Is America Suffering from Bad Data?
I wonder whether the Administration and congressional negotiators are asking the same questions Thomas Redman would ask. Redman loves good data, hates bad data. Writing on Harvard Business School’s blog, the president of Navesink Consulting Group says:
The credit crisis presents an object lesson in data quality. Incorrect data on mortgages, borrowers’ creditworthiness, and securities ratings… missing data about the makeup of CDOs… and balance sheets that could not be trusted all contributed to the current situation.
To make matters worse the company leaders, regulators, and law-makers who are trying to resolve the crisis can’t trust the data available to them now. And, since economies and markets are so intertwined, it is impossible to predict the impact of the decisions they are making.
What actions should leaders, in this or other crises, take in the near-term to improve the data they’re using? First, they should make two lists as carefully as they can:
- The data they’d really like to have as inputs for resolving the crisis
- The data they do have
The first area of focus should be on items they’d like to have but don’t. Decision makers in crisis simply must try to obtain this data. They are often more readily available than expected.
Second, they should carefully evaluate each data item they do have, asking the following questions:
- Where did this data item originate? Can that source be trusted? Does that source have a stake in the decision? Is the source holding anything back?
- Are there independent sources? Do they confirm the data item?
- Was the data item the result of a carefully-managed process? If so, how reliable is the process?
- Is the data item an estimate? If so, how good how similar estimates proven? Is the item an opinion or anecdote?
- How will crisis resolution be impacted if this item is wrong?
Leaders must seek alternative sources on all data that rate low on trust and can seriously impair the resolution of the crisis if wrong.
It’s tempting to say America is in a “data crisis,” except the data quality was far worse in previous dangerous moments in U.S. history. We go to war because of bad data. The Great Depression and all the market panics that preceded it were caused by bad data. It seems to me that we’re on the cusp of an epoch of fabulous data, with the ability to rapidly identify every needle in every haystack through sophisticated data-mining. On Wall Street, that might already be the case.
No, it seems to me that this crisis was brought about by the things that don’t change about human nature: Greed, wishful thinking, and a refusal to face uncomfortable facts. More and better data just means we’ll have more and better data to ignore.
