CRJU/POSC 320: Introduction to Public Administration
This is the last content lecture of the course, and it answers the question every previous module has been circling. We’ve given public administrators an enormous amount of power — to tax and spend, to permit or deny, to decide who’s eligible for benefits, to inspect and license private activity. Modules 5 through 14 were largely about how they use it.
This module is about the check on it: how democratic societies hold administrators accountable while still letting them do their jobs. That "while still" is the whole difficulty. An agency you can’t control is dangerous. An agency you’ve controlled so tightly it can’t act is useless.
Pair this lecture with the Accountability Mechanisms handout — it carries the detailed breakdown of each mechanism, so I’m keeping this short and focused on the logic.
Accountability = the obligation to answer for one’s actions to some higher authority. It has three components, and a system missing any one of them isn’t really accountability:
A city manager has all three: answerable to the council, enforceable through firing, controllable through the charter. Compare that to an agency that files reports nobody reads and faces no consequences — answerability without enforceability is theater.
CJ students: this is why "the department investigated itself and found no wrongdoing" is such a recognizable failure. Answerability was satisfied; enforceability wasn’t.
Administrators answer to five different masters at once, and the demands don’t always agree:
The handout works through each one’s mechanisms, strengths, and limitations. The point for lecture purposes is that these can pull in opposite directions: what your professional judgment says is correct may be what your elected board doesn’t want to hear.
Two models for how oversight actually happens — the terms come from the legislative oversight literature, but they describe every oversight relationship:
Police patrol is continuous and proactive. Regular budget reviews, performance dashboards, scheduled inspections. It catches problems early — and it’s expensive, because somebody has to be watching all the time.
Fire alarm is reactive and event-driven. Somebody outside the agency — a journalist, a complainant, a plaintiff — pulls the alarm, and oversight responds. It’s cheap, because you only pay when something’s wrong, but it depends on someone noticing and having enough power to be heard.
Legislatures overwhelmingly prefer fire alarms. Limited time, limited attention, and the political credit for responding to a scandal is much larger than the credit for preventing one. Which means most oversight arrives after the harm.
Internal mechanisms — chain of command, internal affairs and QA, performance management, professional standards — have the knowledge. Insiders understand the work, know where the bodies are buried, and can act quickly.
External mechanisms — civilian oversight boards, investigative journalism, interest groups, judicial review, inspectors general and auditors — have the independence. They aren’t asking colleagues to indict colleagues.
The tension is structural, not a design flaw: internal knowledge versus external independence. You can’t maximize both, which is why nearly every real accountability system layers them. A public works department reports to the city manager (executive), is funded by the council (legislative), constrained by courts (judicial), and regulated by state agencies (administrative) — four watchers, none of them sufficient alone.
CJ students: substitute "police department" and this is the textbook example — internal affairs plus a civilian review board plus consent decrees plus the press.
Knowing the mechanisms matters less than knowing where they systematically fail. Four recurring gaps:
Information asymmetry. Agencies know more than their overseers, and they control what gets disclosed. Add gaming — hitting the metric without improving the performance — and oversight can be formally satisfied while nothing changes.
Diffused responsibility. When federal, state, local, and contracted actors all touch a program, nobody clearly owns the failure. Privatized service delivery is the sharpest version: private contractors face profit incentives, restrict access to their operations, and often serve populations with the least voice.
Protections that shield poor performance. Civil service protections, adjudicator independence, and self-policed recusal all exist for good reasons — insulating professionals from political retaliation. They also make removing bad performers genuinely hard.
Attention scarcity. Electoral cycles are short, problems are long, and media attention moves on. Sensational incidents crowd out systemic issues.
CJ students: qualified immunity, union protections, prosecutorial discretion, judicial tenure, and private prisons map onto these four directly. Same structure, higher stakes.
None of these gaps have clean solutions, and be suspicious of anyone who tells you otherwise. Every fix — more monitoring, more metrics, more external review — costs autonomy, speed, or money. That’s the trade-off this whole course has been about.
Key takeaways: