Bureaucratic Reputation Analysis

Carpenter and Krause distinguish performative reputation (the agency can do its job and produce results), moral reputation (it is honest, compassionate and protects the vulnerable), technical reputation (it has genuine expert and scientific capacity), and procedural reputation (it follows accepted, fair and lawful processes). Agencies can be strong on some dimensions and weak on others, and threats often target a specific dimension, which is why analysing them separately is central to the framework.

Carpenter argues that when an agency is esteemed across diverse and influential audiences — legislators, professionals, the media, the public — political principals find it costly to override or constrain it, because doing so would provoke backlash from those who trust the agency. A strong, multidimensional reputation thus creates room for discretion and policy innovation that statute alone would not grant. Autonomy is forged over time through demonstrated competence, network-building and protective behaviour, and it erodes when reputation is damaged.

There is no single metric; reputation is reconstructed from evidence about audience beliefs. Researchers code media coverage, legislative testimony and oversight records, analyse interest-group and professional communications, examine archival documents, and where possible use opinion or survey data. Increasingly, scholars apply text and sentiment analysis to large media or document corpora to track reputational dimensions over time. The construct remains partly interpretive, which is why triangulating multiple sources is the methodological norm.