Free Riding and the Financial Crisis

We've heard many people describe contributors to the financial crisis:

• Alan Greenspan kept interest rates too low in the mid 2000s
• Freddie Mac and Fannie Mae's quasi government status provided too much easy credit into housing
• A speculative bubble fueled by greed and "animal spirits"       

I agree that each of these points were important enablers.  I'd like to address one additional contributor, that  worked in conjunction with those points above, that was important in creating the financial crises.

An important economic concept is the notion of "free riding."  Free riding occurs when one person's activities provide as a group benefit. It is difficult to make those parties receiving a benefit from the activity from bearing a proportional cost of the activity.  An often used example of a free riding problem is police or personal safety.  If policing was privatized,  those who purchased protection would likely provide some unintended protection to their neighbors. The security guard that would be on watch would prevent crime for the client, and also the client's nearest neighbors as well.  Many economists think that the total personal safety expenditures would be smaller than we would actually prefer, because many people would hope that others would purchase protection, allowing them to "free ride."  Of course, if too many people feel this way, total expenditures on a valuable service will be quite small.

Free riding can also occur when ambiguity exists.  When two or more parties are involved in completing a complex activity, and there is imperfect information between them, often times each of them will attempt to free ride on the others actions. It is inefficient and costly to duplicate work; it is natural to seek to avoid such duplication.  The risk becomes when all parties, in the name of efficiency, assume another party is addressing a critical activity in the process. Therefore, unless there is clearly defined responsibility, accountability, and transparency, the potential that all parties are trying to "free ride" exists, so that the activity in question does not get performed appropriately.  As the process becomes more complex and involves more parties, the risk of such a free riding problem becomes greater.  


I believe the financial crisis was largely subject to such a pervasive free riding problem.  Fueled by the issues listed in the bullets above, the mortgage finance industry expanded in complexity and the number of involved parties.  For example, the archaic model of mortgage financing is that a bank or savings and loan would make a mortgage loan to a home buyer and then hold the loan until it was paid.  On the financing side, there was one party - the bank or savings and loan.  In the modern mortgage finance model, the following parties may all be involved:

• Mortgage broker
• Mortgage underwriter and initial lender (bank, thrift, or mortgage finance company)
• Wholesale lender
• Mortgage security underwriter (investment bank, Fannie Mae, Freddie Mac)
• Credit rating agency (S&P, Moody's, Fitch)
• Institutional investors

So, we've moved from a process involving a single party to one that can involve up to six parties.  Whereas the underwriters were responsible for determining which loans to make, and suffered the consequences if they made poor decisions, such decisions and consequences could be spread among five parties in the new financing model (the credit rating agency bore no direct financial risks).  Importantly, all six of these parties shared some responsibility to investigate the quality of the loans:

• The mortgage broker and underwriter were responsible for investigating the borrower.  
• The wholesale lender was responsible for performing due diligence on the mortgage underwriter.
• The security underwriter, credit rating agency, and institutional investor were responsible for performing due diligence on the portfolio of mortgages.

Without our hindsight knowledge, it would be easy to conclude that there were robust controls in this process, and therefore the mortgage investments that were distributed must be sound.  However, I'm convinced such complexity and lack of accountability and transparency created a massive free riding problem.  What actually happened was, in many cases, extensive cutting corners by each of these parties, under the assumption the other parties provided a safety net, or at least plausible deniability (as in "I was informed that [fill in other company's name] was responsible for addressing that risk").

My takeaway - when processes add layers of complexity and the number of parties involved increases, the risk of free riding increases.  If there is clear accountability and transparency in all aspects of the process, free riding is much less likely to occur.  Importantly, many times it is in the interest of parties involved to blur accountability and reduce transparency.  Doing so allows them increased flexibility and reduced costs, under the veil that another party shares their responsibilities and can be relied upon.  In the financial crisis, it took massive losses to illuminate what each of the parties involved in the process were actually doing.  Perhaps it's possible to identify environments subject to such free riding problems going forward and thereby avoid participating in the resulting losses.