23 Apr 2009

Dangerous Consultants

In contrast to the careful study above (academics at their best), we (via DB) have this disaster of a "report" [DOC] given to the SFPUC.

Here's what RFC says:
Since the SFPUC is facing large rate increases over the next several years, it was important to get a sense of the price elasticity of water demand in the service area. As part of this analysis, RFC analyzed the per capita consumption and commodity rate increases from FY 1994 through FY 2007. Using only the commodity rate increase is imperfect, but is the best use of the historical information available.
They go on to estimate (by fitting a line between rate changes and changes in use) that total price elasticity is -0.08, which is MUCH lower than the -0.40 to -0.70 range that I have seen here, there and everywhere...

As RFC admits, "attributing all changes in consumption to changes in rates is questionable" but they then go on to commit the sin of all time, i.e., "we cannot recommend factoring in reduced consumptions to accompany the proposed rate increases," i.e., do not consider how price will affect demand when calculating the impact of higher rates!

Excuse me, but the lack of evidence of an impact of higher rates on demand does NOT invalidate the law of demand.

As RFC admits, their "analysis" omits economic conditions and weather patterns, but it also omits important factors such as income (we spend a tiny portion of our income on water bills), block rates, billing frequency, etc.

I would give an undergraduate who produced this "analysis" -- especially considering its weak economics -- a grade of D.

Bottom Line: I hope that SFPUC got this report for free, since it's worth nothing.


  1. Just eyeballing it, it looks methodologically crappy (if I can get technical). What time frame do those elasticities you cite apply to? My father (a civil engineer) has told me that water, unlike gasoline, tends to be more elastic in the short run than the long run.

  2. @dWj -- those are contemporaneous (ST) elasticites. All the studies I've seen say that elasticity is greater in LR. I'd say that he is RIGHT in the sense that ST drops in demand can be big, but LT (permanent) drops only occur when demand destruction happens.

    It's the difference between not flushing and a low flow toilet...

  3. David,

    Overall, I agree with your thought that price elasticity is an important factor to account for in setting water rates. However, the price elasticity factor needs to be adjusted for local demographics. While a -0.4 factor might be reasonable for a some users, residential users tend to be less price sensitive, particularly multi-family users.

    In the case of San Francisco, a couple of issues need to be taken into account when defining the price elasticity factor. First, the mix of customers. Roughly, 65% of housing units within San Francisco are renter occupied/multi-family (2000 census). Because the City does not sub-meter multi-family users, there is no nexus between water usage and person paying the bill. Second, San Francisco’s per capita water usage is extremely low (relatively speaking). Water consumption dropped significantly during the last big drought and has remained level sense then, even with new development. While certain areas of the City are more price sensitive, a lot of water savings have already been achieved. The City invested heavily in a retrofit/conservation program targeting lower-income areas of the City for individuals that could not necessary afford to replace toilets, washers and shower heads.

    Again, overall, I agree with your hypothesis. However, in the case of San Francisco, the price elasticity factor needs to much lower than -0.4. That being said, the analysis that RFC performed in developing their elasticity factor was not robust (i.e., univariate).

    As a disclaimer, I work as a rate consultant, albeit for a different firm (FCS Group, located in downtown SF).


Read this first!

Make sure you copy your comment before submitting because sometimes the system will malfunction and you will lose your comment.

Spam will be deleted.

Comments on older posts must be approved (do not submit twice).

If you're having problems posting, email your comment to me