Culture, Cognition, and Group Dynamics at the FOMC

Cognitive sociology alerts us to the notion that individuals cannot process huge amounts of information and that a model of human agency that assumes full knowledge and rational action is therefore improbable (Martin, 2010).

Cerulo (2006) shows how such schemas are shaped by the nature of social groups. She argues that the structure and culture of most organizations lead ac­tors to either downplay negative information or reinterpret it in a positive light. The latter is similar to another endemic feature of organizations, which Vaughan

Table 8.ι: Voting members of the Federal Open Market Committee, 2000-2008

THE BANKS DID IT

ts∙

Data source: Federal Reservefor names and terms of service and Wikipediafor biographical information.

Why Did the Federal Reserve Miss the Financial Crisis of 2008?

tdecision-making autonomy), presidents who do not vote, and staff who present reports but rarely take positions of their own.

Another way that looking for a positive outcome comes into play is that there is some pressure in the meeting to try to attain consensus as to whether to change interest rates. So, while people may disagree on how to add up the facts and details, at the end, they work to form a consensus opinion as to what the FOMC should do. A consensus generally follows a median path, not too radical in either direction. This means that if things might be looking bad, the tendency will be to not be alarmist and to go along with the majority until the evidence is so clear that no one can deny it. People with outlier positions will back off in the end and support median opinion. That consensus will most often be optimistic that whatever is going on, the FOMC has time to make informed decisions.

Prior research has demonstrated the presence of such tendencies at the FOMC. Abolafias (2004, 2010) analyses of FOMC meetings from the 1980s and 1990s show that meeting participants construct a narrative account of the economy. Much of the give-and-take of the conversation is oriented toward making sure that small disagreements do not turn into large disagreements. He also shows that there is a kind of groupthink whereby discordant facts tend to be ignored. This observation suggests that the tendency toward thinking the same thing allied with a tendency toward positive bias means that the outcome leans toward optimism.

Framing and Cognition at the FOMC: The Role of Macroeconomics in Discussion of Housing Market “Fundamentals”

House prices nearly doubled from 1995 to 2005. Rising home values directly stim­ulated residential investment and indirectly boosted consumption as borrowers extracted equity when they refinanced. By the end of 2003, however, the market for conventional mortgages was saturated, and so mortgage originators began aggressively targeting borrowers unable to qualify for prime loans.

The meeting transcripts reveal that committee members were aware of the his­toric run-up in housing prices. Figure 8.2 shows that they discussed the possibility of a house price bubble more and more over time. But they generally made sense of that run-up by referring to the “fundamentals” of supply and demand in the physical stock of housing. At the August 13, 2002, meeting, for instance, Chairman Alan Greenspan raised but quickly rejected the possibility of a housing bubble:

There is clearly concern at this stage about a housing value bubble that is going to burst. But I think that most of those who look at this in some detail question whether that's a valid notion... the impact of immigration superimposed upon the difficulty of finding viable land for homebuilding is keeping significant upside pressure on home prices over and above the construction productivity issue. So, the likelihood of any really important contraction in the housing area would in my view require a very major contraction in the economy overall. (FOMC, 2002: 74)

Federal Reserve governor Ned Gramlich concurred with Greenspan's assess­ment, adding, “I agree that there is no bubble. I agree that there is no produc­tivity change in housing construction. So, the relative price of housing is rising compared with income and other prices for what we’ll call real reasons” (FOMC, 2002: 82; emphasis added). The causal model of house-price appreciation em­ployed by Greenspan and Gramlich thus assumed the following form: immi- gration—a “real,” demographic factor—was increasing the demand for homes (relative to a fixed supply of land and static construction productivity), which in turn was raising the price of housing.

Committee members were attentive to the wave of mortgage refinancing that preceded the subprime expansion. Yet when the refinancing boom subsided in late 2003, the FOMC showed little recognition that subprime mortgages were re­placing refinancing as the driving force behind mortgage and housing markets. Indeed, a simple search of the transcript corpus reveals that the word subprime appears a mere four times before mid-2005—and not at all in 2004, when the subprime market became the largest part of the mortgage market.⁶ In fact, the first time that the committee made any reference to changes in the composition of the mortgage market was February 2005, after these changes had already oc­curred (FOMC, 2005a: 119).