28 January 2009

Farmers Don't Use Much Water

NB: I updated this post on 20 Aug 2014 to replace and update missing "food dollar" data.

This post is important and perhaps paradigm shifting. Let's see if you agree...

It's conventional wisdom that farmers "use" 70-80 percent of all developed* water supplies. But farmers do not use water in the same way as municipal and industrial (M&I) users do. When I use water to flush the toilet, that water only benefits me and is unavailable to others.**

Farmers, on the other hand, use water to grow products that benefit others. Put differently, we also use that water when we consume agricultural products.

So the right way to calculate "use" is not by looking at how much water one person (or sector) diverts but by looking at each participant's share in the total benefit from that initial diversion.

For instance, say that a farmer diverts water to grow carrots. If he sells those carrots for $0.25/pound to the wholesaler, who sells them for $0.50/pound to the retailer, who sells them for $1.00/pound to consumers who value carrots at $2.00/pound, then we can say that the farmer and wholesaler each get 12.5% of the total value, the retailer gets 25% of the total value, and consumers get 50% of the total value.

So consumers are using 50 percent of "agricultural" water.

So that's the theory. How much water do farmers use in reality?

To find that number, we need to know two things (for each product): What is its value to the final consumer, and who gets paid what along the way?

Although the second part is easier (the USDA tracks this information crop by crop), the first part is much harder. Basically we need to know how much a consumer values an agricultural product. Although we know that they value it at more than the price (because they buy it :), it is hard for us to know how much value they assign above the price. Measuring that "consumer surplus" is more of an art than a science, and it relies on estimations using demand elasticities and/or stated preference surveys. I am now looking for such data. Please email me or comment on where I can find it.

To get back to the second part, consider this USDA statistic: Farmers only keep 19 percent 11 percent of the retail price of food. The average share to each actor in the food supply chain is distributed as follows:

From these statistics, we can see that farmers are using a maximum of 11 percent of 80 percent -- or 9 percent of developed water.

Compare this to M&I's 20 percent share. Are you seeing paradigm shift now?

If consumers enjoy any surplus from agricultural products (i.e., if those products are worth more than they pay for them), then farmers' share of the total is even lower!

Bottom Line: Farmers are not using 80 percent of the water because they are not receiving all the benefits from that water. Use should be attributed in proportion to benefit. Those who receive more benefits are "using" more water -- and that's us, the consumers of agricultural products.
* Developed as in "controlled." If one counts all uses of fresh water, then farmers use about the same amount as the environment (about 40 percent each), and cities use the rest.

** Conventional water accounting defines water as used when it cannot be used by others without processing. Thus, power generators use water by discharging it at temperatures that are too hot for the environment, plants use water when they discharge it through evapotranspiration, and I use water when I flush black water into the sewer system. These discharges can be recovered by cooling, precipitation and cleaning, respectively. (Tomorrow, I'll have more on this.)

A BIG hattip to JR for bringing this idea to my attention.


Ray Walker said...

David, Nice analysis. Another way to analyze is with regard to how USE is defined. Consumptive Use (CU)for your carrot patch is the amount consumed by growing the carrot. It will take a great deal more water to grow the carrot with flood irrigation methods, but much of the water will help recharge the groundwater. Sprikler irrigation of the crop is "more efficient", but the evaporation losses can be high and there is little if any recharge. In dry regions, one man's use is the next man's water right. With flood irrigation, the secondary use would be from groundwater...a change from flood irrigation to sprinkler/drip irrigation often eventually eliminates the groundwater source & use. Domestic use works the same way. Depending upon the sewer treatment, the amount consumed (CU) is only 5% in central sewer systems and 10-15% with septic tanks. Such things as swamp coolers change these numbers dramatically. The irrigation of the lawn can be 100% CU and require 3-5 vertical feet per growing season.
Converting from flood irrigation to sprinkler &/or drip irrigation (best management practices) is why groundwater aquifers are eventually depleted...too many straws and not nearly enough historic recharge.

California will undoubtedly turn to pumping more and more groundwater as their surface water supplies diminish due to drought and curtailment by Court order.

Water in storage also has a CU factor due to 5-6 vertical feet of evaporation per year.

There is a new non-tributary fresh water Source of a million acre feet a year available for CA to develop which will not harm the environment, is legally available and will not damage the water rights of anyone, anywhere. ONE OPTION would be to use the non-tributary water to keep Lake Mead reasonably FULL and producing 2000 megawatts of renewable energy from a facility that is already built and paid for. At the present time, the additional CU from storing an additional million acre feet in Lake Mead would be 20,000 acre feet (2%). There are real water solutions, but somehow the people must get the attention of the bureaucrats and politicians. Ray Walker (Retired Water Rights Analyst) waterrdw@yahoo.com

dWj said...

Actually, my preference is typically to impute all of the inputs to a consumer good to the consumer. In principle, if a consumer good requires the use of some bit of physical capital, it should get a share of the inputs into that capital as well.

This seems to me like the natural epistemological definition of the concept; value ultimately should only be ascribed to consumption, because other activity is only done, in the final analysis, to support that consumption. In practice, you can go about with your input-output tables and try to tease it out, but it's very complicated, especially to allocate fixed costs properly or to deal with good or bad investment. Here's the big thing, though: I don't care.

The great thing about the price system you've been advocating is that the cost of the water gets passed along to the consumer. Fixed costs get dealt with the same way as they do regarding other inputs: entrepreneurs do their best to charge it of whichever of their customers they think are most willing to bear it. The cost of water might make some investments better and others worse, but any economic profits can be straightforwardly attributed to the entrepreneur, with the cost of the water passed along.

In fact, epistemologically again, let me suggest that the "use" of water be attributed as the expense to each party of a marginal rise in the price of water. A producer of low-flow showerheads thereby uses negative water, as his activity becomes more valuable as water prices increase. I would expect production to largely net out to zero, with consumers bearing close to 100%.

gormk said...

David - I disagree. The word "use" and the phrase "deriving benefits from" are two different things. If you want to account resource use this way (by the value value-added approach) for water, so you must for all other resource inputs. At the end of the day, it it the final market value that matters, so you could say that the consumer who's actually eating the carrot, is "using" all the water by your logic.

This raises a problem when you apply the logic to externalties: Is it then the coal plant that is polluting the nearby water-body and the air, or isn't it instead the consumer who enjoys the final commodity (e.g., energy)?

Also, why do you feel compelled to come in defense of farmers? For one, farm-subsidies lead to an overallocation of all resources to food production in the US.

Secondly, grandfathered property rights to water resources have put many farmers in a "slackness" situation, in which too much water is applied too (technologically) inefficiently in food production.

Lastly, water can, and is, reused. The 1 million gallon a Colorado corn-producer diverts for irrigation, seeps back out into ground and surface water. For example, river rafters on the Colorado River still get to enjoy most of the same water for their activities. Ditto for my 3.5 gallon toilet flush... So there are additional accounting issues that complicates the picture.

David Zetland said...

@gormk -- The Saudis don't use all the oil they produce. (We use a lot.) Thus, you can see my logic -- that the "use" of a good be ascribed to the consumer of it. In the case of a carrot, the "use" is divided among beneficiaries....

I am not interested in tracing other inputs, only in attributing the water to those who "use" it.

Your externality example is not exactly on topic, but I'll say that the consumer is responsible. In a capitalist economy, supply does not persist without demand. (I'll concede that advertising can create demand...) Thus, the consumer bears the responsibility for the use of resources (price) and environment (non price).

These are good thoughts, btw....

I am not a defender of farmers as much as a clear debate. As to your comment on grandfathered water rights, I agree. That's why I propose the all in auction.

Your reuse comment is a distraction. We are talking about "economically useful" water. If we discussed consumptive use, the bottom line would be the same...

David Zetland said...

@Ray -- good points that I will repeat tomorrow. Looking forward to hearing where your "non-tributary source" is!

@dWj -- Sound accounting and GOOD point on prices (which I did not emphasize...)

Philip said...

This seems like a sensible way to look at resource use. After all, farmers don't drink all that water, they turn it into stuff people need. And, as Captain Flounder (where is he, anyway?) and others have pointed out, before we greatly reduce the amount of food produced in California, we need to ask ourselves where else that food might be sourced, and what the human and environmental consequences of that might be. Water markets will lead to greater efficiency, more stable farm profits, and a sensible way to prioritize the uses of water we willingly devote to environmental purposes.

Hillary said...

I like this idea, but I'd also like to see another factor: comparing the "optimal" water input for whatever food it is to the actual water input for a given geography. Does the farmer growing alfalfa in southern CA irrigate twice as much as a farmer growing the same product in Iowa? What are the actual transportation costs and opportunity costs of the water he uses?

Philip said...

Hillary, I can address some of your questions. A fair amount of research has gone into measuring the amount of water actually consumed by the various crops we grow. This consumption takes two forms: the water that becomes part of the plant, and the water that is gassed off as part of the plant's respiratory processes. The remainder of applied water either percolates to deeper levels (where it may or may not be able to be re-used), evaporates, or runs off the end of the field(again, this may or may not create waste).
In the San Joaquin, alfalfa is grown on heavy clay soils that have high water holding capacity. The crop has a very deep and extensive root system that enables it to use applied water very efficiently. Yields of 8-10 tons of dry matter per acre are not uncommon.
In the Midwest, the growing season for alfalfa is much shorter, and yields are commonly 2-4 tons. The crop receives fewer irrigations, so it uses less water; but on a harvested biomass per applied acre-foot measure, there is probably little difference between the two areas. The two principal reasons the dairy industry has moved from its traditional base in New England and the Midwest, aside from population growth, is the availability of feed, and energy costs. The cost of keeping dairy cows warm in the wintertime is very substantial in cold climates. I suspect milking your cows in warmer areas, and transporting the milk to the end user is more efficient than keeping cows in cold areas, importing feed, then shipping the milk. Economic reality seems to be proving this out. This brings up the whole idea of whether it is better to devote large acreages to low-input, low yield systems, or smaller acreages to high-input, high-yield ones. It's a debate worth having, and I'll bet the answer is "it depends". Hope this is not too long winded.

Slav Hermanowicz said...

Regardless of the analysis which in principle is right, the farmers physically use water and are in control how much they want/need to use (withing their water rights or allocations). The consumer is using "virtual water" and they have no idea, at least for now, how much that amount is. In theory they could look for products with smaller or larger content of virtual water (like they can look for gas mileage or CO2 emissions - in Europe - when buying a car). However, in practice this is not possible. Hence, water use is rightly allocated to farmers (or power stations or you and me) when they are "primary users".

Jonathan said...

the battle for water in the state of california has the environment and cities getting the majority. i say no worries.. just buy your food from over seas sources and get what all of you deserve. remember chemical tainted dog/cat food, and the baby formula that that killed babies. so while the discussion goes on about water farms are going broke, farm workers are out of work, and eventually food prices will go sky high. needless to say that when these things come to pass the farmers will be blaimed again. Shame on ALL OF YOU.

rdan said...

Way cool David....can I use this!