Analysis of U.S. Grain Standards

Chapter 8 Analysis of . -- - — - . - - - —

CONTENTS

Page Inspection and Testing ...... 189 FGIS Inspection ...... 190 Non-FGIS Inspection ...... ,..... 190 Export Inspection ...... 190 Testing Technologies...... 191 Establishing Grain Standards ...... 196 Evaluation of Grain Standards ...... 200 Historical Review ...... 201 Objectives of Grain Standards ...... 201 Integrating the Four Objectives...... 204 Alternatives to the Present System ...... 205 Applying Economic Criteria to Grain Standards ...... 209 Grade-Determining, Non-Grade-Determining, and Official Criteria Factors ...... 210 Establishing Grade Limits ...... 211 Number of Grades ...... 211 Evaluation of Recent Legislation ...... 212 Optimal Grade ...... 212 The Grain Quality Improvement Act of 1986 ...... 212 Prohibitions. Market Incentives ...... 212 Findings and Conclusions ...... 213 Chapter 8 References ...... 215

Box Box Page 8-1. Testing Technologies for Aflatoxin...... 194

Figure Page 8-1. Alternative Authorities to FGIS for Implementing New Tests ...... 197

Tables Table Page 8-1. Wheat Standards ...... 198 8-2. Corn Standards...... 199 8-3. Soybean Standards ...... 199 8-4. Examples of Quality Measures Under USGSA and AMA...... 199 Chapter Analysis of U.S. Grain Standards

The evaluation of the current U.S. inspection dence that it would benefit the industry as a system, possible alternatives, and proposed whole. Measures of oil and protein in the soy- changes in grain standards requires principles bean standard may logically meet opposition on which to base the criteria for change. This from individual firms whose profits depend on chapter sets forth those principles. No attempt being the first to identify sources of soybeans has been made to assess the economic conse- with oil content above the average for that crop quences of the alternatives because they depend year, on the market response to the actions of many individual companies involved in marketing Uniform grain standards provide all buyers grain, Uniform, accurate, and objective qual- and sellers with equal access to information on ity measurement should be based on the logic value. This forces competition on the basis of and consistency of the system, not on the eco- operating efficiency, rather than on control of nomic benefits to individual companies or sec- information, The inability to gain acceptance tors (2). of voluntary standards prior to the 1916 United States Grain Standards Act (USGSA) can be Marketing efficiency requires a system of de- traced to conflict between market opportunity scriptive terms that enable purchase and sale for individual firms through product differen- by description. The sale of millions of bushels tiation and the efficiency of marketing asso- of grain by telephone and telex would not be ciated with product uniformity. Only through possible without the common language on qual- nationally enforced grain standards could in- ity and value provided by U.S. grain standards. dividual firms reap the benefits of industrywide Since most commercial transactions use only market efficiency emanating from uniform one or two grades, the many diverse qualities standards. produced by nature and varying farm practices are combined into a few relatively uniform The 1916 USGSA and subsequent amend- standardized categories during the marketing ments and regulations have established two process. The move toward a uniform product, areas of responsibility for the Federal Grain In- however, conflicts with the profit-maximizing spection Service (FGIS) of the U.S. Department principle of product differentiation, of Agriculture (USDA): private gains from product differentiation are to establish uniform grades and standards, offset by aggregate losses in the efficiency of and market transactions, It is not surprising to find to implement national inspection pro- individual exporters and domestic grain han- cedures to assure accurate and unbiased dlers unwilling to support change despite evi- results.

INSPECTION AND TESTING

Grain can be inspected many times as it tests, such as breakage susceptibility in corn, moves from the farm to its ultimate destination, to laboratory tests like dough handling proper- as demonstrated by figure 2-2 in chapter 2. Nor- ties of flour. mally grain is tested for one or more important characteristics each time it is moved into No single national policy outlines what tests or out of a grain elevator. The number and type will be performed or who will perform them. of tests performed vary from those provided The USGSA requires that grain standards be for in the grain standards to specific end-use developed and used when marketing grain. The

189 standards provide tests covering such items as In addition to developing standards and pro- moisture content, bulk density, and amount of viding inspection services, FGIS: impurities, but do not specify who will perform ● the tests or what tests will be conducted on develops and publishes inspection pro- grain moved within the United States. In fact, cedures, ● two USDA agencies have been authorized by evaluates and approves equipment for use Congress to provide testing services, using the during inspection, grain standards, on grain moving domestically. ● monitors inspection accuracy of FGIS em- Except for protein content in wheat, other tests ployees and licensed inspectors, ● such as for protein/oil quantity and quality and periodically tests sampling and inspection specific end-use tests are performed at the dis- equipment for accuracy, cretion of the ultimate user. The only manda- ● provides appeal inspection, and tory testing is performed by FGIS on export ● responds to complaints regarding service. grain. FGIS also audits its own activities to ensure that In practice, grain traded between two com- service is being provided on a nondiscrimina- panies is normally tested by FGIS or one of its tory basis and that no licensed individual has affiliated agencies, using tests contained in the a conflict of interest. grain standards. However, some domestic processors and nearly all grain companies that buy Non-FGIS Inspection farmer-owned grain purchase it on the basis of tests performed by their own personnel. Inspection services using grain standards These groups, except in cases where a particu- established under USGSA may be performed lar buyer requires additional tests, normally use by grain company employees or by private com- some or all of the tests provided for in the grain panies not affiliated with FGIS. Grain received standards. In other cases, in-house testing is from farmers is seldom graded by FGIS or FGIS- used by grain companies on shipments mov- licensed inspectors but by employees of the ele- ing between their own facilities. vator or processing firm. The standards established under USGSA are FGIS Inspection generally used as the basis for inspection by company employees. FGIS procedures may or The inspection system mandated by USGSA may not be followed, based on individual com- currently consists of FGIS offices with Federal pany policy. Equipment and inspection ac- inspectors located at major ports, 72 designated curacy are not monitored unless the company State and private agencies located in the interior has established an internal monitoring pro- of the United States, and 8 delegated State agen- gram. In many cases, company inspectors com- cies at ports not serviced by FGIS. FGIS admin- pare their test results to those obtained by FGIS isters field offices throughout the country to or FGIS-licensed inspectors on the same grain oversee the activities of State and private in- in an effort to ensure accuracy. Either buyer spection agencies. or seller may request FGIS or an FGIS-licensed All nonfederal employees employed by State inspector to check the grain if the results of the and private agencies authorized to perform in- private inspection are in question. However, spection on behalf of FGIS must pass exami- neither party is required by law to abide by the nations on grading proficiency and must be inspection results. licensed. These individuals can be licensed to inspect one or more of the grains for which Export Inspection standards have been established. In no instance, however, can individuals perform official in- USGSA requires that all grain being exported spections unless they hold a valid license for be inspected by FGIS or a FGIS-delegated State that grain. agency. The only exceptions are for grain mov- . .-

191 ing into Canada and Mexico by land carrier and societies such as the American Oil Chemists’ for small exporters who ship less than 15,000 Society and American Association of Cereal metric tons in a given year. Chemists develop criteria for approving tests, publish performance procedures, and establish Notwithstanding this requirement, importers programs to ensure equipment accuracy. Many often request private companies in the United tests covered by professional societies are not States to represent their interests and inspect in the grain standards. the grain as it is being loaded. Such inspections can include checking for grade as defined by Regardless of which tests are performed and USGSA grain standards or for factors not cov- who performs them, several factors are impor- ered by these standards, such as falling num- tant. These include instrument precision and ber in wheat or oil and protein in soybeans. standardization, calibration, the choice of refer- When private companies perform inspections ence methods and traceability to standard refer- using the grain standards, two groups issue ence methods when developing rapid objective certificates—FGIS and the private company. tests, and natural error resulting from sampling. Samples obtained by private companies are Standardization often submitted to FGIS for analysis and grade, and results from FGIS are then used as the ba- The term standardization means that a meas- sis for the private company’s certification. In uring device has been adjusted to be in fun- other instances, private company inspectors damental agreement with a universally ac- actually perform the inspection. Settlement in cepted standard and that ongoing efforts are most instances is based on the results provided made to keep it in agreement (4). Standardiza- by FGIS or a FGIS-delegated State agency. In tion is vital to fair trade and will be even more rare cases settlement has been based on the pri- important as technologically advanced testing vate inspector’s results or destination grades. equipment is introduced into the marketplace. The validity of a commercial measurement is Testing Technologies judged by comparing it with a more stringent method that is accepted as determining the true Since no single policy exists for inspecting value. The standard is the base method defined grain, no one group is responsible for develop- as being the true value. Working standards are ing and overseeing the tests and equipment devices and methods used to actually validate used. FGIS provides independent, third-party an individual test instrument. For dimensional services using tests contained in the standards. measures such as mass, time, and volume, the Other tests such as for protein content and fall- reference standards are very precise. The pro- ing number tests in wheat, aflatoxin in corn, cedure of matching routine devices against work- and ethylene dibromide residue are also pro- ing standards and working standards against vided, All tests done under the authority of FGIS reference standards introduces little variability. are regulated in that the equipment must be ap- Probably the most visible example of stand- proved, procedures for its use developed, and ardization is the weights and measure program the accuracy of results monitored. coordinated by the National Bureau of Stand- Tests provided under the authority of the ards (recently renamed the National Institute Warehouse Division of the USDA, on the other of Standards and Technology* (NIST)) of the hand, are not regulated to the same degree. No Department of Commerce in conjunction with requirements for type of equipment, procedures the National Conference of Weights and Meas- for its use, or monitoring of equipment accuracy ures. NIST develops specifications for instru- have been developed under this program. In some instances, individual States have de- *The National Bureau of Standards was recently renamed the National Institute of Stand arcls and Technology (NIST) \t.ith the veloped criteria for approving equipment and passage of the Omnibus Trade and Competitiveness Act of 1988 monitor the equipment’s accuracy. Professional (Public Law 100-418) as of August 1988. . . . .

192

ment precision and accuracy along with scale In the case of moisture, choosing a reference tolerances, and maintains national standards. method is more difficult since no one method Scale testing agencies follow NIST procedures is universally accepted. As on electronic mois- in performing periodic testing using field stand- ture meters, FGIS has approved the Motomco ards that can be traced back to the NIST na- brand meter for its testing program and cali- tional standard. In the case of grain measures brates it to the air oven. In States that do not other than weight, no single national organiza- enforce moisture meter accuracy or that allow tion exists for standardizing tests. different sets of calibrations to be used, several types of meters test 1.0 to 1.25 percentage points Measuring grain quality is difficult to stand- higher than the air oven and the Motomco. In ardize because the true answer is not always addition, some meters used on farms are less known, as in the case of characteristics such accurate than those used by industry (5). This as moisture, protein, and oil content. The refer- could result from the fact that not all manufac- ence method is therefore defined rather than turers standardize to the air oven since there proven. Choosing the reference method, however, can be difficult since it can also be as variable as the instruments themselves. For moisture, the standard reference is the air-oven method; for protein, it is the kjedahl procedure. The kjedahl procedure for determining protein is internationally accepted and used. Cur- rently, protein can be determined rapidly by using near-infrared-reflectance analyzers (NIR). These instruments measure reflectance read- ings at various wavelengths. The precision (repeatability) of the kjedahl procedure is+ 0.15 and for NIR it is ± 0.10. Therefore the NIR is more precise than the kjedahl, but after standardizing NIR to the kjedahl procedure, the results obtained with NIR are ± 0.2 to the kjedahl.

Photo credit: OTA Staff Photo The near-infrared-reflectance analyzer (NIR) is the most advanced rapid objective testing technology. It currently is used to measure protein content in wheat. It will soon be used to measure oil and protein content in soybeans and corn protein. - .- . —-

193 is no legal requirement to do so. As all oven uted to the recent controversies over the ac- methods are empirical, relying on weight loss curacy of FGIS wheat protein testing. that is assumed to be water, varying oven pa- The chain between the instrument used in rameters will give varying results. Most inter- the field to the standard reference method is national buyers use an oven method standard- referred to as traceability. The more steps there ized to the Karl Fisher titration method. In the are in the traceability chain, the more chance United States, this is not the case, and some there is for compounding random errors from feel that the United States is underestimating one step to the next. In the case of moisture, corn moistures by as much as 0.7 percent. for example, a standard meter in the main laboratory is standardized to the air oven, stand- Further difficulties arise in standardizing ard meters in the field can then be checked to tests when subjective measurements are in- the standard meter in the main laboratory, and volved. Results for many of the current tests field-standard meters can be used to check in- contained in the grain standards are performed dividual meters. Minimizing the number of by visual and sensory evaluation and rely on comparison steps in the traceability chain may human judgment. FGIS monitors its own ac- or may not maximize accuracy, depending on curacy and has developed visual aids as the ba- the actual size of the random variations. sis for determining many visual tests, such as the degree of damaged kernels. These types of Source of Testing Errors tests are difficult to standardize, and accuracy can vary widely especially when the same tests Since any test result is based on a small sam- are performed without using FGIS visual aids ple that represents the entire population, test or being subjected to FGIS oversight. sample portions are subject to bias and variability, and any test result is really only an esti- mate of the properties of the entire population. Calibration The types of variation can be described as ran- In addition to having standardized tests, the dom and nonrandom. Nonrandom variation equipment used to determine grain quality must occurs from uneven distribution of grain prop- be calibrated to standard reference methods. erties, improper sampling procedures, and in- The calibration must always contain the full accurate measurement. Random variation is range of properties and equipment variations natural and unavoidable, since each grain ker- that will be encountered in general use, so that nel differs from all others. the instrument will not be overly sensitive to If a load of grain is homogeneous, the close- inevitable variations. However, the major cali- ness of the test result to the actual condition bration issue in grain testing is the pervasive is governed by the laws of probability. The sam- variability of these tests. Calibration is further ple size required to produce a result that has complicated by having to use actual grain sam- the desired probability of approaching the ac- ples to calibrate instruments, which introduces tual condition of the grain can be calculated. sampling variation independent of analytical To increase the probability, additional quanti- error. As discussed, both the instruments to be ties of grain must be obtained or the size of the calibrated and standard reference methods are sample actually tested must be increased. For subject to error. Changes in grain properties example, at a 90-percent confidence level a test due to climatic variables complicate the prob- portion size will produce results within a de- lem of obtaining truly representative sample fined range. To narrow the range, a larger por- sets. In addition, as with NIR analyzers, units tion is required or more than one analysis must of the same brand are not identical, which be performed to increase the accuracy of the means that the same calibration constants can- result. not be universally used. It should be noted that NIR calibrations require continuous monitor- When setting portion sizes, the frequency of ing for accuracy. Lack of monitoring contrib- occurrence within a grain mass must also be 194 considered. For example, aflatoxin in corn can corn. Other factors, such as fine material, seg- affect only a few kernels in a grain mass and regate and cause uneven distribution within a in order to detect levels of 20 parts per billion, load. The method and type of sampling is there- which is the limit established by the Food and fore critical to obtaining a truly representative Drug Administration (FDA), 10 pounds of corn sample. Other nonrandom errors involve in- should be examined. This compares to only 2.5 accurate measurements from incorrectly cali- pounds of grain required to determine the brated and maintained instruments, from weight per bushel. (For additional information human error, or from not following correct on aflatoxin testing technologies see box 8-l.) procedures. Uneven distribution in a load of grain is more Knowledge of the source of the variation is of a nonrandom error problem with some char- critical for assessing and improving the ac- acteristics than others. For example, variance curacy of test results. For example, improving in weight per bushel—even though it can fluc- moisture meter precision is an unnecessary ef- tuate within a load—is normally not that great. fort because it contributes less than 10 percent Moisture, on the other hand, can vary due to of the total variability associated with the test mixing or flow characteristics of damp and wet (4). Moisture measurement errors arise mainly

Box 8-1.—Testing Technologies for Aflatoxin Aflatoxin, a known carcinogen, appeared in a large proportion of the corn crop for the first time in many years due to the extremely dry weather conditions in 1988. The principles discussed in this section are very germane to the ability to test for aflatoxin. Aflatoxin is a secondary metabolize produced by the fungus, aspergillus flavus, which infects the corn during field growth. Environmental conditions that favor the production of the mold are high temperatures coupled with dry, drought type conditions during kernel maturation. Aflatoxins are particularly important metabolizes because they are toxic and potent animal carcinogens in excess of certain threshold levels. It is not uncommon for some of the corn crop in the South and Southeast to be infected with aflatoxin at levels that exceed FDA guidelines. Due to the stress this year’s crop underwent, the inci- dence of aflatoxin extended well beyond these regions into the corn belt, especially the Eastern corn belt. At the present time, testing technologies are not adequate. The rapid test used at the country elevator or terminal is not always reliable. And the more reliable tests are not conducive to elevator environments. The most common and rapid test is examination of corn under an ultra-violet light. This is a screening method which does not quantitate the aflatoxin. Contaminated corn will have spots on the kernel that fluoresce a bright greenish-yellow (BGY). But the presence of BGY does not necessarily mean aflatoxin is present. The possibility therefore exists of false positive test results. Corn can also be tested with the Holaday-Velasco minicolumn or thin-layer chromatography methods. The minicolumn test, which takes about 45 minutes, gives indications of whether the corn exceeds the 20 parts per billion guideline established by FDA. Thin-layer chromatography, which takes between 3 and 4 hours, provides quantitative results of aflatoxin levels. The minicolumn and thin-layer chromatography tests are most suited to laboratory environments. Both use chemicals that must be controlled and are not suited to the normal grain elevator environ- ment. Recently, several new technologies, such as methods based on enzyme immunoassay or rocket immune assay techniques, have been developed to detect aflatoxin that require less chemicals and are more suited for use in grain elevators. They also produce results in a more timely manner. These technologies are currently being reviewed by the American Association of Cereal Chemists. As with any method, adequate sampling must be used because aflatoxin is not uniformly distributed among kernels. 195 from differences in electrical properties of grain established testing procedure errors when load- samples, not from the meter’s precision. When ing grain of a desired quality. examining the variability of any test made with an instrument, it is essential to know the rela- Stationary sampling is usually performed tive contribution of instrument error. with a grain probe. Because a probe obtains samples from only one point in the grain mass Sampling at a time, multiple probings are crucial in obtaining representative samples. Probing pat- Grain samples are obtained with either on- terns have been developed to ensure represen- line or stationary methods. On-line sampling tative samples are obtained and to counteract can be done manually, using an Ellis cup or the segregation of fine material in grain at rest a pelican sampler, or mechanically, using a in a carrier. However, probes cannot reach the diverter-type mechanical (D/T) sampler. This bottom of barges or hopper railcars, which af- equipment allows samples to be drawn from fect the representativeness of the entire grain a moving stream of grain being carried on a mass. belt or within a spout, or from a free-falling stream. Samples drawn on-line are generally Probing grain is time-consuming and labor- considered to be most representative since the intensive, and current probing patterns only grain is sampled more frequently and is more obtain about 5 pounds of grain. Mechanical homogeneous in nature than stationary grain. truck probes have been developed and are being used in some locations to reduce the cost Export cargoes must be sampled with a D/T. and labor requirements. But, in many locations, Many barge and railcar shipments are also sam- such as country elevators, only one or two prob- pled with this method even though there is no ings per truckload are taken or a pan full of requirement to do so. D/Ts provide quite large grain is taken as the truck is unloaded. This samples that must be reduced in a secondary compares to the five-to-nine probings required sampler to more workable sizes. The smaller under FGIS procedures. Limited sampling samples are further divided in a laboratory makes the test results more vulnerable to divider to the prescribed test portion sizes. It nonuniformity within the grain mass. is important to recognize that every subdivi- sion as well as the initial sample collection con- As indicated, the sample size used has a direct tributes potential errors. Shippers must there- bearing on the test result’s accuracy. For fore allow for sampling variations as well as aflatoxin, the 10 pounds required to accurately

Photo credit: U.S. Department of Agriculture

The diverter-type mechanical (D/T) sampler allows grain to be drawn from a moving stream. Samples drawn from a D/T sampler are considered to be most representative since grain is sampled more frequently and is more homogeneous than stationary grain. to measure protein content in wheat. This technology has been discussed to some degree throughout this section. Considerable work is being done to develop additional tests with NIR, which will be particularly important at the first point of sale. Calibrations for barley protein are being developed. Work is also being done on developing calibrations for determining soybean oil and protein along with corn protein. The ability of NIR to determine wheat hardness, along with other important tests for wheat, is also being investigated. The first point of sale will probably be the most difficult place to introduce new technol- Photo credit: OTA Staff Photo ogies, The time constraints are severe and the Stationary sampling is performed by a grain probe. resources, both human and capital, are limited. Multiple probings are crucial in obtaining a Yet, the demands of testing at this point should representative sample. In many locations, such as country elevators, only one or two probes are taken be paramount in designing new tests. As these compared to the five to nine probes required are developed and introduced, the impact of under FGIS procedures. the amount of sample required to perform not only the particular test but more importantly detect aflatoxin requires that a truck be sam- all the tests required at the first point of sale pled twice using the current five-to-nine prob- must be evaluated in terms of practicality. ing pattern, which only yields 5 pounds. Get- ting 10 pounds from a D/T sampler is simpler Many of the potential new tests require the since large quantities are obtained through the grain be ground or processed before testing, normal course of sampling. If increased ac- while the tests currently in the grain standards curacy is required, or as additional tests are are performed on the grain as a whole. As more adopted requiring larger sample sizes, the im- tests are introduced that require processing, the pact on the test’s accuracy must be weighed impact of the sample size required to provide against the cost of obtaining the sample. This accurate results versus the quantity of sample will be especially relevant to samples obtained obtained through stationary sampling becomes at the first point of sale from trucks. critical. For example, FGIS introduced a test for sunflower seed oil content, The sample size Criteria for New Technologies required to predict oil content accurately was determined to be 250 grams, which would have As additional tests on an ongoing basis be- required double probing of stationary grain lots come more relevant, criteria must be estab- and consequently increased testing time and lished to govern the design of rapid test require- costs. Thus a trade-off between accuracy and ments. Yet, development of rapid tests must cost became necessary. To overcome this prob- meet the basic criteria associated with stan- lem, a smaller sample size (45 to 50 grams) was dardization, traceability to standard reference established but duplicate tests were required methods, and calibration. In addition, rapid to help minimize the impact of lowered accu- tests must be evaluated in terms of speed, cost, racy due to smaller sample size. accuracy, durability, and capability of handling wide ranges in quality. Establishing Grain Standards The most notable advance in rapid objective Standards are established by FGIS under the testing technology has come from using NIRs authority of Section 4 of USGSA. In the case 197 of corn and wheat, the factors contained in the Figure 8-1.– Alternative Authorities to FGIS for standards were selected in the early 1900s. Soy- Implementing New Tests bean standards were established in 1922 as New test voluntary and brought under the USGSA by congressional amendment in 1941. No changes were made in the number of factors included until moisture was removed as a factor from wheat in 1934 and from corn and soybeans in 1985, The grade limits and factor definitions have been changed frequently during the years, however; tables 8-1,8-2, and 8-3 contain the current standards for wheat, corn, and soybeans. The procedure for changing grades is clearly specified in USGSA. Proposed changes require publication in the Federal Register, solicitation of comments, and, in the case of the new or amended policies, a 1-year waiting period before becoming effective. Grade-determining Non-grade-determining

Introducing new factors requires an under- SOURCE: Office of Technology Assessment, 1988 standing of the alternatives available to FGIS for implementing a new test or quality factor. The agency operates under two authorities— corn is one example. It was an FDA ruling that USGSA and the Agricultural Marketing Act the addition of water to grain to increase its (AMA)–and new tests can be implemented un- weight or value was adulteration and subject der either one (figure 8-l). to prosecution. FDA also regulates color addi- Three methods for implementing tests are tives mixed with grain for identification pur- available under USGSA. The category “official poses and the adhesives and coatings that may criteria” is used for tests provided only at the come in contact with grain transported in rail- request of buyer or seller. Factors contained cars and barges, in the category of “standards” must be determined for each inspection. However, non- The FDCA prohibits food products contain- grade-determining factors are always reported ing whole insects, insect parts, and excreta. Fu- but have no maximum or minimum associated migation or treatment of grains already infested with assigning a numerical grade, whereas does not make grain legal under the act. Chem- grade-determining factors establish a numeri- ical treatment may be used as a preventive to cal grade according to the lowest factor ap- keep grain from becoming infested, but residues proach. Examples of current tests are given in from these chemicals must not exceed permis- table 8-4. sible tolerance levels. Grain is illegal if it contains residues of pesticides not authorized or The Food and Drug Administration also has in excess of safe tolerances set by the Environ- responsibility for grain quality issues as they mental Protection Agency and enforced by relate to health and safety. Under the Federal FDA. Food, Drug and Cosmetic Act (FDCA), grain is deemed to be adulterated if it bears or con- The separation of responsibilities between tains an added or a naturally occurring poison- FGIS, FDA, and other government agencies is ous or deleterious substance that may render somewhat vague in principle, For example, it injurious to health. Aflatoxin-contaminated grain dust is a health and safety issue covered Table 8-1.—Wheat Standards

Minimum limits of— Maximum limits of— Test weight per bushel Hard Red Damaged kernels Spring All other Shrunken d wheat or classes Heat- and Wheat or other classes White Club and damaged Foreign broken Contrasting wheata subclasses kernels Total b material kernels Defects c classes Total e Grade (pounds) (pounds) (percent) (percent) (percent) (percent) (percent) (percent) (percent) U.S. No. 1 ...... 58.0 60.0 0.2 2.0 0.5 3.0 3.0 1.0 3.0 U.S. No. 2...... 57.0 58.0 0.2 4.0 1.0 5.0 5.0 2.0 5.0 U.S. No. 3...... 55.0 56.0 0.5 7.0 2.0 8.0 8.0 3.0 10.0 U.S. No. 4...... 53.0 54.0 1.0 10.0 3.0 12.0 12.0 10.0 10.0 U.S. No. 5...... 50.0 51.0 3.0 15.0 5.0 20.0 20.0 10.0 10.0 U.S. Sample grade: U.S. Sample grade is wheat that: a. Does not meet the requirements for the grades U.S. Nos. 1, 2, 3, 4, or 5; or b. Contains 32 or more insect-damaged kernels per 100 grams of wheat; or c. Contains 8 or more stones or any number of stones which have an aggregate weight in excess of 0.2 percent of the sample weight, 2 or more pieces of glass, 3 or more crotalaria seeds (Crotalaria spp.), 2 or more castor beans (Ricinus communis L.), 4 or more particles of an unknown foreign substance(s) or a commonly recognized harmful or toxic substance(s), 2 or more rodent pellets, bird droppings, or equivalent quantity of other animal filth per 1,000 grams of wheat; or d. Has a musty, sour, or commercially objectionable foreign odor (except smut or garlic odor); or e. Is heating or otherwise of distinctly low quality. ~These requirements also apply when Hard Red Spring or White Club wheat predominate in a sample of Mixed wheat. Includes heatdamaged kernels. cDefect~ include damaged kernels (total) foreign material, and shrunken and broken kernels. The sum Of these three faCtOrS may not exceed the limit for defects ‘Or each ‘Umerical grade ‘Unclassed wheat of any grade may contain not more than 10.0 percent of wheat of other classes. elncludes contrasting classes’

SOURCE: Federal Grain Inspection Service, U.S. Department of Agriculture, 19SS. 199

Table 8-2.-Corn Standards

Maximum limits of— Minimum test weight Damaged kernels Broken corn and per bushel Heat-damaged Total foreign material Grade (pounds) kernels (percent) (percent) (percent) U.S. No. 1 ...... 56.0 0.1 3.0 2.0 U.S. No. 2 ...... 54.0 0.2 5.0 3.0 U.S. No. 3 ...... 52.0 0.5 7.0 4.0 U.S. No. 4 ...... 49.0 10.0 5.0 U.S. No. 5 ...... 46.0 15.0 7.0 U.S. Sample grade: U.S. Sample grade is corn that a. Does not meet the requirements for the grades U.S. Nos. 1, 2, 3, 4, or 5; or b. Contains 8 or more stones which have an aggregate weight in excess of 0.20 percent of the sample weight, 2 or more pieces of glass, 3 or more crotalaria seeds (Crotalaria spp.), 2 or more castor beans (Ricinus communis L.), 4 or more particles of an unknown foreign substance(s) or a commonly recognized harmful or toxic substance(s), 8 or more cock- leburs (Xanthium spp.) or similar seeds singly or in combination, or animal filth in excess of 0.20 percent in 1,000 grams; or c. Has a musty, sour, or commercially objectionable foreign odor; or d. Is heating or otherwise of distinctly low quality. SOURCE Federal Grain Inspection Service, U.S. Department of Agriculture, 1988.

Table 8-3.—Soybean Standards

Maximum limits of— Minimum Damaged kernels test weight Heat Foreign Soybeans of per bushel damaged Total material Splits other colors Grade (pounds) (percent) (percent) (percent) (percent) (percent) U.S. No. 1 ...... 56.0 0.2 2.0 1.0 10.0 1.0 U.S. No. 2 ...... 54.0 0.5 3.0 2.0 20.0 2.0 U.S. No. 3a...... 52.0 1.0 5.0 3.0 30.0 5.0 U.S. No. 4b ...... 49.0 3.0 8.0 5.0 40.0 10.0 U.S. Sample grade: U.S. Sample grade is soybeans that: a. Do not meet the requirements for U.S. No. 1, 2, 3, 4; or b. Contain 8 or more stones which have an aggregate weight in excess of 0.2 percent of the sample weight, 2 or more pieces of glass, 3 or more crotalaria seeds (Crotalaria spp.), 2 or more castor beans (Ricinus communis L.) 4 or more particles of an unknown foreign substance(s) or a commonly recognized harmful or toxic substance(s) 10 or more rodent pellets, bird droppings, or equivalent quantity of other animal filth per 1,000 grams of soybeans; or c. Have a musty, sour, or commercially objectionable foreign odor (except garlic odor); or d. Are heating or otherwise of distinctly low quality. ‘Soybeans that are purple mottled or stained are graded not higher than U.S. No. 3. !Soybeans that are materially weathered are graded not higher than U.S. No. 4. SOURCE: Federal Grain InspectIon Service, U.S. Department of Agriculture, 1988.

Table 8-4.—Examples of Quality Measures by the Occupational Safety and Health Admin- Under USGSA and AMA istration (OSHA) but prohibiting its reintroduc- Authority Example factors tion into grain falls to FGIS. A Memorandum USGSA: of Understanding between FGIS and FDA has Grade-determining standards . . Foreign material enabled the two agencies to work together in Damage several areas. FDA has established action limits Test weight Non-grade standards...... Moisture for many factors, such as aflatoxin in corn and Official criteria...... Protein in wheat pesticide residues in grain. The agreement be- AMA: tween agencies requires that FGIS report to Official criteria...... Aflatoxin FDA whenever inspection results for the items Falling number Pesticide residue (EDB) identified in the agreement exceed FDA’s SOURCE: Office of Technology Assessment, 1989. limits. 200

EVALUATION OF GRAIN STANDARDS

Recent debates over grain standards, foreign recognized that the industry as a whole would buyers’ complaints, and the numerous propos- benefit. als for regulatory and legislative changes echo earlier calls for change. Recent complaints con- After many years of educational efforts and cerning certificate final contracts, excess for- attempts to obtain support, the National Grain eign material at destination, spoilage, and heat- and Feed Dealers Association agreed to endorse ing during transit are nearly identical to those a compromise bill for Federal standards. The of the late 1800s. (Certificate final in the ex- compromise garnered reluctant support from port/import contract means that grade is de- the two opposing positions of voluntary adop- termined at the point of loading and the buyer tion of inspection and grain standards versus has no legal recourse regardless of delivered full Federal control. The compromise allowed quality, unless it can be proved the origin grade Federal supervision of uniform standards and was incorrect at the time of loading.) inspection by private firms, many of which were already in business in 1916. These problems have been documented re- peatedly in congressional hearings in 1908, More than 150 bills and amendments have 1928, 1975, and 1986 (3). Even the wording is been submitted to Congress since Senator Pad- almost identical, despite a span of some 75 dock first proposed Federal legislation in 1890 years. The Senate Report 988, of July 27, 1912, to establish measures of quality that would be criticized the current system for rewarding uniformly and objectively applied. Yet the basic adulteration: premises and procedures of the Grain Stand- ards Act of 1916 remained intact until 1986. Under the present conditions an enormous Further, the complaints and persistent prob- system of mixing or adulteration of grain is forced on all the home markets and also the for- lems regarding quality have continued for over eign market, destroying all confidence in our a century despite all the legislative, adminis- grades and working to the detriment of the trative, academic, and industry efforts spent grain trade. on improving quality measurement. Several ob- stacles to a permanent solution must be recog- In 1986, Representative Byron Dorgan (D-ND) nized and dealt with if current efforts of Con- used similar phrases: gress and industry are to succeed:

In short, the U.S. system being what it is al- ● lows and encourages dockage to be added back The changing nature of the industry in into wheat shipments . . . unless we confront terms of technology and genetics results this problem, we will further erode our export in quality characteristics that change from potential, one crop year to the next and from one part of the country to the next. All these diverse Historical Review qualities enter the marketing channels and must be handled. The legislation establishing Federal grain ● Industry and government have often re- standards was passed in 1916 following more fused to accept that problems exist. Faith than 50 years of debate and repeated attempts in the infallibility of a market system has by the grain industry, trade associations, and made it difficult to accept that government boards of trade to establish private and regional agencies have a function in setting uniform measures of grain quality. Private firms can in- grain standards in order to facilitate an ef- crease profits by individually differentiating ficient market. quality from that of other firms. As noted ● Most of the participants in the debate have earlier, this self-interest blocked every attempt seen themselves as adversaries since the at obtaining voluntary adoption of uniform beginning of the discussions. Each group– standards even though the trade associations farmers, processors, grain merchandisers, 201

and government—has expended consider- trary inconsistencies. For example, test weight able public effort trying to blame one or limits in corn for grades No. 1 and No. 2 were more of the other groups. in the original standards in 1916; were added The USGSA and the regulatory agencies to grade Nos. 3, 4, and 5 in 1918; were lowered implementing the act have focused on one in 1934; and were raised in 1959. Throughout objective— “to facilitate marketing.” Eco- all these changes, researchers questioned nomic principles have not been incorpo- whether test weight was a relevant measure of rated into the grade factors, factor limits, value in any grade. Those who argue that the or the many seemingly arbitrary changes. only objectives of standards is to describe physical and biological characteristics have been un- Objectives of Grain Standards able to find a criterion by which they could justify the characteristics chosen. For example, Since the beginning of Federal regulation in why measure kernel density but not kernel size? 1916, the official purpose has been to provide Why measure the percent of split and broken uniform standards for promoting and protect- beans but not the percent of whole kernels in ing grain moving in interstate and foreign corn? Grain has many physical and biological commerce “so that grain can be marketed in properties, and without additional criteria for an orderly and timely manner and that trading guidance, the factors and factor limits become in grain maybe facilitated. ” It has seldom been arbitrary numbers. recognized that the lack of an economic justification in that directive has caused much of the In practice, few markets quote prices for No. difficulty in arriving at a consensus on grain 1, 2, and 3 corn, soybeans, or wheat. In nearly standards and standardization within the grain all cases, a base price is given for one grade industry. Although early researchers frequently and discounts or premiums attached to devia- referred to “intrinsic value” and “value in end tions on each factor from that base. At the coun- use,” these criteria were seldom in evidence try elevator level, for example, nearly all corn when developing standards, or in the numer- is purchased on a No. 2 base and farmers’ prices ous changes that were introduced in subsequent are established by discounting each factor that years. falls outside of the No. 2 limits. International markets generally specify No. 2 soybeans or No. Much of the disagreement over specific 3 corn, and prices in the contract are usually changes is the result of explicit or implicit dis- quoted for that one grade with adjustments for agreement over the objectives of standards. deviations. Each group held a different view. Many in industry and government viewed standards as a A set of clearly stated objectives for stand- convenience for merchandisers. Processors ards based on sound economic principles was wanted the standards to indicate yield of proc- absent in legislative and administrative changes essed products. Farmers wanted assurance that between 1916 and 1986. The many changes dur- discounts associated with grade factors were ing that period did little to resolve the basic based on differences in real value in use. Con- problems and issues. In 1986 a relatively sim- gress often viewed the standards as a cause of ple change in the language of the USGSA opened foreign complaints and lost export markets. the door to a new era in the identification of Clearly, if grain exporters, grain processors, quality as a means of efficient communication Congress, and farmers hold divergent views on of information about value. what grain standards are intended to accom- In June the North American Export plish, they cannot be expected to agree on 1986, which grade factors would meet their diverse Grain Association submitted a report emanating from a series of industrywide workshops. and conflicting purposes. This report, entitled “Commitment to Quality,” The lack of clear goals, objectives, and cri- presented a consensus to serve as guidelines teria inevitably led to reversals as well as arbi- for Congress and FGIS in revising standards 202

(6). One of its most important aspects was a new The technology and terminology for meas- definition for the objectives of grain standards. uring current factors have been in use for The four objectives were identified as: enough time to make the trading of grain on these factors and grades limits extremely effi- 1. to define uniform and accepted descrip- cient. Domestic and international contracts tive terms to facilitate trade, have been written using this terminology for 2. to provide information to aid in determin- many years. Trading in the marketplace is read- ing grain storability, ily handled with a minimum of specification. 3. to offer end-users the best possible infor- These terms are adequate for basic contracts mation from which to determine end prod- and quotes in the futures market as well as in uct yield and quality, and international trade. Basing price on the numer- 4. to create the tools for the market to estab- ical grade, millions of bushels change owner- lish incentives for quality improvement. ship with a simple phone call. These objectives were incorporated in the 1986 Grain Quality Improvement Act with the Aid in Determining Grain Storability passage of Public Law on November 99-641 10, The factors that influence storability of grain 1986. FGIS for the first time had a set of criteria on which to base changes in standards and have been well documented by many scientists, to evaluate the numerous proposals for change. dating as far back as the grain storage studies Each criterion provides a basis for assessing done in the early 1900s by Dr. Duvel of USDA. current standards and the recent efforts by These factors are moisture, temperature, air FGIS to improve measurements of quality. flow, mold, and insect infestation. The more technical attributes of these characteristics are covered elsewhere in this assessment. The Facilitating Trade length of time grain has been held in storage and the condition under which it has been Almost any set of factors can meet this cri- stored are also important criteria. teria. It is important that grain be grouped into a relatively few number of categories to make Current tests for factors contained in stand- buying, selling, and classification efficient and ards provide little information on direct meas- inexpensive. Trade is facilitated by having a ures of storability. Moisture content is provided small number of grades determined by a mini- in terms of averages, but it has been demon- mum number of factors. From the standpoint strated that the range of moisture among indi- of trading simplicity, three numerical grades vidual kernels may be more important than the in the corn standards may in fact be more average. The standards identify damage, but desirable than five. this is an arbitrary determination of a stage of The factors in the current grades are for the development and storage deterioration that most part easily measured and provide a basis does not provide information on how much on which buyers and sellers can communicate longer the grain may be stored before it goes price and appropriate discounts. Some have the out of condition. Numerical grades alone cer- additional advantage of being commonly used tainly provide no information on storability. in international trade and thereby serve to fa- Foreign material, test weight, damage, and cilitate communication within the international moisture are thus indirect indicators at best as well as domestic market, even though they are reported on every inspection certificate. Since almost any set of factors and grade limits can be handled equally well by the grain Changes in inspection practices and stand- industry as long as they are universally ac- ards have done little to improve measurement cepted, recent FGIS changes have made little of storability since the 1920s, when Dr. Duvel improvement on this criterion. suggested acidity as a measure of storage life 203 and an indication of mixtures of old and new from corn standard factors is often unrelated crop grain. The 1986 changes in the interpretive to its feeding value or starch yield. In general, line slides used to identify damage in soybeans those characteristics of raw grain most closely did not improve the ability of the standards to associated with value of products derived are predict storability. However, the correlation be- not currently covered by the standards. Purity tween damaged kernels and levels of free fatty and cleanliness in terms of foreign material, acid better equipped standards to indicate oil numbers of insects, or other grains provide one quality. of the few indications of value in the current standards. Even here, however, there is lack Most current measures for storage life are of clarity; the term “foreign material” means based on laboratory procedures. No rapid com- different things in each grain, and the term mercial test is available to provide a quick in- “broken corn and foreign material” (BCFM) dication of the stage of deterioration or the time does not differentiate between broken corn and remaining before grain’s condition worsens. foreign material despite the difference in value.

Measuring Value in Processing FGIS in the last few years has moved toward Products improving the measurement of end-use value. For example, even the simple step of rounding Because most grains are used for more than dockage percentage to the nearest one-tenth in one purpose, it is not a simple matter to iden- wheat gives a better indication of the amount tify the characteristics that influence value (see of grain versus nongrain being purchased. Com- ch. 4). The nutritional composition, starch con- pared with the previous method of rounding tent, and recovery rate of corn and baking char- down to the nearest half-percent, the new ap- acteristics of wheat are all end-use properties proach better reflects true value. FGIS has also desired by processors. Few of these, however, taken steps to differentiate more critically on can be converted directly to measures of phys- damage in soybeans. The change in the inter- ical or intrinsic properties of the raw grain. Re- pretive line slides has been linked to the level search in recent years has identified some fac- of free fatty acid, which is in turn a direct meas- tors that relate to value that can be measured ure of the quality of oil derived from the soy- in the commercial market channel. For exam- beans. Progress is therefore being made toward ple, hard endosperm corn provides a higher finding better measures of value. yield of flaking grits. Obviously, high oil and high protein soybeans provide higher yields of Additional measures are available that have oil and soybean meal. Breakage susceptibility not been incorporated in the standards. There tests identify the ability of corn to withstand is a lack of total industry agreement that these handling without increasing breakage. Many measures are sufficiently accurate and relia- tests—such as falling number, farinograph, ble to be introduced. Continued commerciali- hardness, and baking tests–are available that zation of measures of breakage susceptibility indicate baking and milling characteristics of in corn and soybeans; intrinsic properties of wheat and are frequently used in laboratories corn, soybeans, and wheat; new measures of in the United States and Europe. But the in- baking properties and classification of wheat dustry does not completely agree that these at- varieties; and rapid measures of oil and pro- tributes always clearly indicate value. tein in soybeans are all candidates for inclusion in the standards or should at least be made Factors currently contained in the standards available as information. do a poor job of meeting the criterion of end- use value. Soybean standards include little in- Providing Market Incentives dication concerning oil and protein content. New varieties of wheat have diminished the ef- Measurement of factors that result in price fectiveness of class and grade for evaluating differentials in the market provide an incen- flour and baking characteristics. Information tive for each point in the market channel to 204 make decisions to improve quality and avoid processor. Yield becomes the primary and in discounts. This incentive works back through most cases the only criterion on which to select the market channel to producers, who may the variety to be planted, A second example is change harvesting, drying, and storing prac- the drying temperature of corn. Although most tices (see ch. 7) or may select different varie- corn processors object to the use of high tem- ties (see ch. 6). Preference for varieties with cer- peratures during drying, the market does not tain quality characteristics will be conveyed to differentiate between corn dried at high and plant breeders who will in turn generate bet- low temperatures. The premiums paid by a few ter varieties. In a competitive market, profit is dry millers for low-temperature-dried corn in- a strong driving force in any company’s deci- dicate that farmers do respond to these incen- sions, whether they be exporters, individual ele- tives when offered. At present the incentives vator managers, producers, or plant breeders. are not offered by means of the standards, but While the market itself sets incentives in terms by means of contracts in localized areas. It is of prices and price differentials, the standards especially difficult for buyers some distance are not neutral in this scenario. For example, from the production point to obtain qualities farmers have an incentive to select corn varie- they desire when those qualities are not incor- ties that weigh at least 54 pounds per bushel porated in uniform standards and terminology. in order to avoid discounts in the market. This producer incentive translates into incentives Recent FGIS efforts have had limited effect for plant breeders, who have spent significant upon incentives within the market channel. The research funds to develop varieties that will removal of moisture as a grading factor reduced produce a high test-weight corn under normal the number of factors on which blending was required when purchasing on grade only, but conditions. the market still generates income from blend- Current grain standards provide incentives ing wet and dry corn. The change in rounding in several ways. For example, the allowance procedures for dockage percentages in wheat of 3-percent BCFM in No. 2 corn, in conjunc- removed the incentive to blend dockage just be- tion with the market practice of paying top price low the next higher break point. However, the for No. 2 corn, gives farmers an incentive to number of grades and the steps between grades incorporate 3 percent BCFM in their deliver- have not been significantly altered, and incen- ies. By the same token, elevator managers have tives and disincentives still fall short of the an incentive to clean or blend in such a way ideal. as to deliver 3 percent BCFM throughout the Measurement and sampling technology or market channel. Any shipment containing less testing is a problem in terms of incentives only than 3 percent BCFM is a lost profit opportu- so far as the accuracy of the equipment will nity. Shipments containing more than 3 per- not accommodate a finer distinction between cent BCFM will usually receive a discount. The qualities on certain characteristics. Increased step functions between the grades create incen- accuracy will permit narrowing the spread be- tives for blending when grain is purchased on tween numerical grades, thereby reducing the grade alone. The wider the range between incentives for blending when purchasing on grades, the greater the number of factors, and grade only. the greater the number of grades, the greater is the opportunity for blending when purchasing on grade without specifying factor limits. integrating the Four Objectives The other side of this coin is the lack of in- Changes in grade-determining factors, or fac- centive for improving quality on those charac- tor limits, should meet the four objectives of teristics omitted from the standards. With no grain standards. But, not all alternatives will price differential for soybeans with high oil and contribute to all four objectives, and it is likely protein, farmers have no incentive to select vari- that conflict between the objectives will de- eties that will represent greater value to the velop. For example, complex and lengthy tests 205 for measuring end-use value will slow the in- sis for any dispute or arbitration regarding qual- spection process, increase marketing costs, and ity. This certificate final system, which was in thereby detract from the purpose “to facilitate use prior to the USGSA, provides advantages trade. ” Similarly, incentives are best created in terms of efficiency and costs. Other alterna- by a continuous discount schedule on each fac- tives are currently used in other countries (see tor, starting from zero. But a zero base requires ch. 10 and companion report) and variations each buyer to specify factor levels in the con- of the official U.S. system are being used even tract. Numerous contract specifications in- in the domestic market. Understanding the crease the number of segregated lots and re- basics of each alternative can help develop pol- duces the interchangeability of shipments, icy for directing future quality regulations, Increasing the amount of information available to users increases the difficulty of merchandis- Total Defects ing uniform lots. Under the lowest factor approach, all factors The “perfect standard” that optimizes all four can be at the maximum limit for the determi- objectives may not exist. The final solution will nation of a particular grade, For example, No. be a compromise among some of the objectives 2 corn can have 54 pounds test weight, 3 per- (e.g., sacrificing complete information for all cent BCFM, 4 percent damage, and 0.2 percent users in order to facilitate orderly efficient mar- heat damage. This provides the opportunity to keting) and among various buyers and sellers. blend on one or more factors to achieve the maximum allowable limit within that grade. A Alternatives to the Present System system of “total defects” introduces one more limiting factor that sums the defects in that An evaluation of the alternatives to the cur- grade across all factors and places a maximum rent system in the United States illustrates some that would be more restrictive. Thus the accu- of the trade-offs inevitable in establishing grain mulated value of the defects would become standards. The system today is generally known more restrictive than the individual factors, In as “numerical grades determined by the lowest factor.” the example just given, the defects are BCFM, total damage, and heat damage, Under the to- The “lowest factor” approach requires that tal defects system, a maximum value of 6 per- the grade be set by the factor representing the cent might be established, for example, and the lowest quality. For example, if the test weight corn would be considered No. 3 if the sum of of corn is 53 pounds per bushel, that sample damage, heat damage, and BCFM exceeded 6 is graded No. 3 even though all other factors percent. Currently, the wheat standard includes are equal to No. 2 or better. This method has a total defects factor with a maximum limit of the advantage of simplicity and is used by many 3 percent for No. 1. If shrunken and broken ker- of the major grain-exporting countries. How- nels, for instance, are at the maximum No. 1 ever, it does not always reflect true value since limit, which is 3 percent, all other defects would it fails to consider factors above the minimum. have to be at O. This approach provides addi- Considerable variation in quality can occur be- tional information about quality and an incen- tween shipments without a grade change. The tive to increase quality if the total defects fig- lowest factor system also encourages blending ure is more restrictive. to bring all factors down to the quality of the lowest factor determining the grade, and there- The total defects approach differs from the fore does not generate incentives for quality im- lowest factor only by accumulating the sum of individual factor results. The greater restric- provement unless limits are established by con- tion it entails could also be achieved by chang- tract that are more restrictive than the grade limit. ing limits on the present grade factors. Total defects also contains a logical inconsistency: In export trade, most contracts specify that It establishes limits on three factors for No. 2 the grade determined at origin is the legal ba- wheat but if the sample meets all these criteria, it will be classed as No. 4 on the basis of total Contract Specifications defects. The issue of whether standards are required or whether each buyer may simply specify Absence of Imperfections terms in their contract has been debated since The “total perfect kernels” approach also the idea of numerical grades was first presented sums the percent of defects, but subtracts the to merchandisers in the late 1800s. So long as total from 100 to obtain the percent of perfect contracts are legally enforced, buyers and kernels. Each grade has a minimum limit on sellers may agree to any set of factors, charac- percent of perfect kernels. Corn grades in China teristics, and conditions of shipment they de- are based on this system, and U.S. standards sire. The advantage of numerical grades over for oats include a factor for “sound oats.” This contract specifications is that they facilitate system differs little from the “total defects” con- communication. When the grain trade became cept except that it reports absence rather than too large for buyers and sellers to physically presence of defects. and simultaneously view the grain being sold, terminology was required that would permit Weighted Factor description by factors. The numerical grade was chosen as a way of describing several fac- The “weighted factor” approach would in- tors in one number. Each buyer could identify clude a set of grade-determining factors and the characteristics and factor limits that best factor limits, but each factor would exert a meet the conditions in a particular plant. If the different effect in terms of grade determination. buyer’s needs are unique, however, resale in This is in contrast to the lowest factor approach, the market becomes difficult and segregation in which the grade is lowered when any one is required for each one’s specifications. The factor result exceeds the limit. Instead, defects efficiency of the current marketing system re- would be divided into major and minor cate- lies heavily on uniformity for the majority of gories and each assigned a weight to be multi- the crop being marketed. plied times the percent of that defect present. The sum of the defects times the weighting fac- Although most foreign buyers now use nu- tor would determine the grade. Grade would merical grades for purchasing corn, soybeans, be influenced by the number of factors below and wheat, they almost always include addi- grade, the distance each of those factors fell tional factors besides the numerical grade in below the grade limit, and the relative serious- their contracts. Numerical grade alone seldom ness of the defect. conveys sufficient information to satisfy a buyer’s needs. Other factors are therefore speci- This system would incorporate more infor- fied. It is a small step to go to an entire factor mation in the numerical grade and allow finer basis system and eliminate numerical grades. distinction among the combinations of factors. The primary disadvantages are a much wider Its disadvantage is its complexity, the arbitrar- range of quality characteristics between ves- iness of the weighting factor assigned to each sels and thus difficulty in resale if other buyers defect, and the lack of clear criteria on which do not want the same specifications, greater to set the break points between grades. In or- difficulty in segregation and blending to a wider der for the numerical grade to indicate end-use range of specifications, and increased complex- value, the weighting factors would have to dif- ity in writing contracts. fer among uses. The complexity of such a system would probably eliminate this alternative None of these disadvantages is insurmount- on the criteria of the first objective of stand- able. The problem of reselling uniform lots has ards—” to facilitate trade,” become much less important in recent years. -- . . . —

207

The number of times a vessel is resold after leav- are based on delivered quality. The standard ing port has fallen compared to the 1970s sce- varies between months, countries of origin, and nario. The complexity of identification is also crop years. The quality characteristics are few less daunting since most buyers are already in number and consist primarily of visual ob- specifying grade plus two or three factors; they servation. would be specifying only one or two more. The advantage of FAQ is its simplicity and There would be greater problems in keeping its flexibility for adapting to changing crop qualities segregated throughout the market years. This advantage to the seller is, of course, channel if there were no numerical grades. a disadvantage to the buyer. The contract does One of the purposes of blending is to achieve not cover all factors on which buyers might like a uniform product, and the first objective of information, and the floating standard leaves standards is to facilitate merchandising. This the buyer uncertain as to what quality may be purpose is not accommodated by a system received for processing. It is often a destina- where all buyers develop their own set of qual- tion contract (advantage to buyers, disadvan- ity factors, their own definitions, and their own tage to sellers) and covers factors not likely to limits. The complexity and inefficiencies of change in transit. Since half the vessels are such a market would result in a major reduc- mathematically below the FAQ standard, it is tion in welfare for both buyers and sellers conceivable that many contracts would require throughout the market chain. Numerical grades court arbitration. In fact, the system is man- therefore are a means of creating uniformity ageable because FAQ factors and sampling of quality in the market channel. They are most methods are not sufficiently specific to support useful if they reflect the needs of the majority arbitration action except in cases of extreme of buyers in the market so that additional con- deviation from the standard. tract quality specifications are the exception The FAQ system provides no incentive for rather than rule. improving quality since it only describes what- ever quality is produced. The extent to which Fair Average Quality the FAQ system describes value to users de- The contract for fair average quality (FAQ) pends on the factors included. In the case of specifies that the grain shall be equal to the aver- the GAFTA corn contracts, the factors describe age quality of grain exported or imported at a primarily condition of the grain rather than in- specific location. The most common FAQ con- trinsic value characteristics. Some of the soy- tract in previous years has been the GAFTA bean contracts include oil and protein. The sim- (Grain and Feed Trade Association) of London, plicity of FAQ facilitates trade but the potential which specifies that quality delivered will be for disputes and arbitration can reduce mar- equal to the average of the quality from the keting efficiency. country of export to the importing country for Variations on the FAQ system include con- the month in which the delivery was made. tracts that specify origin instead of destination. Samples are taken from each vessel and stored Canada and Australia have sometimes gener- by GAFTA in their London laboratories. A com- ated a fair average quality based on harvested posite sample at the end of each month is cre- quality for that crop. The FAQ standard is usu- ated as the standard, and shippers who think ally established with respect to selected char- that the quality of their shipments did not meet acteristics of special importance. FAQ submit samples to be judged against the GAFTA standard. Arbitration decisions are In recent years the FAQ contract has been based primarily on visual inspection by an ar- less frequently used. Argentina, South Africa, bitration committee. Most GAFTA FAQ contracts and Brazil all report exporting primarily on nu- 208 merical grades. However, European buyers through warm water, how can shippers guar- have reported continued use of destination antee quality at destination? FAQ on soybeans. The buyer has always had the alternative of using a contract that specifies destination qual- Destination Grades v. Certificate Final ity. Few shippers are willing to take that type of risk at a price the buyer would be willing Most contracts originating in the United to pay. With no control over unloading or sam- States specify that grade is final at origin. This pling, shippers are in a poor position to guar- means that the grade certificate, issued imme- antee destination quality and the number of diately after the vessel is loaded, is the final doc- contracts taken into arbitration court would un- ument on quality and the buyer has no recourse doubtedly rise dramatically, resulting in a sig- beyond proving incorrect inspection on the ori- nificant increase in cost and delays in set- gin samples. Foreign buyers must prove that tlement. the vessels were improperly loaded or inspected at origin before they can claim restitution for Origin and destination contracts can be used quality. Grain received in poor condition, badly to guarantee quality under any set of standards. broken, spoiled, sprouted, or insect-infested is Yet, the two systems incorporate different in- not sufficient evidence on which to base claims centives. Destination guarantees place addi- for damage because the contract specifies ori- tional responsibility, and thus economic incen- gin certificate final. tives, on the exporter to load grain that will maintain quality during transit as well as meet Certificate final is a highly efficient market- the contract at time of loading. This alters the ing technique, and enables minimum cost qual- loading strategy. Under origin grades, for ex- ity guarantees up to the point where the example, a 14. O-percent moisture contract could porter transfers responsibility for quality to the be met by blending 8 and 16 percent moisture captain of the vessel or to the importer. It has corn. Under destination quality guarantees, this the disadvantage of the ultimate customers’ dis- blending would not be a good strategy under satisfaction with the delivered quality and their most time and temperature conditions. The inability to control quality when several han- high-moisture corn would probably result in dlers stand between shipper and user. It also damage levels and spoilage at destination ex- enables the shipper to load closer to the qual- ceeding the contract limit. ity limits without regard to the inevitable con- sequences of placing that quality in the vessel. The issue of origin and destination grades must also be considered in the context of do- The alternative is destination grades (often mestic markets. The same issues exist as in the in conjunction with a cargo, insurance, and export market, but the results are different for freight sale). In this case, the seller guarantees two reasons: delivery to the foreign port and guarantees quality at that destination. This alternative has often 1. the time between origin and destination is been suggested as a solution to the problem of usually much shorter in the domestic mar- foreign complaints. But many unanswered ket; and questions remain about such contracts. For ex- 2. sampling and inspection procedures at ori- ample, who will take the destination samples gin and destination are subject to a single and how? Second, what type of guarantee can set of regulations in the domestic market, shippers make for a vessel that will be distrib- whereas FGIS has no control over sam- uted among 10, 20, or even 50 different buyers, pling and grading at foreign destinations. each one getting only a small portion from a vessel with a highly variable quality among Domestic contracts specify origin or desti- holds? Third, since the buyer specifies the qual- nation grades as well as whether settlement will ity characteristics and frequently requests a be based on official or private inspection re- moisture content unsafe for long voyages sults. Confidence in accuracy is developed with 209 a large number of transactions over a period especially with respect to sanitary quality fac- of time and by the option of calling for FGIS tors, and could include more information on inspection and an appeal if needed. This con- value in processing. fidence often results in acceptance of non-FGIS Separate standards for different users could origin inspection where direct contact between provide more information on value. The food buyer and seller has generated mutual trust. grade or export grade could have price differen- Trade rules and arbitration procedures estab- tials that would generate market incentives for lished by the National Grain and Feed Asso- improved quality. The market would direct the ciation also provide an important alternative higher qualities into the higher valued uses. The and supplement for domestic trade. dual system would create a more complex marketing system and would probably increase the Dual Grades cost of segregation and transport. Probably the A dual grading system could be based on sep- greatest difficulty would be determining which arating domestic and foreign markets or sepa- standard and discount to apply to the producer. rating use, Separate export standards have often Since ultimate use would not be determined been suggested as a way to compete in inter- at the time of farmer delivery and, in fact, in- national trade with more restrictive quality tended use might be changed more than once specifications. The Canadian system establishes in the market channel, the higher discounts on separate grade requirements for export wheat food grade or export grade would have to be and controls purchases, movement, and clean- applied against the producer. This would gen- ing procedures between country elevators and erate incentives for quality improvement on all the ports to administer this system (see com- grain at a cost that would not be justified for panion report). Separate food grades and feed feed use in domestic markets. Dual standards grades have also been suggested for corn and would thus not facilitate an efficient market. wheat. Food grades would be more restrictive,

APPLYING ECONOMIC CRITERIA TO GRAIN STANDARDS

As the four objectives of grain standards This purpose would be met by tests that specified in the 1986 Grain Quality Improve- reflect storage history as well as predict- ment Act do not lead directly to a system of ing remaining storage life. Information on grades and standards, an intermediate set of infestation by molds, fungi, and insects guidelines is required to: needs to be accompanied by the extent of the development and deterioration. Kind ● Define uniform and accepted descriptive of infestation is also an important meas- terms to facilitate trade. ure of storability as a guide to actions re- This requires a small number of categories established by clearly defined factors. quired to inhibit further deterioration. The factors must be readily measured in ● Offer end-users the best possible informa- commercial trade and objectively deter- tion from which to determine end-product mined by technology that gives repeatable yield and quality. results at each point in the market chan- The characteristics of raw grain that in- nel. The factors must be acceptable to and dicate the quality and quantity of processed used by most participants in the market. products differ with different industries. Trade is also facilitated by stability and ab- Factors selected for inclusion in standards sence of change, since any change results should either be common to several indus- in uncertainty and adjustment. tries or be important to an industry con- ● Provide information to aid in determining suming a significant portion of the crop. grain storability. The more directly the factor measures the 210

desired end product, the more efficiently such as broken kernels, moisture, oil and pro- will the standards reflect value. tein content, and other intrinsic characteris- Create tools for the market to establish in- tics or physical properties that influence value centives for quality improvement. for the major processing uses. Higher or lower Incentives in standards are created in percentages for these do not necessarily mean part by including factors that are economi- higher end-use value over the entire range. For cally important. To provide the market the example, the required level of protein in wheat maximum opportunity to establish price depends on the ultimate product to be made incentives the standards should: 1) mini- from the flour. Lower moisture content means mize the distance between factor limits for more dry matter per pound, but 5 percent mois- each numerical grade; 2) report all values ture corn is not generally of greater value than as accurately as measurement technology 12 percent because of the effects of overdry- allows, using standard mathematical pro- ing. Usually some optimum value is indicated cedures for rounding to the nearest signif- for each of these factors, but the optimum varies icant digit; and 3) convey important eco- with the use and location in the market channel. nomic information to producers that will enable them to respond to producer prefer- Under Section 4 of the USGSA, factors con- ences related to value. - tained in the standards must be measured during any official inspection. Those considered official criteria are measured upon request. Al- Grade-Determining, though the advisability of that particular part Non-Grade Determining, and of the law is a matter of debate, in its present form it leads to the conclusion that characteristics most important to the largest number of The factors selected as indicators of value users would be incorporated into the standards. may be included as grade-determining factors, Those of lesser importance, or important to only as non-grade-determining factors, or as official a few users, would be considered official cri- criteria. As described earlier in this chapter, teria available upon request to buyers who need grade-determining factors set numerical grade them. Thus moisture and basic intrinsic prop- according to the factor limits established. Non- erties—such as protein content, kernel hard- grade-determining factors contained in the ness, and falling number tests in wheat; pro- standards do not influence grade but must be tein and oil in soybeans; and starch in reported as information whenever an official corn—might be incorporated as non-grade- inspection is made. Factors defined as official determining factors. Breakage susceptibility criteria are measured and reported only when and kernel hardness in corn and kernel size in requested. soybeans are examples of factors to be made available under official criteria. Assigning each factor to one of the three categories requires a guideline that can be used The advantage of putting the major factors objectively. As noted previously, standards in as non-grade-determining factors rather than should serve the needs of a majority of users official criteria is that the characteristics would and should reflect value for those uses. This move into the market channel much more read- suggests that grade-determining factors should ily. Obligatory measurement throughout the be those that relate to sanitary quality, purity, market would spread the cost across the entire and soundness (absence of imperfections). Us- industry. The cost per unit would be insignifi- ing this guideline, the grade would be based cant and therefore the information would be on factors such as impurities, foreign material, readily available as an incentive. A character- total damage, and heat damage. The lower the istic that must be specified by a separate re- values of any of these defects, the greater is the quest from each individual buyer would in- value of the product. Non-grade-determining crease the cost of information. For example, factors would be those related to properties the true value of information on test weight is 211

irrelevant under the present system since every- ing applied for additional levels within the abil- one is required to measure test weight. The mar- ity of sampling technology to differentiate. ginal cost to the buyer for that information is nearly zero. In contrast, if only one buyer speci- The base for the non-grade factors is of course fies oil content in a soybean contractor falling immaterial, since it is not grade determining, number in a wheat contract, the cost of infor- and the market is now free to choose what, if mation would be much greater because the cost any, price adjustment is to be made for differ- of measurement throughout the market chan- ent levels of those factors. nel would be borne by the single buyer and The zero base concept is limited by the free- would be spread over only the bushels that dom of the market to respond. Unless (or until) buyer purchased. export contracts and prices are established at zero base, merchandisers could start discounts Objective criterion based on the four purposes of standards has thus provided a basis at any level they desired, including the current for choosing factors and for dividing them factor levels for No. 2 corn and for No. 1 wheat among grade-determining, non-grade-deter- and soybeans. mining, and official criteria. Number of Grades Establishing Grade Limits The final question in setting standards is the number of different grades required. The num- As noted earlier, the grade limit on various factors is an automatic incentive throughout ber of numerical grades differs among grains— the market channel to add materials or to blend malting barley has three, soybeans and sorghum have four, and corn and wheat have five—and to reach that limit. Blending damaged soybeans with good ones does not increase the value of all grains have a sample grade designation. His- the damaged soybeans, but it does increase their torical records provide no rationale for these price, for they may now be sold as a higher numbers and the justification for the different grade. The criterion of providing incentives for numbers between standards is not clear. The improving quality dictates that the base be set fewer the numerical grades, the simpler is the marketing and the less space required to seg- at zero. The overall objective of a standard is regate these grades in storage. “to describe the value of the lot of grain being sold. ” If the percent of damaged kernels can A single grade would force the foreign buyer change from 5.0 to 6.9 percent without chang- to specify the quality characteristics and the ing grade, then numerical grade alone does not level of those characteristics desired. Buyers provide complete information on differences would no longer receive 4 percent BCFM by in value. Current soybean standards allow 1.0 default when ordering No. 3 corn. They would percent foreign material and no discount is ap- be forced to specify the levels desired and would plied by the market, implying that any level be- know in advance the trade-off with price. tween and 1.0 percent represents equal value. O The market seldom uses more than two The first 0.5 percent of foreign material in a grades. More grades increase the complexity shipment of soybeans has no more real value for the market and provide no increase in in- than the third 0.5 percent, even though the third formation. The disadvantage of a single grade is discounted and the first is not. is that nearly every buyer must use one grade Tighter limits on existing grade factor limits or specify levels on each factor. This would re- would reduce the incentives for blending and sult in increased diversity of contracts, less op- provide a more accurate measure of value as portunity to resell uniform lots, and increased long as discounts continue to be applied on the transaction costs. The final number of grades same grade. However, it is as difficult to justify to be established for each grain must be a com- an arbitrary limit of 0,5 percent as it is to justify promise between the purposes of providing in- an arbitrary limit of 1.0 percent. The only ob- centives, identifying value, and facilitating jective limit is zero, with market discounts be- trade. 212

EVALUATION OF RECENT LEGISLATION

Optimal Grade Prohibitions v. Market Incentives The optimal grade concept proposed a sin- Developing solutions to the problems and is- gle grade, with low values for selected defects sues raised by grain standards faces two basic and discounts on any defects above the base. choices: The object was to reduce the incentive for 1. legislative prohibitions against practices blending built into the current standards by that are detrimental to quality; and lowering the limits and thereby better meeting 2. changes in the economic incentives of the fourth objective of standards. It used the standards in pricing practices to allow the principle of grade factors being required to re- market to discipline offenders. flect cleanliness, soundness, and purity. It set low levels for that grade but not at zero. It would The first alternative focuses on controlling the have met the criteria of simplicity, facilitating process by which grain is marketed; the sec- trade, and removing incentives for blending. ond, on accurately evaluating the product and However, it did not meet the criterion for meas- value of different qualities. uring quality for various uses. Its greatest defi- ciency was a failure to identify the non-grade- Throughout history numerous bills and determining factors that would be incorporated amendments have tried to legislate specifics in in the standards. It also failed to eliminate com- grain standards. Nearly all legislative attempts pletely the incentive to blend off large quanti- have failed. The few successful activities since ties of poor quality grain by not setting the base the late 1800s have focused on setting policy values at zero. Congressional rejection of the and creating a framework for administrators optimum grade was probably more a reflection to implement rather than legislating specifics. of problems of implementation than of failure During the 1980s numerous bills or amend- to meet the criteria established by the purposes ments have been submitted to restrict the way of grades. in which foreign material, dockage, or dust can be handled, particularly in the export markets. The 1986 Grain Quality Improvement Act in- The Grain Quality Improvement cluded a prohibition against reintroducing dust Act Of 1986 or foreign material into the grain stream once it has been removed. The intent was clear: to Perhaps the most important contribution of improve the quality of the grain being exported this legislation was introducing into the USGSA and especially to improve the U.S. image in in- the four explicit objectives for grain standards, ternational grain markets. The success of this including three relating to economic value. type of legislation is not yet clear. Yet several Without these objectives, the standards had pri- difficulties can be identified in implementing marily reflected response to pressure from vari- prohibitions while leaving intact the standard ous groups. Without support from one or more structure that generated the incentives for major associations or organizations, it was ex- blending when purchasing on grade only. tremely difficult to make any changes in the standards because the only criterion was that The prohibition in the 1986 Grain Quality Im- of facilitating trade. As noted, FGIS now has provement Act focuses on controlling the proc- criteria on which to base changes and can ess rather than the product. Consequently, its justify those changes in terms of what is best enforcement has presented numerous problems, for the industry. not the least of which is a definition of what 213

constitutes foreign material, which the stand- and foreign material in the grain, nor does it ards define in many different ways and iden- prevent blending of different lots of grain con- tify with many different names. For example, taining various levels of foreign material to in corn only one factor encompasses all non- achieve the maximum allowed. It only limits grain material: BCFM, which is defined as any the procedure by which the maximum maybe material passing through the 12/64-inch sieve achieved. In addition, some if not most of the plus noncom material remaining on top. Wheat “foreign material” in overseas processing includes a factor for foreign material and one plants is broken grain created during unload- for shrunken and broken kernels as well as a ing of vessels. Consequently, it is unlikely that category called dockage. Barley standards con- the actual amount of “foreign material” deliv- tain three factors: dockage, foreign material, ered to the foreign buyer will change, but it may and broken kernels. In nearly all cases, foreign be more expensive to attain the contract levels material is defined in conjunction with parti- at the export elevator. cle size, meaning materials that got through a The most basic problem is one of trying to certain size sieve. The sieve size varies among legislate restrictions to counter the economic grains. In almost all cases, not all foreign material is actually removed, and in almost all incentives built into the standards themselves. Having an allowance of 4 percent BCFM in No. cases the foreign material contains small grain particles that are not “foreign” at all. (For in- 3 corn builds in an automatic incentive to add that much foreign material to the load when formation on problems in defining grain, see corn is purchased on the grade alone. Remov- ref. 1.) ing that incentive would be more efficient than Another problem with the prohibition ap- a prohibition. If a set of standards can be estab- proach is that the quantity of foreign material lished that creates incentives to achieve the or broken kernels incorporated into a grain desired end product in terms of quality, the shipment is controlled by the contract in con- legislative prohibition would become unnec- junction with the grade limit. Therefore, the essary. prohibition does not prevent leaving dirt, dust,

FINDINGS AND CONCLUSIONS

Continual review of grain standards since ● include arbitrary grade limitations on their inception in the 1916 United States Grain many factors that do not always reflect real Standards Act has generated numerous changes differences in value, and in some cases are and proposals for change in factors and factor not consistent with statistical principles. limits. These have not resolved the problems or foreign complaints related to quality. U.S. The many regulatory and legislative standard changes have failed to move the industry grain standards today: in a consistent direction. In fact, there have ● create incentives for practices not consist- been numerous reversals of previous changes. ent with good management and efficiency; What has been lacking is a clearly defined goal ● fail to identify many of the characteristics toward which the system is being moved. With related to value in use; the four objectives for standards now estab- ● fail to reward producers and handlers for lished by legislation, it is possible to develop improved drying, harvesting, handling, for each grain a set of “ideal” standards that and variety selection; and could provide a direction for future changes 214 and a yardstick against which to evaluate alter- of properties to be included. The criteria for natives. If each change adopted moves the oper- inclusion should be that the cost of obtaining ational standards closer to the ideal, repeated the information is less than the value of that reversals should no longer create unnecessary information to users who need the information. adjustment costs and confusion in domestic By starting with the major products generated and foreign markets. from each grain, a list of physical and intrinsic properties can be developed that correlates with Conflicts between purposes can be explicitly value in use. New, rapid objective testing tech- identified and the trade-offs and economic con- nology is also a requirement prior to inclusion. sequences calculated and recorded as guidance for a long-range consistent policy. The ideal sys- The list of factors to be measured under offi- tem should include grade-determining factors, cial criteria is almost unlimited except by meas- non-grade-determining factors, and definition urement technology. Any properties that can and measurement technology for official be accurately measured can be requested by criteria. buyers and sellers. These would be developed only after evidence of sufficient demand to Existing research is generally adequate to identify grade-determining factors—sanitary cover the cost. Information on the factors of interest for the various users could be provided quality factors, damage that reduces yield and by private laboratories and would be added to quality of processed products or value in use, official criteria only after rapid objective tech- and foreign material including dust. For No. nology is developed and when there is suffi- 1 grade, these factor limits should be set as close cient demand by domestic or foreign buyers to zero as measurement technology will allow. to justify including them. Any value above zero violates the third and fourth purposes of grades. The exact definition Standards should be designed to reward posi- of these factors, including sieve size for foreign tive actions, such as genetic improvement, and material, still requires additional research to sound harvesting, drying, and marketing prac- evaluate the alternatives against the criterion tices. They should also be designed to provide of reflecting value accurately. the best information available on the value of each shipment by descriptive terminology. All Non-grade-determining factors should meas- changes must be evaluated against the criterion ure value for a majority of users. The preferred of providing information that is worth the cost level may differ among uses. For example, splits in soybean reduce oil quality only in propor- of obtaining it. Optimum information, not maximum information, is the goal. Proposals for tion to the time in storage. Thus, domestic proc- change must also be tempered with current ca- essors buying for immediate use may allow high levels of splits with no discounts. Foreign pabilities of the industry, the cost of adjustments v. potential benefits, the realities of in- buyers, whose processing of the soybeans may ternational trading rules, and the historical lag harvest by 12 to 18 months, may place much sequence by which the industry has attained more restrictive limits and discounts on splits. the present situation. Measurement and de- Many of the intrinsic and physical proper- scription of quality is only one part of the prob- ties that influence the quantity and quality of lem. Quality must be evaluated in the context products derived from the grain have not been of technology, competition, foreign demand, identified. More research may add to the list and processing requirements. - - .- . -. - .—

215

CHAPTER 8 REFERENCES 1. Evans, Cooper, and Livingston, Kristi, Grain ground paper prepared for the Office of Tech- Quality: Positioning Ourselves for the Future, nology Assessment, U.S. Congress, Washington, Iowa Quality Grain Study Final Report, Cedar DC, 1988. Falls, IA, 1987. 5. Hurburgh, Charles R., Paynter, L. N., Schmitt, 2. Hill, Lowell D., “Analysis of the United States S. G., and Bern, C. J., “Performance of Farm-Type Grain Standards,” background paper prepared Moisture Meters,” Trans. American Society of for the Office of Technology Assessment, U.S. Agricultural Engineers 29(4):1118-1123, 1986. Congress, Washington, DC, 1988, 6. North American Export Grain Association, 3. Hill, Lowell D., Grain Grades and Standards: His- “Commitment to Quality,” Washington, DC, June torical Issues Shaping the Future, submitted to 1986. University of Illinois Press. 4, Hurburgh, Charles R., “Technology for the In- spection of Corn, Soybeans, and Wheat, ” back-