Learning from the Leaders: Results-Based Management at the Federal Emergency Management Agency

Summary

Viewed by many as one of the most incompetent agencies of the federal government in the 1980's, the Federal Emergency Management Agency began a reorganization effort in 1993 and emerged as one of the most highly complimented agencies. Specific results measures highlight the agency’s success:

  • In fiscal 1998, FEMA took an average of 8 days to get relief checks to disaster victims, down from 10 days in 1997 and a high of 20 days in 1992.
  • Between 89 and 97 percent of disaster aid recipients rate FEMA favorably on ease of access, clarity of information, promptness of aid, compassion, and overall quality of service.
  • State, local, and nonprofit officials who have received disaster aid from FEMA give the agency positive ratings in the 70-80 percent range, exceeding baselines of 60-75 percent.
  • FEMA's mitigation programs prevent more than $2.00 in disaster losses for every dollar spent.
  • Enforcement of the National Flood Insurance Program's building standards prevents flood losses of $750 million annually.
  • In fiscal year 1998, between 76 and 87 percent of enrollees in FEMA training courses said they learned things that improved their job performance. 

A careful analysis reveals that no single variable explains FEMA's success. Rather, FEMA improved its results through a collection of different but related management reforms. The agency did a number of things right at the same time, and the results for the whole exceeded the sum of the parts. These kinds of changes are not only applicable to FEMA. They illustrate general prinicples that can guide any government agency seeking to transform itself from a rule-driven bureaucracy to a results-driven organization.

To speak with a scholar or learn more on this topic, visit our contact page.

Mercatus AI Assistant
Ask questions about this research.
GPT Logo
Mercatus AI Research Assistant
Ask questions about this research. Mercatus Chatbot AI More Details
Suggested Prompts:
Ask us anything. We use OpenAI's ChatGPT 4o base model to answer any question about Mercatus research.