and committees, on pathways between the Union Station and Capitol South Metro stations, and outside on the grounds of the U.S. Capitol. Interviews were secured in a variety of locations and over five consecutive business days in order to secure a representative sample. The survey instrument included 28 questions. The questionnaire was designed and approved by the client prior to data collection. Interviews were conducted June 22-28, 2005.


The margin of error for the survey is +5.9% at a 95% confidence interval, meaning that in 19 out of 20 cases, the data obtained would not differ by any more than 5.9% in either direction were the entire population of Capitol Hill interns surveyed. Margin of error for subsets of the population is higher.