The problem with “drought”

I’m increasingly at a loss about how to do useful journalism because of the gap between the technical world I try to understand (risks associated with water contamination, for example, or the problem of drought) and the public reaction to my efforts to provide a nuanced but accurate explanation of the thing at hand.

Mark Lubell at U.C. Davis gave nice voice to my queasiness in a piece this week on, among other things, the public reaction to water contamination in Toledo. He maps nicely maps the issue onto the difficulty of communicating about drought:

Take the current California drought. Yes, it really is bad. But as I’ve pointed out elsewhere, there have been bad droughts before. Droughts activate what Kahneman and Tversky call the “availability heuristic”, where dramatic events increase our estimation of the actual probability of the events. For example, news about a grizzly bear attack causes most people to over estimate the risk–you’re much more likely to die in a car accident on the way to work than be attacked by grizzly in Glacier National Park.

Hence, drought is a source of “panic politics”, with knee-jerk policy recommendations often overwhelming more reasoned analysis. For example, one of the first recommendations is “build more surface storage” but in reality even building out all of the potential surface storage sites in CA will only be a band-aid solution. A much more integrated portfolio of drought management strategies is required, including both demand and supply side tools. Groundwater management is one obvious issue that deserves attention.

An oft-heard phrase among water policy stakeholders is “don’t let a good crisis go to waste”, and perhaps there is some wisdom in this saying (although its overuse annoys me). The “water planner” in me wants to respond “but let’s do rational planning BEFORE the drought”. But that goes directly against human psychology, so maybe we should figure out ways to maximize the policy change that is possible from agenda-setting events like a drought. On the flip side, the availability heuristic also explains why rain literally dampens the water policy agenda–the risk doesn’t seem as obvious. But in all of these cases, the underlying probabilities are driven by climatic variables, while the risk perception is driven by psychology.

This resonates. I try to explain drought, and people freak out. It’s been raining for a month in New Mexico, and now people are ready for the freakout to be over. Neither is entirely appropriate, and Lubell’s model nicely explains why, but also why the results I’m seeing are probably inevitable.


  1. The word “drought” is already defined in a very negative way and used by most people such that it is almost impossible to use it in a nuanced way. Wiki states its a “deficiency”, Google defines it as “abnormally low”. You almost need a different word or define it right from the beginning. ” The low side of the long term average resulting from the expected natural variability of rainfall.”.

    This may be a result of in this country there not being long term rainfall records and few studies using proxies, or that people just like to have drama (“panic politics”). Everybody thinks the next 20 years is going to be like the last twenty. It’s a major reason we are in this situation, i.e., using the available flow records of 1905-1922 to define the average flow of the Colorado.

    I appreciate your use of the term and the information you give on the Colorado River and how while the upper basin is being impacted by the natural variability in rainfall, the lower basin is being impacted by over-allocation. Even Secretary of Interior Salazar made the same point on his discussion of Reclamation’s 2012 supply and demand study.

  2. Pingback: Bloggers on the water bond, drought, legislation, habitat restoration and more, plus will conserving water contribute to global warming? » MAVEN'S NOTEBOOK | MAVEN'S NOTEBOOK

Comments are closed.