Trust, but Verify

ShipOfFools_web_07292016“Behold, I am sending you out as sheep in the midst of wolves,
so be wise as serpents and innocent as doves.” (Matt 10:16)

Trust, but Verify [1]

By Stephen W. Hiemstra

My experience as an intern in Western Europe Branch in 1979 led me to aspire to a career as a European analyst. During my graduate school years, this aspiration gave me a reason to invest more in language study than my peers because I saw an immediate application as a Western European analyst, even if it was not my primary motivation for the study. So along the way, Cornell University sent me to study both in Puerto Rico (before my federal experience) and in Germany, and Michigan State University aided me with six weeks of intensive French, courtesy of the U.S. State Department, presumably so that I could begin work also in West Africa. By the time I joined Western Europe Branch as a full-time analyst in 1985, I could read and converse at various levels of proficiency in Spanish, German, and French, which was, in part, why I was recruited to join the Western Europe Branch after about a year with World Trade Branch.

In discussions over my transfer, my focus was to be researching feed manufacturing in Spain as lead investigator for a joint research project between USDA and an experiment station in Zaragoza, Spain. The project with Spain focused on improving our understanding of the prospective entry of Spain into the European Community (EC), which occurred the following year (1986). The project also had a tricky administrative goal of converting the project budget of several hundred thousand dollars from an administrative travel fund into a fund earmarked for research, as promised in the project proposal. As lead investigator, I would have a budget greater than most managers in the building which offered research flexibility, but it also would make the project a prospective “takeover” target for jealous competitors. To prevent any such tampering with the project, I asked to report directly to the branch chief, which would cut out a layer of management, and my request was granted.

Broadly speaking, it was expected that corn import demand in Spain would decline as import levies helped French corn to compete better with U.S. imports—the reasoning was simple, the levies would prevent U.S. corn from being imported until the French corn supplies were exhausted. Domestic barley would also substitute for U.S. imports, but primarily in swine rations, not broiler rations where corn was preferred. Spanish consumers, much like their U.S. counterparts, preferred a yellow chicken—the yellow color came either from corn consumption or from introduction of marigold flower pedals into the rations; chickens fed barley, which offered similar nutrients, turned the chicken meat a sickly bluish-white color. Unfortunately, we had no studies of the Spanish broiler industry from which to assess possible impacts of EC ascension. This deficiency motivated a trip to Spain to confer with my counterpart, Luis, and to see if a broiler study might be undertaken by the Spanish team.[2]

While the Spanish project got started, I continued to publish trade papers based on my statistical work in World Trade Branch. While I was proud to publish a study of EC trade, my methods study attracted more attention, in particular, because the chief economist of a large rice corporation enjoyed the case study of the Thai rice trade that I had used to illustrate my points.[3] That study and my outlook report on the rice trade were apparently unique in giving attention to the rice trade. While other minor crop reports could easily have been written, interest dropped off after the corn, soybean, and wheat markets were reported.

The Spanish project generated an interesting six-week trip in 1985 in which I traveled to Malaga on the Mediterranean coast to attend a conference of the International Association of Agricultural Economists.[4] During this ten-day conference I spent time with many classmates from Göttingen Universität in Germany and made contacts with agricultural economists from all over the world. I mostly remember the awkwardness of seeing former professors eying the topless women on the beach and being targeted by foreign intelligence officers via other women—I never realized how political agricultural trade could be. During the rest of my time in Spain, I needed to check into the embassy—a required courtesy call which mostly annoyed the attaches who were already busy—and to spend time in Zaragoza with my counterpart.

The most productive part of the trip was visiting different feed manufacturing plants to hear first-hand about their procedures and concerns. Their procedures included sample rations, which showed substituting energy (corn, barley, tapioca, etc) and protein (soy meal, sunflower meal, etc) components and proved helpful in a later study of Spanish import demand for corn (1987). Their primary concern back then was the low quality of U.S. corn exports, a problem that was later corrected.[5]

After five weeks in Spain, I traveled to Germany for a week to visit old friends and to confer with them about what I had learned in Spain. Although Germany was also part of my responsibility as a country analyst, the interest in trade with Germany was much less than Spain, in part, because trade with northern Europe and the policy environment were well-established.

Although I expected to return to Spain for a follow up visit, it never happened. My branch chief was diagnosed with lymphoma in the winter of 1986/87 and was quickly unable to function, even though he continued to come into the office.[6] In his absence and after his death, my role as chief investigator came under fire and I suffered fairly arbitrary criticism until I gave up the project. At that point, I requested reassignment to the European situation and outlook unit, but my research responsibilities—just not the project leadership responsibilities—followed me into my new job making the whole arrangement rather stressful.

As a country analysts working in situation and outlook, I had both country (Spain, Germany, and other EC countries) and commodity (cotton and oilseeds) responsibilities. I really enjoyed the outlook work, which included making a quarterly export forecast for roughly 40 commodities, and I began developing a quantitative procedure for estimating exports. I did these estimates with a Lotus spreadsheet macro program which took 40 quarters of export, price, and export sales data and computed three estimates of exports (an elasticity estimate, a linear projection, and a percent change over the previous year) and a graph depicting the forecasts and historical data. I worked with another analyst to write a report outlining the procedure so that the procedure could be used by other country analysts.

Resistance to this export model arose from two quarters. The first point of resistance came from the other country analysts who were primarily former state department analysts with master’s level training, but no quantitative training—at the time, spreadsheets, like Lotus, were new and scary to many people. Most analysts estimated quarterly exports as simply the previous year’s number; only one other colleague routinely used the export sales figures to compute a percent change of the previous year.  Furthermore, employing this model would require that historical data be accumulated and analyzed, which would require time and effort even beyond learning to use the spreadsheets.

The second point of resistance came from policy analysts who had trouble accepting the results from back-testing which showed that the elasticity estimates were the most reliable.  EC imports were not believed to be price sensitive because of the EC variable levies. In fact, the back-testing suggested that a two-step decision process was involved. In the first step, imports were totally restricted until domestic EC production was sold. Then, in the step, imports were purchased from the lowest cost supplier. Hence, price sensitivity in the second step essentially explained the results from the elasticity estimates.

As far as I know, follow up studies of price sensitivity were never completed because later in 1987 senior agency management  announced a reorganization with the stated objective of eliminating the country analyst program. The world of trade was changing as improvements in transportation and communication reduced the need and the growth of large international trading firms reduced the desire for specialized country analysts in the public sector—why listen to a country analyst when you can pick up the telephone or hop on a plane to speak directly with your counterpart elsewhere in the world?

During the reorganization, positions throughout the Economic Research Service were opened up for competitive bidding so I applied for and was granted a transfer out of the International Economics Division and into the Rural Economy Division where I began a new career in finance in Finance and Tax Branch. Finance was entirely new for me so this was a huge move at the time—both professionally and emotionally—because I had spent years preparing for work in European affairs and had almost no training in finance.  Yet, the move into finance proved to be one of the most important career decisions that I ever made. The move led to a series of promotions which made it possible to buy a house and to afford to have my wife, Maryam, stay home to raise our kids.


Hiemstra, Stephen W. 1985. “U.S. Share of World Rice Market Declines,” Rice: Outlook and Situation. U.S. Department of Agriculture (USDA). Economic Research Service (ERS). March.
Hiemstra, Stephen W. 1986. “U.S. Farm Exports to EC Continue Falling,” Foreign Agricultural Trade of the United States. USDA. ERS. November/December.
Hiemstra, Stephen W. and Arthur B. Mackie. 1986. Methods of Reconciling World Trade Statistics. USDA. ERS. Foreign Agricultural Economics Report No. 217. May.
Hiemstra, Stephen W. 1987. The Effect of Spain’s Entry into the EC on the Demand for Imported Corn. USDA. ERS. Staff paper No. AGES870916. October.
Hiemstra, Stephen W., and Stephen MacDonald. 1987. Forecasting U.S. Agricultural Exports Using the Trade Estimates System. USDA. ERS. Manuscript. May.

[1] “Trust but verify” is an expression made famous by President Ronald Reagan who used it to characterize his negotiation strategy with the USSR.

[2] This proved to be an elusive goal because the researchers in Spain were dedicated livestock researchers.

[3] Over the years, the United States and Thailand have competed for the honor of being the world’s largest rice exporters, but rice exports are small compared with corn, soybean, and wheat exports here in the United States.

[4] August 26-September 4, 1985.

[5] Export corn was sold by grade. If the Spanish purchased number 2 corn, then the exporter would purchase U.S. corn (which was typically number 1 quality) and add foreign matter to lower the quality to number 2 grade. This addition of dust and water to the corn lowered the quality and rendered the corn a mess by the time it was imported in Spain. Complaints about such practices to Congress eventually forced changes to the grading standards to remove the incentive to add foreign matter.

[6] When he heard his diagnosis, he knew that he would soon die. During his career in the Air Force, he worked in a nuclear storage facility and was exposed to excessive radiation. All of his colleagues in the facility suffered the same fate. He was close to being eligible to retire and preferred to retire rather than leave government on disability, but he did not live that long.


You may also like

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.