WikiSummary, the Social Science Summary Database

Moe: Control and feedback in economic regulation

From WikiSummary, the Free Social Science Summary Database

This summary needs formatting (i.e. "wikification"). Can you help us improve it? (Formatting help.) Please volunteer.

Moe. 1985. Control and feedback in economic regulation: The case of the NLRB. APSR 79: 1094-1116.


Bureaucracies are influenced by a variety of forces. Models that focus only on one-way, diadic relationships (e.g. Congress and the bureaucracy) in the name of parsimony do so at the cost of great losses in accuracy. In fact, several forces affect bureaucratic behavior. Moe proves his point through a statistical analysis of voting at the National Labor Relations Board. All the factors below matter.

Two-tiered hierarchy:

  • Top tier: political authorities, who will bring influence down on their appointees
  • Second tier: political appointees will influence their staff
  • Unclear just how much influence the president has; there's a shift with each new administration, but it's moderate.
  • Congress appears to have more influence than the president, but staff has more influence than Congress
  • And staff is influenced by the board, which is influenced by the president
  • So it's hard to tell who has the most influence, and where the influence is coming from.


The NLRB makes rulings for/against complaints submitted by either labor or business. Step 1: A complaint is filed. Step 2: NLRB staff decide whether the complaint has enough merit to get a vote. Step 3: The five NLRB directors vote on the complaint. These three steps all affect one another. Step 1 is influenced by the expected outcome of Step 3 (does the current board favor labor or business?). Step 2 is influenced by both Step 1 and by the staff's expectations about what will happen in Step 3 (staff will forward more labor complaints to the board if the board is more pro-labor). And Step 3 is affected by what happens at Step 1 and Step 2. So Steps 1, 2, and 3 are all endogenous factors.


The president can affect the NLRB, primarily through his appointment, but also through executive orders and other mechanisms (table 3). The Congress can affect the NLRB through committee hearings, publicitiy, and other mechanisms. And the courts can overrule the NLRB, deterring it from deviant decisions.


Unemployment, inflation, and unionization can affect whether labor or business is more likely to submit more claims (table 5). They also affect whether NLRB staff favors labor over business (table 4).


  • Exactly how much does each factor matter in relation to the others? (sort of answered in table 2)
  • Okay, so all these things can constrain the NLRB. But exactly how much do they constrain it? How much autonomy remains?
  • Moe presents evidence that presidents can constrain the NLRB, but this is based primarily on evidence that conservative presidents appoint conservative board members (and vice versa). How much influence to presidents have after the appointments are made?