Opportunities for Statisticians in the Federal Government

1
Opportunities for Statisticians in the Federal
Government

• Clyde Tucker
• Bureau of Labor Statistics

2
Overview

• Pros and Cons of Government Employment
• The Recruitment and Hiring Process
• Divisions of Labor
• Expectations and Training
• Career building
• Succession

3
Pros of Government Employment

• Job Security
• Great Benefits (defined pension, matching 401K,
  lifelong health insurance, generous vacation and
  sick leave)
• Pleasant surroundings in relatively low-pressure
  environment
• Reasonable hours
• Work on important problems

4
Cons of Government Employment

• Money not as good (especially at higher grades)
  as in the private sector
• Many jobs restricted to the Washington, DC area
• High cost of living (but jobs for two earners)

5
Recruitment and Hiring

• U.S. citizenship required
• Cumbersome application process
• Protracted hiring period
• Limited negotiation room
• Fair and impartial
• Truth in advertising

6
Divisions of Labor

• Production vs. Research
• Survey statistics vs. other applications
• Mathematical statisticians vs. other statisticians

7
Production vs. Research

• Production
• usually less technical training
• majority of government statisticians
• day-to-day operations
• assigned to specific programs
• Research
• more likely to have Ph.D.
• housed in research units
• may work across programs

8
Survey Statistics vs. Other Applications

• Survey statisticians
• involved in data collection
• sampling primary activity
• concern about minimizing sampling error
• focus on survey statisticians
• Other applications
• biostatistics
• agriculture
• quality control
• engineering
• operations research

9
Mathematical Statisticians vs. Other Statisticians

• Mathematical statisticians (1529s)
• usually the research statisticians with at least
  masters in statistics
• focus on sampling error and estimation but also
  consider methods of dealing with nonresponse
  (Hansen and Hurwitz 1958) (imputation and
  weighting)
• some consider other forms of nonsampling error
  like the interviewer (Hansen, Hurwitz and Bershad
  1961)

10
Mathematical Statisticians vs. Other
Statisticians (Cont.)

• Other Statisticians (1530s)
• those involved in operations
• over time, social scientists interested in
  nonsampling errors
• some statistics background and often classified
  as 1530 statisticians
• often Ph.D.s and called survey methodologists
• research-oriented like mathematical statisticians
• often work across programs in research units

11
Expectations and Training

• Desires of agencies
• blend of theory and application
• computer programming
• sampling (survey agencies)
• how to approach real-world problems
• writing and communication skills

12
Expectations and Training (cont.)

• Realities of university education
• hard for a statistics program to live up to this
  ideal
• not so well-rounded in the arts
• statistics and math attracts those interested in
  the abstract and solitary pursuits
• professors publish in theoretical areas
• advice is likely to be to take more rigorous
  courses
• who would teach writing and communications beyond
  basics

13
Expectations and Training (cont.)

• Ways to prepare students
• statistical consulting (Snee 1982 and Zahn 1982)
• more interaction between government and academia
  (Savage 1978)
• internet helps students learn about government
  statistics

14
Expectations and Training (cont.)

• New programs
• government agencies do the teaching (Currie et
  all. 1986 and Lyberg 2002)
• some of government programs more on day-to-day
  operations
• Lyberg describes new university programs, but
  some may be stronger on social science side

15
Career Building

• Situation
• salaries at higher levels less competitive, and
  relative expenses increasing
• can be offset by other aspects of job
• Retention
• retention bonuses
• educational benefits
• tuition
• release time or sabbaticals
• rotational assignments
• promotions and more responsibility (production)
• wealth of data and no publish or perish
  (research)

16
Career Building (cont.)

• Future career direction
• if stay, ultimately greater responsibility
• likely to lead in some capacity
• more exposure outside own office
• need training in communication skills (Killion
  1995)
• often neglected
• increasingly important as move up
• first step in developing better interpersonal
  skills
• critical for senior-level positionsotherwise
  positions go to non-technical staff

17
Succession

• So much now written in this area
• Walsh (2006) says particular problem in Federal
  government because workforce older
• Right now fewer younger statisticians to draw on
  and hard to recruit mid-level ones from outside
• Good time to get in on ground floor