Measurement Error in Agricultural Economics

Agricultural economist is criticized by Steven Payson in his book "Economics, Science and Technology" that, 'price per pound' is a variable with a strong historical precedence in agricultural economics. My first reaction is that Steven should be right about this assertion. I have not read many published papers in agricultural economics and applied economics where sort of measurement unit is used to digitalize (numerically) variables. But my experience tells me that, in general economic theoretical research, economists do not have to specify the unit of a variable after a standardization on everything (or simply with an assumption of unit as 1 at the beginning or by default). In this case, the objective of economic modelling and associated mathematical manipulation is to tell an economic logic or reasoning story (not necessarily an easy to understand answer). As long as the direction of specification (assumed economic relationship) is correct, then there is no need to worry about numbers. The potential issue of measurement error has been taken away in structural research, a skeleton without flesh never needs to worry about the issue of obesity, the probability is no larger than 0.

In agricultural economics, one of the biggest applied area of economics, research is dominated by empirical application of economic analysis. However, often time, the data employed by these research is not collected by researchers themselves (including agricultural economists in different research institutions and graduate students). People who collect the first hand data are usually people who are not supposed to do economic analysis (and they may not be interested in doing so). They are staffs in statistics bureaus, USDA extensions, marketing research companies and so on. They follow the tradition and associate every column of data with an unit as 'per pound', 'per gallon' or 'per acre'. By this way, they also find it is easy to communicate with people who actually work on farm daily. Economists barely bother to do field data collection-the input of research activity. So there is an inconsistency, the academic researchers try very hard to explore and understand the laws and mechanisms behind the economic activities of non-academia. When the data collectors proceed their job, however,  there is a communication need to be considered as I mentioned above. The numbers and measures they are using have to be easy to understand for different people, almost all of them are not economists.

So when people outside of academia criticize the research output of economics, they do not always realize the difference between applied economics and theoretical economics. What usually happens is that, when there is a critique, theoretical economists tend to defend for entire economics. Why? because they are more famous than applied economists, and they won Nobel prizes, so people know them pretty well (at least better than they know about applied economists). When people get questions about their economic life, they come to theoretical economists for answers. But they barely know that relatively precise answer can and should only be provided (more or less, good or bad) by applied economists. And another important fact is that, theoretical economists are good at telling story and making arguments.

This inconsistency and the silence of applied economists make a big distinction between economics and natural science. In physics, for example, people know Bill Gates, Henry Ford, Thomas Edison very well. Why, because they are applied physicists (or engineers). However, when you ask people do you know any theoretical physicists?  I am afraid they can only tell you Einstein, and the reason they know Einstein is not because they want to understand or they have already understood the Theory of Relativity. It is because that Einstein spent almost 1/3 of his life to make himself (or the Theory of Relativity) famous.

Therefore, what an applied economist should do is obvious. It is not trying to adjust or reduce your standard error of estimators, since you never get large enough sample size, and you never know how much is the measurement error involved in your data. What you should do is that, to start answering question from people who do not understand economics and people who are not interested in understanding economics. Do not let theoretical economists take your position and suppress you in a corner; it is too bad that is the case now, but you can change your silent status by speaking out. And do not always ask or even make up a question by yourself, and then answer it by yourself. That is what theoretical economists should do, do not take their position because you would have a more suitable and more pleasant position if you do something different and something more important.

Reference:
[1] "Economics, Science and Technology", by Steven Payson, pp.43, Edward Elgar, Cheltenham, UK.