Tuesday, March 4, 2014

New Temperance Flat Feasibility Study Claims Salmon Benefits and Delta Earthquake Risk Reduction Justify the New Dam and a Big Taxpayer Subsidy

I spent a good part of the afternoon reviewing the new feasibility study for the Temperance Flat dam and compared it to the one released in 2008.  The Bureau of Reclamation's claimed benefit-cost ratio in the new feasibility study is much higher than the one from 2008 that infamously found a B-C ratio of 1.0 to 1.06 despite the fact that the estimated water yield is lower.

Some observations about the benefit-cost estimates.

1.  Estimated construction costs of the dam dropped by nearly $1 billion (new estimate is about $2.5 billion compared to original of about $3.5b), even though the current estimate is in 2013 dollars and the old one was in 2006 dollars.  I am told the cost reduction is attributed to a change in the hydroelectric mitigation required.  Apparantly, it isn't true that estimated construction costs always go up.

2.  The new feasibility study justifies the dam for its ecosystem benefits to salmon.  It values the ecosystem benefits 2-10 times higher than the water supply benefits.  In addition to economically justifying the dam, this finding also is convenient for justifying a much higher taxpayer subsidy of the dam than proposed in 2008 (more on that later).  These multi-billion dollar ecosystem benefits (annual benefit estimates ranged up to $500m per year) result from the report's estimate that the dam will increase long-run average abundance of salmon from between -0.7% and 4.9% per year.  I'm not a biologist, but that doesn't seem like a huge benefit to me for a river that is projected to have relatively small salmon populations.  This recasting of the dam as a salmon project is very surprising to me as I am not aware of any environmental groups or fishery experts pushing Temperance Flat dam as a priority, and there are even some environmental groups who are opposed.

3.  The report is quite honest that the traditional water supply, flood control, hydropower, and recreation benefits that are associated with dams are not nearly high enough to justify the construction costs of this project.  And that's even after the report inflates these traditional benefits...

4.  In the "best" scenario, the report estimates $19 million in annual agricultural water supply benefits from an average increase to ag. water supply of 41,000 af.  That's a healthy $461 af in current dollars, a value that is about 3 times higher than typically used for incremental ag water in benefit-cost assessments.  Given the special role of agriculture as the economic base of the Valley, I have sometimes argued for using a more generous economic development measure that includes multiplier effects.  Like many of these assessments, this feasibility study also calculates the economic development value in a separate section.  In Table 5-12, the report estimates this annual value at $10.8 million for agriculture, which seems about right for 41,000 af of annual yield.  The strange thing is that this economic development value is lower than the value used in the benefit-cost estimate, and it is usually the other way around.  This seems to confirm my suspicion that the $19 million value associated with ag. water supply reliability is an error.  Bottom line, the agriculture water supply benefits are overestimated by a factor of 2-3, at least $10 million per year.  [Update:  I have now seen the technical appendix, and it turns out that this huge agricultural value results because the model they are using allocates the majority of new agricultural water produced by Temperance Flat to recharge groundwater where it has a much higher economic value than growing crops. This is an interesting finding and if it accurately measures the value of recharging groundwater and/or the external cost of pumping groundwater on other users of the aquifer, it makes a powerful argument for regulation of groundwater.]

5.  Delta earthquake and flood protection benefits.  The feasibility study estimates $25 million in annual benefits from emergency water supplies Temperance Flat would provide in the event of a catastrophic Delta flood.  This benefit is inflated due to ridiculous assumptions about levee failure probabilities among other issues.  As a point of comparison that shows the foolishness of this number, it is almost identical to the risk-reduction benefits the BDCP estimates for the Delta tunnels which are thought to preserve several million acre feet of water exports in the case of these catastrophic events.  [The BDCP estimates 50 years of this benefit has a present value of $364m to $460m, use the present value formula to solve for the annualized value and it is in the neighborhood of $25 million annually.  This Temperance Flat study allocates over $400m of construction costs to taxpayers due to this benefit.]   

6.  The benefit-cost analysis uses annual costs and benefits.  It annualizes capital costs over 100 years with a 3.75% discount rate.  That is a very generous assumption, and it understates the annual costs.

Some observations about the proposed cost allocation for Temperance Flat.

1.  Only 26% of the cost of the dam would be allocated to water users (12% ag, 14% municipal/industrial).  In contrast, the 2008 study of the dam allocated the majority of costs to water users.

2.  About 73% of the cost of the dam would be paid by federal and state taxpayers.  The 73% allocation can be broken down into three general categories of claimed public benefits: 49% ecosystem, 8% recreation/flood control, and 17% emergency water supply benefits from a Delta flood (see #5 above).  This 73% share is only direct construction costs, and does not count the subsidy in the Bureau's 0% financing of agricultural users contribution.

So that is about $1.25 billion in taxpayer dollars towards dam construction for claimed salmon benefits (direct costs, this doesn't count interest costs on the water bond that would finance the state's share).  I wonder what a salmon expert would do if you gave them $1.25 billion of taxpayer funds and said spend this to improve salmon habitat.  

In addition, it allocates 17% of the dam's costs (nearly $500 million) to state/federal taxpayers due to the Delta flood risk reduction benefits (see #5 above).  Taxpayers might prefer to spend $500 million in Delta risk reduction would be better spent directly on Delta levees themselves - providing flood protection benefits for water supplies and protecting property, other infrastructure and lives in the Delta. 

Some observations about financial feasibility calculations for water users.
Unlike BDCP, this report correctly proposes a cost allocation before making any conclusions about financial feasibility. 

1.  Agricultural water supply is allocated $264 million of construction costs.  Assuming 40 year repayment period with no interest, and operating costs comes to $8.7 million per year.  The report estimates the cost of the incremental water supply to the agricultural users is $212 af.  That's a hefty cost for agricultural water, and note that this is the cost even with the Reclamation's generous no-interest financing and taxpayers picking up 73% of the estimated construction cost of the dam.

2.   Municipal and industrial water supply is allocated $362 million of construction costs.  The report assumes a 40 year repayment of capital costs and 5.37% interest.  Principal, interest, and operating costs come to $27.4 million annually, and the incremental water supply cost to M&I users is $1,305 per acre foot. That's a pretty expensive municipal and industrial water supply, even with taxpayers picking up 73% of the estimated cost of the dam.

Overall, it's not a very convincing feasibility study, and I don't believe it provides strong economic justification for Temperance Flat dam.  It's disappointing, because I believe in the value and need for storage and I would like to be able to support storage projects in the Valley.  But there are better uses of taxpayer dollars for these and other purposes, and the water it is still an expensive option for water users even with the large taxpayer subsidies.  


  1. Hi Jeff

    This analysis is great. I hope to use some of it in my env. econ class on CBA.
    I would like to clarify a few things if you can -

    For the agricultural reliability numbers - are they claiming that water put into the ground is worth $461/AF? Is this all therefore reflecting reduced pumping costs? Seems way too high. Or is it that the water can then be more available in times of extreme shortage when it is indeed worth that much or more?

    Also, water stored in Temperance Flat has an opportunity cost, right? How valuable were those flood releases to the Delta that will now be captured, and why aren't they listed as a cost if this new reservoir will stop them?

    And same thing for recreation visitation - isn't some of this bound to be reallocation from other nearby sites? Like from Millerton? So they are missing out on that cost as well (or overestimating the benefit).

    And recreation in general seems overvalued - they used the mean value for different recreation activities from Loomis' study. That could be OK (maybe the median is better, I didn't read Loomis' study closely), but when I see a camping value of $45 per day, that seems a bit high. Correct me if I am wrong, but these are total benefit values, and so even if the numbers are correct, the correct recreation value to include here is consumer surplus, which would be $45 minus the cost of camping (the fee, getting there, etc.). And hiking has a value of $37? That seems way too high - most hikers in that area are local, and so suppose it costs a car $20 to get there - wouldn't this imply that this car would be willing to pay up to $17 in fees just to hike this area? Despite the other close substitutes?

    Thanks for your help here - I hope to share this with more people so they understand the problems with this particular project.

  2. Nice post and such a remarkable information that you gave to us. Thank you so much for it.

    VAT Consultancy in UAE