What do cows want?

As many of you know, I've worked a lot on animal welfare issues over the years (e.g., see this book with Bailey Norwood).  One of the biggest challenges is knowing what changes will make a farm animal better off, particularly given the fact changes in housing conditions often improve one dimension of animal welfare while lowering another (e.g., giving animals more space and room might also expose them to aggression from other animals; providing access to outdoors might involve exposure to predators or more variable temperatures).  

It's too bad that animals can't tell you want they want.  Or can they?  Probably because the approach is so similar to the way economists model human behavior, I'm a big fan of a stream of the animal behavior research that looks at what animals choose to infer what they want.  

To illustrate, consider a a variation on the experiment in this 1977 paper by Marian Dawkins.  Imagine an egg laying hen that is given the choice of entering one of two pens with different flooring materials: wire or straw.  

What would the hen choose?  Not surprisingly, hens prefer straw to wire.  So, when presented with this choice, hens will choose to enter the pen with straw flooring.  

But how much do the prefer straw to wire? Well, we can try to make the wire cage a bit more attractive by adding something to it that we know hens like: food.  Now, the hen's choice is as follows:

Now it's not quite as obvious which option is preferable: the wire floor with good tasting food or the straw floor?  Let's suppose our hen still chooses the straw floor.  What now?  Well, we can continue to try to make the wire floor more attractive by adding more food:

We keep adding food until the hen switches from straw to wire, and the amount of extra food required to get the hen to switch from straw to wire is their minimum willingness-to-accept (WTA) compensation for living on a wire floor.  But, the units are in terms of food rather than in dollars.  But this doesn't mean it's not a useful metric.  Once we know the hen's WTA, we can compare the WTA (in units of food) for wire vs. straw floor to WTA (in food) for other things like outdoor access vs. no outdoor access to get a sense of what is most important to the hen.  Heck, because food has an economic value (in dollars) in the human word, we can even convert the hen's WTA to dollars if we really wanted to make comparisons in monetary units.  

Now let me ask a different question: what would you think of a law that required dairy farmers to milk their cows more often?  My presumption would have been that this would reduce milk cows' welfare.  However, this quote in a post from Tyler Cowen at Marginal Revolution suggests cows think otherwise: 

There’s something very interesting in someone else’s vantage point, which might have a truth to it. For instance, the tagging of cows for automatic milking machines, so that the cows can choose when to milk themselves. Cows went from being milked twice a day to being milked three to six times a day, which is great for the farm’s productivity and results in happier cows, but it’s also faintly disquieting that the technology makes clear to us the desires of cows – making them visible in ways they weren’t before.

So the next time you wonder what an animal wants, you might conjure up a creative experiment to let them tell you.

BTW, economists have been studying animal choice behavior for a long time, and it seems animals' behavior is often quite consistent with our "rational" economic models (e.g., see this book by Kagel, Battalio, and Green)