Friday, October 07, 2005

Timothy Wise: Six Common Errors in Presenting Farm Statistics

In casual chit-chat before the debate last month at the Consumer Federation of America, former Congressman Charlie Stenholm (D-TX) showed me a working paper (.pdf) he liked and was carrying in his brief case, by Timothy Wise from the Global Development and Environment Institute at Tufts University. My graduate students have also been reading it with interest.

The tartly written working paper criticizes statistics from the Environmental Working Group and others, who disparage farm subsidies and emphasize the concentration of farm payments to the largest farmers. Wise's paper is entitled, "Understanding the Farm Problem: Six Common Errors in Presenting Farm Statistics."

For example:
  • One commonly hears that only 40 percent of farms get federal farm payments. Wise argues that, if one excludes "rural residence" farms (including "hobby farms"), then more than half of farms get farm payments.
  • One commonly hears that farm household income is higher than the average U.S. household income. Wise argues that, if one excludes very large commercial farms, the average household income falls below the U.S. average.
  • One commonly hears debates about farm household income that (as in the previous example) count both off-farm and on-farm income. Wise argues that farmers appear much poorer and more deserving of support if one considers their on-farm income alone.
And so forth for the six cases. Wise has harsh words for those who quote the commonly heard statistics: "All of the above statements are true – and truly misleading. The same data present a very different story when treated more carefully."

That's a bit too rough. Like those he criticizes, Wise chooses statistics carefully to make rhetorical points.

In each of the examples above, there are equally legitimate reasons to quote the original statistic. One could equally well argue, in response to Wise's three examples above:
  • It is circular to consider a class of farmers most likely to get farm payments -- by excluding "rural residence farms" from consideration -- and then make a big deal of the fact that many of those farmers do indeed get farm payments.
  • It is circular to consider a class of farmers with modest income -- by excluding the largest and richest farmers -- and then make a big deal of the fact that these farmers do indeed have modest household income. Why exclude the richest farmers from the farm half of the computation, but not exclude the richest non-farmers from the non-farm half of the computation?
  • It would be a recipe for bad public policy to consider on-farm income alone and ignore off-farm income. We might not want to subsidize Hollywood stars with Montana farms, just because those farms grow some wheat. Or, to take a more common case, consider a farmer in any part of rural America who has both on-farm skills and off-farm skills (say, as a mechanic or a school-teacher). If her overall household income is adequate, why should we send a price signal that her farming work is worthwhile, but her work as a mechanic or schoolteacher is not?
One could read Wise's paper and think that the critique of current farm subsidy programs is based on misleading statistics. But, the critique has much merit. The major farm programs encourage over-production of field crops (used largely as inputs to meat and dairy production and, in the case of corn, for high-fructose sweeteners), while offering little support for fruit and vegetable production. They damage the interests of developing country farmers, harm the environment, and largely benefit farmers with household incomes above the U.S. average.

Note to readers: Short break. Will post again on Tuesday.

No comments: