OCTOBER 1960 TISTICAL AN LYSIS or m: . euum GRAINMINDUSTRY ' y\_,_. SUMMARY AND CONCLUSIONS Sorghum grain has been under price support operations since 1941. The seasonal average price was above the average support price in every year during 1941-52 except in 1949. The seasonal aver- age price was below the average support price (except in 1955) during 1953-59. This was a result of increased sorghum grain acreage through di- verted cotton and wheat acreage and increased yields through hybrid sorghum grain seed which cause supplies to be in excess of domestic and foreign demand particularly during 1956-59. The total cost to the government from the be- ginning oi the program until September 30, 1959 was 235 million dollars. Five million dollars of this total cost were for the disposition of sorghum grain under Title I and Title 11 programs. Production for sorghum grain is concentrated primarily in Texas, Oklahoma and the Corn Belt. Kansas was second to Texas as a maior producing state. Texas produces about 42 percent of the na- tion's supply. Sorghum grain harvested acreage trends for Texas and the United States were similar during 1930-55. However, after 1955, Texas production in- creased at a considerably faster rate. This was a result of Texas producers using primarily hybrid sorghum planting seed, fertilizers and irrigation practices on diverted cotton and wheat acreage on the High Plains. The number of animal units on feed in the United States was compared with Texas and U. S. sorghum grain production. There was no particular relation between animal units on feed and sorghum grain production. The number of animal units on feed has been rather constant: whereas sorghum grain production has shown a consistently upward trend since 19.55. particularly in Texas. This per- haps can be attributed to guaranteed support prices and to the fact that sorghum grain was a more profitable alternative crop in utilizing diverted cot- ton and wheat acreage on the High Plains. The index of seasonal price variation for sor- ghum grain was developed using weighted average seasonal prices for 1947-56. Prices were adjusted for cycles and trends. A’ range of an expected monthly variation from the seasonal index of prices also was computed. Sorghum grain prices usually remained above normal from November to Iune and below normal during the remaining months. Sorghum grain prices had two seasonal highs in Ianuary and May and one seasonal low in August. The widest spread in price from a low in August to a high in May amounted to about 36 cents per hundredweight. The smallest variation in price from the seasonal average usually occurred in March. The optimum period for storing sorghum grain was between October and December and the optimum time to sell was between February and May. The seasonal index of pric, cated that the most inopportune ~ was between May and ‘August. i The difference in pfice variatih ghum grain and corn in Texas d ff about 8 percent; it was about 3 United States. During 1946-58, were less than 1 percent for both? United States, which reflected a b’ tween the demand for the two ~~f grains. 1' An analytical model was fitted 4 Texas harvested acreage data for tain the influence of diverted cottons age, a preceding year's produ - grain and price supports on the v' and U. S. harvested sorghum of these factors accounted for 62 - ation in Texas sorghum grain w. Diverted cotton acreage accounted tion than any of the other factoi model was fitted to U. S. harvest and the variables accounted for i variation in harvested acreage. .§ acreage on a national basis acc erably more variation in U. S. '_ than did the other factors. ' A statistical price model was ' Texas sorghum grain prices forth ‘ and 1946-57, to ascertain the p" fluencing the variation in sorgh ‘ Sorghum grain production, cotton a age combined. total of the other ' number of animal units on feed prices to sorghum grain prices wer to account for sorghum grain prici combined effect of these variabl period in Texas accounted for 5 variation; whereas in the postwar counted for only 44 percent. The i to the U. S. data and the same fact period accounted for 71 percent - i sorghum grain prices. All factors A percent of the variation in prices d A period. * General recommendations as ’ analysis are: ( 1) that the Federal‘. tablish for all grain producers a :-~ program for all feed grains and or quantity produced; (2) that r' tions and agencies continue and p’ programs to develop sorghum - ’ (3) that research agencies con‘ research relative to nutritive v Y grains, and (4) that the United S] of Agriculture continue to fu w»: sonnel and funds for investigm. of increasing foreign demand for 1' N has been under government 'ces since 1941, with the percent g from 35 percent in 1941 to 85 54. Sorghum grain has become f» commodity in Texas with the '_0r the 10-year period, 1948-57, bout 10s million dollars. The ‘iin1958 amounted to about 235 phich was about 2 percent of the 0m all agricultural commodities ghum grain production ranged ' bushels in 1948 to 238 million with a low of 54 million bushels ‘ averaged $1.19 per bushel during riod, While production averaged , bushels. tion attempts to ascertain the _. price patterns in the sorghum in relation to government pro- ‘feet of price programs for Wheat he supply of sorghum grain. A ount of statistical analyses Were i. factors responsible for changes . ted acreage and prices for sor- _ _.* ithin these analyses, the effect lfprograms for Wheat and cotton j ‘d into the sorghum grain model diverted acreage for Wheat and ly to determine the effects of 7» support price programs on the f‘. grain. .1 grain industry under govern- rice programs was analyzed on a i; well as on a separate state basis “as produces 42 percent of the grain supply. g s 01-" FARM PROGRAM OPERATIONS t to single out any specific legis- um grain because most of the Ybeen under the classification of an under a specific commodity- grain, rye, oats and such. How- “statistics on Commodity Credit rations were obtained relative to éstocks, value and cost of disposi- ’ grain since the initiation of the 1. Corn usually has set the price i other three principal feed grains n1 oats and barley. Even though ysproduced a large segment of rghum grain supply, there never A STATISTICAL ANALYSIS g 01-" THE SORGHUM GRAIN INDUSTRY I. A. KINCANNON. Associate Professor Department oi Agricultural Economics and Sociology has been a supply problem until the past several years, particularly since the advent of hybrid seed. In the past 2 or 3 years, however, the CCC inventories of sorghum grain have been ratherlarge, particularly for 1958-59 crops- The quantity of sorghum grain under contract to pur- chase as of September 30, 1959 was 275 million hundredweight. The cost of this contract amounted to 695 million dollars with a reserve for losses of 220 million dollars, resulting in a net book value of 475 million dollars. The quan- tity of. sorghum grain pledged for outstanding loans as of September 30, 1959 was about 14 million hundredweight, or a CCC investment valued at 25 million dollars. As of this same date there was 275 million hundredweight of CONTENTS Summary and Conclusions . . . . . . . . . . . . . . 2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .. 3 Highlights of Farm Program Operations _ 3 Sorghum Grain Production . . . . . . . . . . . . . . 5 U. S. and Texas Harvested Acreage and Yields . . . . . . . . . . . .. 6 U. S. and Texas Sorghum Grain Production Trends . . . . . . . . . . . . . . . . . . . . 6 Sorghum Grain Production and Livestock on Feed in the United States and in Texas . . . . . . . . . . . . . . . . . . . . . .. 6 Sorghum Grain and Corn Prices in the United States and in Texas . . . . . . . . 7 Adjusted Seasonal Index and Zone of Price Expectancy for Texas Sorghum Grain . . . . . . . . . . . . . . . . . . . . . . . 7 Index of Seasonal Variation . . . . . . .. 8 Zone of Price Expectancy . . . . . . . . . . . 9 Variations in Sorghum Grain and Corn Prices ior 1930-40 and 1946-58 . . . . . . . . . 9 Statistical Analysis of Harvested Acreage and Prices . . . . . . . . . . . . . . . . .. 9 Harvested Acreage Model . . . . . . . . . . . 9 Texas Harvested Acreage Data, 1946-58 . . . . . . . . f . . . . . . . . . .10 U. S. Harvested Acreage Data. 1946-58 . . . . . . . . . . . . . . . . . .10 Price Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Texas Data. 1924-40 . . . . . . . . . . . . . . .12 Texas Data, 1946-57 . . . . . . . . . . . . . .12 U. S. Data. 1924-40 . . . . . . . . . . . . . ..12 U. S. Data, 1946-57 . . . . . . . . . . . . . ..13 Acknowledgments . . . . . . . . . . . . . . . . . . . . . .13 TABLE 1. U. s. SORGHUM GRAIN INVENTORIES ACQUIRED UNDER THE PRICE-SUPPORT PROGRAM BY i SITION FOR SEPTEMBER 1959 AND FISCAL YEAR 1'0 DATE - _ _ _ Unit o} 191d Sales tor dollars Public Law 480—export Ccimnwdity‘ pemd and “em measure dispesitiens Domestic‘ Exporti Title 13 Title n“ ( I — — — — — — — — — - — - - 1.000 4911a... _ _ _ _ l Sorghum Grain , Current Month 4 Quantity Cwt. 2,898 1.503 3 189 i’ 122 Cost value $11,122 3.685 18 571 462 Proceeds $ 5.376 2,398 7 558 401 Fiscal Year Quantity Cwt. 6.953 4.852 3 425 309 c951 value $29,795 19.202 1s 1.299 1.02s s‘ Proceeds $13,846 9,038 7 1.262 983 l y ‘Includes inventory gains. losses and related recoveries. Also includes quantitative gains and losses in pr ilncludes some sales which may be applied subsequently to barter contracts or P. L. 480. Title I authoriz -»' classification will cause downward adiustments in “Sales for Do1lars—Export." aProceecls represent the Corporation's lull investment amount charged to the statutory limitation. ‘Proceeds represent exchange value of strategic or other material to be delivered under contract. ‘Amount delivered as payment in kind for exportation under P. L. 480 is included in P. L. 480 (Export) Titlel I included in this column. sorghum grain in CCC inventory. This Was an investment of approximately 695 million dollars. There was, as of September 30, 1958, a total CCC investment, before a reserve for losses, of 185 million hundredweight of sorghum grain which Was valued at 433 million dollars. The total investment in 1959 was almost twice that of the 1958 investment in CCC inventories. Farm programs for wheat and cotton have affected the supply of sorghum grain consider- able, particularly in Texas. As a result of cut- backs in acreage for wheat and cotton and high support prices for sorghum grain, farmers gave considerable impetus to diverting acreages from these two crops to that of sorghum grain. The use of sorghum grain as an alternative cash crop on diverted cotton and wheat acreage has been practiced primarily in the Rolling Plains cotton area and in the southern portion of the Panhandle wheat belt. In the past several years, sorghum grain has been planted as an alternative cash crop in the Gulf Coast area. Within this area, however, sorghum grain was produced primarily because of the ease with which it could be har- vested with rice combines and the profit that could be realized by producing the crop relative to other agricultural commodities. Sorghum grain production in the Gulf Coast area was not necessarily a result of diverted acreage from any basic commodity as was true in the High Plains and Panhandle areas. The disposition of CCC inventories of sor- ghum grain under Title I and II contracts TABLE 2. ACQUISITIONS OF SORGHUM GRAIN UNDER THE PRICE SUPPORT PROGRAM FROM PRO CENT YEARS THROUGH SEPTEMBER 30, 1958 ’ amounted to nearly 2 million h “ an investment cost of 3.8 million i stocks were accumulated from t @ the program through September‘ gross cost to the government of, portation under Title I and the transportation under Title II a F’ proxlmately 1 million dollars. i. of sorghum grain under Title I 1i has cost the government approxi - dollars. However, this cost is ..__ compared to other commodities J cotton, corn, tobacco and rice. v of inventories acquired under th programs, by type of disposition f 1959 and fiscal year to date, is sh The table gives a breakdown on - domestic export, Title I and Title for sorghum grain. The acquisi gram under support price prog ' duction during 1955-58 are giv‘ Table 2 shows that the larger " sorghum grain stocks and loan; operations took place since 1957. I that these figures in Table 2 wi the 1959 crop since 1959 was a r year, particularly in Texas. ' The total loss from the incepti program operation for sorghum ‘- September 30, 1959 amounted to. 235 million dollars. However, L the fiscal year’s 1958-59 stoc :1 roughly 18 percent of the total w gram from its beginning to 1959. Commodity and Unit of Pumhases (3011016161 production year measure Quantity value Quantity I Sorghum grain 1955 Cwt. 125.419 221,272.00 51,735,179 1956 Cwt. 200.347 392,627.31 18,000,608 1957 Cwt. 3.935.988 8,044,309.45 151,517,060 1958 Cwt. 3.897.927 7,458,839.75 133,857,051 GRAIN. AVERAGE PRICE SUPPQRT rice. These mandatory support prices were to gfiogglgg%pngiligggn BY FARMERS begin with the 1959 crop. Average seasonal The percentage of_ parity, average support out support level average price level for sorghum grain and the seasonal aver- i ‘iv for sorghum reseived by age price received by farmers since the govern- 9mm “w” i°“'m°'s ment support price program was initiated in -- - - Dollars --—- 1941, are given in Table 3. The seasonal aver- 535.0 .54 .87 age rice received b farmers since the start -° ~71 1-03 of the price prograih was greater than the } r0 .98 1.41 ' ' d ' ; _o L70 L53 1nc us1ve, w1 e excep 1on o W en e ~l8.0 1.65 ‘ 2.14 seasonal average price was only 7 cents below the average support level. The seasonal average :0 2:31 2:29 price was below the support level during 1953-59 _g m9 gm inclusive, with the exception of 1955 when the ‘l .0 1.87 1.88 two prices were equal. "8 a; 1g SORGHUM GRAIN PRQDUCTION i133 g2 Sorlghfum gralin fprodgétiog ‘yvvas, Ifor many . . . ears e o an a er or ar , - -° 1-35 1-94 lzrrated prirrlgrily in Texas and Oklahoma. coTllrlgge fig fi-tllg two states produced roughly two-thirds of the nation’s supply of sorghum grain until the latter part of the 1940’s. Since sorghum grain can be produced in semi-arid as well as irrigated re- n listed under “other nonbasic gions, and since the crop is a very profitable cgbut was reclassified effective alternative enterprise, the production in the early ,959 on the basis of Public Law fifties began to spread north into the Corn Belt, ,w provided for mandatory sup- all through. the Southeast and through the South- §barley, sorghum grain, oats and west as far as California, Figure 1. Texas and y? .. 5,585 -» '° 8| 552 '3=°9° ' I l,200l,59 12,454 W |28,964 35° . 3598 L980 l . 3 445 I8 45o "888 ' 45G 4.836 8,085 ' 3,28 |,58 9 I2 8 88 O Q . O x‘. .. i . I . . ' 0.09 g .0‘. O .l$gf‘b‘.fg ‘o! . ' - °.'-.‘ .'-".'?-f-".'-.-:i=;‘.E~s.'-%.*I;~ " " ' ' o .ovq¢~sb, 02f 7"%'v- 0' ‘ . . . .0.0:::3.: Q...I:\'.\',I‘ ‘:5 .00 . O Q ‘itz-{Ygfitv -° 215,055 a a’. sorghum production in thousands of bushels,_ by states. 1958. The dots represent areas of concentration oduction. in Texas and are based on the 1954 census data. Each dot equals approximately 10,000 bushels. 5 MIL. ACRES I ' —- TEXAS - --- u. s. 2o ' I :“~. .. 1 \_ - .-' 15 _ , " 1 I I I’ \‘ i 1o ' ‘~' _ 1‘ I \ 1' ' .. f s‘ f \ I _ l’ lax I,\" \\ I‘ f _ ...._/ \\, \ 1 5 __- ’__-,1, ‘\ \/ v i’ \\/l\”r J /\/\/ V \/§-?“ p 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 1 I 1 1 I930 I935 I940 I945 I950 I955 I960 Figure 2. Sorghum grain harvested acreage, United States and Texas. 1930-58. Kansas are the major producing states, with Texas producing about 42 percent of the nation’s supply. U. S. and Texas Harvested Acreage and Yields Sorghum grain harvested acreage for Texas and the United States during 1930-58 has a comparable trend pattern, but with the trend in harvested acreage for Texas increasing consid- erably during 1950-57, Figure 2. The U. S. har- vested acreage trend increased about as much during 1950-58 as it did during 1947-50. The rapid rate of increase in the production trend of sorghum grain, Figures 3 and 4, was more a result of rapid increases in yield than in the number of acres. the use of hybrid seed since 1955. This was due primarily to Before 1956 the trend in acreage and yield increased at about the same rate. Since 1956, U. S. sorghum grain yield has increased faster, Figure 5, than has Texas yield, but Texas sorghum grain harvested MIL. ACRES I 40 | BUSHELS — ACREAGE --- YIELD 30 \ \~_ ‘n s 20 I c- “I \ I\ I’ \\ 1' 'I \vl \'I llll llll llll lllllll I l l j I930 I935 I940 I945 I950 I955 I960 Figure 3. Texas sorghum grain harvested acreage yield. 1930-58. 6 40 3'0 20 and acreage increased at a faster rat harvested acreage. Figure 5 1 trend in yield for the United S, for 1930-58 was about the same. vested acreage trend during 193i to the Texas trend, increased on a more rapid rate. than did perhaps was due to lthe rapid sp a grain production all over the So west after the introduction of a 1955; yield did not increase as '31 nation as a whole as it did in t_i tions of the Rolling and High Texas. u. s. AND TEXAS sons p PRODUCTION T The trend in production for? United States for 1930-58 sho Q duction trend for Texas and the was almost the same until hybri seed came into widespread use, that time there has been a phei. production trend for Texas sorg '_ was perhaps a result of consi acreage in cotton and wheat, most of Texas’ sorghum grainiis a the cotton and wheat areas. trend increased considerably : therefore sorghum grain was a? alternative crop for these divert ers were free to plant any cotton in 1951-53 and for wheat a have been acreage allotmentsf cotton in every year since 194 spectively. A’ larger proportio ghum grain is planted in the in the Wheat Belt, and perhaps there has been a greater propel in production since 1956 in Te, nation as a whole. it SORGHUM GRAIN PROD i uvzsrocx ON FEED 1N STATES AND IN Since annual numbers of li for Texas were not available, 11f stock on feed for the United ‘a 7 and 8, were compared with , grain production. There has no relationship between the nu 1 on feed and the production of w; Texas, Figure 7. Actually, number of livestock on feed dur almost a horizontal line; wher‘ for Texas’ sorghum grain prod _ increased at a rather rapid rate, advent of hybrid seed) andth almost a vertical rate. a Figure 8 illustrates the sa between sorghum grain and num_ on feed for the United Statesi‘. I940 I945 I; S. sorghum grain harvested acreage and BUSHELS —ACREAGE _ ---YIELD 4O I ’ _—l I II I 30 I I . I " I lr-\ /' ‘l; a‘ I 0-’ I, ‘VI/xxx,’ \’ ,- /\ ,. I ._ *\ 4 \I ~ llllllllllllllllll I950 I955 I960 een U. S. and Texas sorghum fon and livestock 0n feed is only increase of sorghum grain pro- j the number of livestock on feed, (8, is for the United States. How- sorghum grain production trend l did not increase quite as rapidly jgéproduction trend. The faster in- sorghum grain production rela- i oduction, particularly since World rhaps a result of Texas sorghum tars invariably utilizing diverted for sorghum grain production. production has been the most frnative crop on Texas diverted In other principal sorghum states, in which diverted wheat I d the greater amount of available éthe resources were devoted to the orn and forage crops. I RAIN AND CORN PRICES IN STATES AND IN TEXAS Texas sorghum grain seasonal .followed a close pattern during ,8 was to be expected since U. S. illum grain prices followed a price l arily by corn prices. U. S. and es varied slightly during 1930-58. e price-determining feed grain, a was expected between the Texas price and the U. S. average an- e only major deviation in prices o commodities was in 1944, at S. average price took a sudden ilTexas average price actually in- During 1930-58, however, Texas iétuated on the average, slightly corn prices. y, um grain prices during 1930-58 ;.c0rn prices throughout every year The corn-hog program and the fie “A” program might have af- BUSHELS | 4O 2 O I _ A Fxvl/fi; Ix f /\v \_ f‘! V’ lllllllllllllllllllllllll I930 I935 I940 I945 I950 I955 I960 Figure 5. Sorghum grain yields, United States and Texas. 1930-58. fected corn prices more than sorghum grain prices during these 2 years. The two prices were about equal in 1935, 1936 and 1939, with the greatest difference occurring in 1947. The 1947 crop year was the best season, pricewise, that farmers have experienced since World War II. U. S. sorghum grain and corn price patterns were similar to the Texas sorghum grain and corn price patterns except for more price stabil- ity in U. S. corn prices than for Texas corn prices. However, this was to be expected since the U. S. average price for most commodities usually will be more stable than the price of a commodity in a particular state. ADIUSTED SEASONAL INDEX AND ZONE OF PRICE EXPECTANCY F OR TEXAS SORGHUM GRAIN Many factors such as weather conditions, business activities and trends in farm prices and farm policies, here and abroad, alter the seasonal M I L . B U. I / _ --- TEXAS f, 50o — u. s. I I I 40o ; _ I I 300 20g Ema , II‘. I" ‘t: / '\ l’ \\ I’ . T I’ \ Ix’ ‘\ I’ lQQ .r~~-_/ ‘ \ ‘V’ f l l I l l l I l l l l l l I l l l l l I I l l l4 I930 I935 I940 . ,I945 I950 I955 I960 Figure 6. Sorghum grain production. United States and Texas, 1930-58. 7 MIL. BU. I i MIL. UNITS - TEXAS PROD.’ _ --- u.s. LIVESTOCK / 250 f 20o _ I l, X l .. I \\ _ I - ’_\ g ‘NJ! "\\\”,’ \\_‘,¢~_ z I50 ' \ ,' 10o’ V“, ' V I 5o- _ N1111111111111111111111 I930 I935 I940 I945 I950 I955 I960 Figure 7. The relationship between U. S. numbers oi livestock on teed and Texas sorghum grain production. 1930-58. price movement of a commodity in any particular year. The variation in price patterns for some commodities remains approximately the same year after year, while for other commodities the varia- tion is pronounced. Therefore, the average ad- justed seasonal variation for Texas farm com- modities should not be applied as a criterion of price change for any one year until adjustments are made, for current as well as probable future changes in economic conditions. Any particular analysis based on past years is a relative concept and should be relied on with discretion when used as a forecast of future years. The index of seasonal variation for sorghum grain was developed, using the weighted season average monthly prices, for 1947-56, Figure 9. The prices were adjusted for cycles and trends. The zone of price expectancy was calculated for 1947-56 to afford a measure of the monthly variation in prices from the seasonal average price for the 10-year period. The zone is the range of the average seasonal price that can be expected for MIL. BU. l I _ — u.s. Imoo. ' _ --- u.s. uvssrocx 500 I 400 ’ , I 300 J 200 ,,-\ ':’I—~\\\'n__-—¢~ p’ ‘_~_--’ "&prf“-._" _ I00 A v lllllllllllllllllllllllll I930 I935 I940 I945 I950 I955 I960 Figure 8. The relationship between U. S. numbers of livestock on feed and U. S. sorghum grain production. 1930-58. MIL. UNITS I any particular month, in approxi I of 10 years. Index of Seasonal Vari I Sorghum grain prices followed pattern, Figure 9. They remained a, from mid-Novemberte} mid-June an mal for the remaining“ months. S0 prices were at normal only three this period. These three normal pe sonal variations in prices were in m' between mid-June and mid-July mid-November and mid-December.‘ grain prices had two seasonal hi seasonal low. The two seasonal hi January and May, while the low wa The index of seasonal variation j sorghum grain reached a low of 931 and a high of 109 in May. This m_ ence of 16 points or 36 cents per hu The widest spread above and below seasonal price variation was" f through October. This was to be view of the heavy marketings of so during these months and competitive, reaching the market. This relatio depending on the supply of sorgh well as the potential supplies of com I The smallest variation in price w when the smallest quantity moves in and the demand is relatively stable. most profitable time to store sorgh when the supply is large or prices Therefore, the best time to store w tween mid-October and mid-Decem, best time to sell, between mid-Februai May. The least profitable time to s ‘ grain, relative to the seasonal p a would be between mid-May and Decisions that concern when to sto should be compared with average c0 vailing relative to the general econ I prlce levels, economic conslderatioi INDEX I I I20 --- SEASONAL? —— PRICE z I00 nob //\ I 90 80 JFMAMJJAS? Figure 9. Sorghum grain adjusted so prices and zone of price expectancy. Texan, i, and government programs closely ap- ~ those that prevailed during the base i‘ ne oi Price Expectancy _ ount of variation in the price for a th from the average price prevailing nth for the 10 years, 1947-56, was Pas follows: if the average price for , := 120 percent of the annual price and of price variation was 5 percent, this ‘n that the price in that month varied 4- index of 115 to about 125 in approxi- ;0ut of 10 years. The narrower the e index of the price zone, Figure 9, r" the stability and closeness of the .sonal price to the average price for ‘r period. Conversely, if the value is a monthly prices in individual years siderably from the average monthly a NS IN SORGHUM GRAIN AND _CES FOR 1930-40 AND 1946-58 tical technique was used to determine tage variation between corn prices ‘ m grain prices during these two pe- e 4. These percentage variations in orghum grain and corn were computed nd the United States for both periods. al difference between the percentage If sorghum grain compared with corn during 1930-40 was that sorghum _;.- varied about 8 percent more than 'ces. During this same period, U. S. _ 'ces varied only about 3 percent more prices. The greater variation in Texas _ ain prices varied only about 3 percent _corn prices. The greater variation in ‘hum prices as compared with corn v g 1930-40 was perhaps a result of um grain being produced primarily as ‘crop. Consequently, Texas sorghum h»: followed as closely as did U. S. sor- (.1 prices the commercial grain price ».Was determined by corn prices. _. 1946-58, Texas sorghum grain prices ut .5 of 1 percent more than Texas ,gwhile U. S. sorghum grain prices .f 1 percent more than corn prices. _ variation in Texas sorghum grain mpared with corn prices during 1946- ‘CENTAGE VARIATION OF TEXAS AND U. S. M THE AVERAGE PRICE DURING THE . PERIODS, 1930-40=AND 1946-58 Texas I United States 1930-40 1946-58 1930-40 1946-58 — — — — —-—Percent—————- 9.1 19.6 37.6 19.4 31.1 19.1 34.5 18.7 58 resulted primarily from the shift in the pro- duction of Texas sorghum grain from a “catch” to a “cash” crop and to a greater balance between the demand for Texas and U. S. sorghum grain and corn supplies in meeting the needs, both in quantity and nutritional values, for an ever in- creasing postwar livestock feeding industry. STATISTICAL ANALYSIS OF HARVESTED ACREAGE AND PRICES An analysis was made to determine the fac- tors responsible for the variation in Texas har- vested acreage and the price of U. S. and Texas sorghum grain. Texas sorghum grain harvested acreage model (following) was considered for 1946-58; the U. S. and Texas sorghum grain price model (shown on page 12) was made for 1924-40 and 1946-57. Harvested Acreage Model The following model was fitted to U. S. and Texas harvested acreage data during 1946-58: Y Z Bo + B1X1 "l" B2Xz + B3X3 + B4X4 Y = harvested acreage for sorghum grain (1,000 acres) X1 = diverted acreage in cotton as percent of 1951, 1952 and 1953 average X1 = diverted acreage in wheat as percent of 1947, 1948 and 1949 average X3 = 1 year lagged production in sorghum grain X4 = U. S. sorghum grain support prices More emphasis was placed on the analysis for harvested acreage than on the analysis for sorghum grain prices, since one of the basic ob- jectives of this manuscript was to determine the effect of Federal price programs, which were initiated at the beginning of World War II, for cotton and wheat on the supply of sorghum grain. However, the importance of knowing the price behavior of sorghum grain is not dis- counted; but since corn prices generally set the pattern for sorghum grain prices, a more thor- ough analysis on harvested acreage was deemed necessary to arrive at possible effects of cotton and wheat programs on the supply of sorghum grain. Since a detailed analysis was needed on har- vested acreage and factors responsible for varia- tions in harvested acreage of sorghum grain, Equation I, for example, was fitted with Texas data and a significance test was run to detect the factor that contributed least to the cause of of variation in harvested acreage and therefore could be eliminated from the equation. Then Equation II was fitted using the remaining variables and again the factor that contributed least to the variation in harvested acreage was eliminated. This procedure was repeated (see Equations III through VIII) until only one fac- tor remained in the final equations (Equations IV and- VIII) that contributed more than any other single factor toward the cause of the variation in harvested acreage of sorghum grain during 1946-58. However, this procedure of deleting variables was not followed for the analysis 0n prices. Only total variation and partial correlation coefficients were estimated. A partial correlation coefficient is a measure of the extent to which that part of the variation in the dependent variable (price), which was not explained by the other independent factors, can be explained with the addition of a new factor. Texas Harvested Acreage Data, 1946-58 EQUATION I Y = 10360.49 — 3872.03X1 — 766.39X2 + .0089X5 — (2458.47) (1131.68) (.00765) 1453.73X4 (1917.26) These four factors accounted for 62 percent of the variation in harvested acreage during 1946-58. The figures in parentheses are stand- ard errors of the regression coefficients. As attested by the “t” values, the partial correlation ry2.134 had the smallest value. The “r” value means that out of the 38 percent unaccounted for, diverted acreage in wheat accounted for less of the unexplained variation than any of the other three variables. As noted in Equation II, with diverted acreage in wheat omitted, the R2 is 64 percent, or a difference of 2 percent improvement in the variation of harvested acre- age. The difference (R2y_1234 -— R2y_134)’ between the 62 percent accounted for in Equation I and the 64 percent accounted for in Equation II, when divided by 38 percent (the difference between the value of R2 in Equation I and 100 percent), gives the squared value of the partial correlation coefficient. If the “r” (—.23) value is squared, a 5.26 percent reduction in the unexplained var- iance is ascribable to diverted acreage in wheat. Since by2_]34 was negative, the sign of ryms, was also negative. A similar explanation of the meaning of the r’s and the adjusted R’s hold for each successive equation. EQUATION II Y = 9280.21 — 4119.92X, + .0111X. —- 1157.08X. (2387.25) (.00707) (1787.77) The reason the analysis was improved, that is the R2 value, in Equation II compared with that of Equation I, was because by eliminating diverted wheat acreage some intercorrelation or serial correlation between independent variables in the equation could be reduced. Therefore, by eliminating an intercorrelated or partially cor- related independent variable, there would be some improvement in the percentage variation in the dependent variable accounted for by the remaining independent variables. A variable that may show no correlation with the dependent vari- able would show significant correlation after the relation of other variables has been allowed for in the equation. The value of ry4_13 (-.22), for example, indicated that support prices accounted for less of the unexplained variation in prices and l0 was therefore deleted before fitting E, to diverted cotton acreage and lagged data. a EQUATION III = 7082.07 — 4955.08X. + .0120x. (2134.80) (.0007) As shown in Eqiuation III, wit wheat acreage and support prices re 7-. the equation, the percentage variation for, was improved to slightly over 66v shown by the R2 values in this equ verted acreage in cotton and 1 year l, duction in sorghum grain contribut‘ the variation in sorghum grain harv age. The lesser importance of sorg lagged production relative to divert in cotton in Equation III was verifiedj value of .59. ‘- EQUATION IV Y = 9555.20 — 0108.98X. (2440.00) With 1 year lagged production i grain eliminated, 53 percent of the sorghum grain harvested acreage Was. for by diverted acreage in cotton. percentage of the variation accoun ,_ diverted cotton acreage was to be e o! the larger proportion of sorghum 1 .5 duced in the Cotton Belt of Texas. n remaining 47 percent that was not ac . can perhaps be attributed to the fact v able proportion of the total quantity |_ grain was produced in the Gulf Co, Texas where diverted cotton acreage significant factor. . The reduction in unexplained v Texas harvested acreage, by the ad new factor in Equations IV through tively, is given in Table 5. P U. S. Harvested Acreage Data, 1946-58 =§ EQUATION v _ Y = 10078.05 -- 4448.58x. - 170087720 +5 (7182.97) (9745.01) 452.09x. (4870.80) The harvested acreage equation same independent variables was fi _ U. S. data during the same period With all four factors included in 4 the correlation was 89 or, in other _ percent of the variation in U. S. sor harvested acreage was accounted for factors. In this equation, as estima “t” test, the support price contri (ry4_123 I —.037) to the variation in acreage and was therefore deleted. EQUATION VI Y = 26006.30 —— 4869.89X1 ~— 17000.86X= (5625.36) (9156.94) ‘UPI TIVE IMPORTANCE OF INDIVIDUAL FACTORS AFFECTING THE DEPENDENT VARIABLE (HARVESTED » ACREAGE) TEXAS. 1946-58 f (Support prices) h VI was fitted to the data again using cage in cotton, diverted acreage in year lagged production in sorghum result, after elimination of support j= the independent variables, was roved and acounted for 72 percent ion in harvested acreage. Similarly, indicated that diverted acreage in uted less than did diverted acreage é lagged production of sorghum grain. ted by the partial correlation co- ,1. and is equal to —.29. Diverted otton was therefore eliminated and ,I was fitted to the data using only ‘age in Wheat and lagged production. EQUATION VII I — 21313.76X2 + .0099X.. I (8595.39) (.0069) ating diverted acreage in cotton ‘g only lagged production and di- ge in wheat, the percentage varia- roved by only one half of 1 percent. _- “t” test indicated that of the two riables (diverted acreage in wheat roduction), lagged production con- ’ , as is verified by the partial cor- iicient, ry... and is equal to .456. EQUATION VIII F 25664.67X. (9178.20) VIII was fitted to the data using ~ acreage in wheat and this single acreage) . . Reduction Coefficient gggré: Factor The ggeégftlfafit in unex- of mu_ltipl_e considered added r s correlation vpgggllzei deteragahon 1 (Diverted cotton acreage) None In.» ——-753 5-57 -53 --‘ 1 (Diverted cotton acreage) X3 (lagged production) r1111 .590 .348 .66 1 (Diverted cotton acreage) t X1 (support prices) t...” — .22 0.48 .64 (Lagged production) 1 1 (Diverted cotton acreage) » (Lagged production) X1 (diverted Whefli Tyzuiat — .23 0.53 .62 factor accounted for a little over 68 percent of the variation in harvested acreage for U. S. sor- ghum grain. The results of these series of equations fitted to U. S. data were to be expected since diverted acreage in wheat for the nation as a whole would be more significant than diverted acreage in cotton, particularly since diverted acreage in cotton is primarily utilized for the larger pro- portion of the Texas sorghum grain production. Diverted wheat acreage alone accounted for about 16 percent more variation in U. S. sor- ghum grain harvested acreage than did diverted cotton acreage alone account for the variation in Texas sorghum grain harvested acreage. This was to be expected since the U. S. model included diverted wheat acreage in the High Plains of Texas Where most of the Texas sorghum grain is produced. Also, a sizeable quantity of sor- ghum grain is produced in the Texas Gulf Coast area in which there is no diverted wheat acreage and in which diverted cotton acreage is not as significant as in the High Plains area. In fitting the equations to the U. S. data, recognition of the contribution of Texas diverted Wheat acre- age and diverted cotton acreage to the results was necessary so that more logical interpretations of U. S. analyses could be made. The propor- tional significance of Texas diverted wheat acre- age relative to Texas diverted cotton acreage also must be considered when interpreting the results of Texas analyses. For example, Kansas, other than Texas, produces nearly as much sor- ghum grain as the other principal producing TIVE IMPORTANCE OF INDIVIDUAL FACTORS AFFECTING THE DEPENDENT VARIABLE (HARVESTED ‘ ACREAGE) UNITED STATES, 1946-58 pf (Lagged production) . . Reduction Coefficient Zigzag: Factor The cggeégiiecfit in iinex- of multiple considered added r s correlation plained determination h Y variance (R2) (Diverted wheatiacreage) None r,“ —.843 .711 .68 " (Diverted wheat acreage) X1 (lagged production) r,“ .456 .208 .73 (Diverted wheat acreage) ( X1 (diverted cotton fyLza —— .29 .084 .72 Lagged production) acreage) ' (Diverted cotton acreage) J (Diverted wheat acreage) X1 (support prices) r1111” ——-037 301 39 11 states combined and Kansas is primarily a Wheat state. When lagged production was eliminated from U. S. data, leaving’ only diverted Wheat acreage, and from Texas data, leaving only di- verted cotton acreage, the variation in Texas harvested acreage was considerably more than was the variation in U. S. harvested acreage, Tables 5 and 6. This perhaps was attributable to a stronger relationship between U. S. diverted wheat acreage and U. S. sorghum grain produc- tion than was true for the relationship between Texas diverted cotton acreage and Texas sorghum grain production. Texas sorghum grain is a major farm commodity and is grown primarily as a cash crop and therefore Texas producers would perhaps pay more attention to a prior year’s sorghum grain supply and alternative crops than would be evident for the nation as a whole. The reduction in unexplained variance in U. S. harvested acreage, by the addition of a new factor in Equations VIII through V respectively, is given in Table 6. Price Model The following model was fitted to U. S. and Texas sorghum grain price data during the pe- riod, 1924-40 and 1946-57. All pertinent coeffi- cients for the U. S. and Texas data are shown in Table 7. Y 7i B0 "l" B1X1 + BZXZ + B3X3 + B4X4 + B5X5 Y = price of sorghum grain (cents per bushel) X. = production of sorghum grain (1,000 bushels) X2 = cotton and wheat acreage combined (1,000 acres) X3 = total corn, barley and oats ( 1,000 bushels) X1 = number animal units on feed (1,000 units) X5 = ratio of corn prices to sorghum grain prices TABLE 7. RELATIVE IMPORTANCE OF INDIVIDUAL FAC- TORS AFFECTING THE DEPENDENT VARIABLE (PRICES) Coefficients Equ°ti°n 4.1121211111111011 (R2) variance‘ Texas 1924-40 Ix .52 131.2345 23 1946-57 ' 134.1235 -44 X .44 132.1345 J2 United States 1924-40 133.1245 -31 XI .71 135.1284 -1 9 1946-57 1' 312.1845 -24 1' y3.1245 -57 XII .77 1' 514.1235 =18 135.1234 .32 ‘All r's not listed were zero. 12 Texas Data. 1924-40 EQUATION 1x _ Y = 1.8042 - .0014x. - .00021x. - L (00009) (.0087) .00081x. + .1100x. (00078) (5802) Equation IX was-jfitted to Te termine how much‘ of the vari 7 sorghum grain prices was attribu independent variables. All fiv, variables accounted for 52 percen tion in Texas sorghum grain pric 40. The production of sorghum 1 i, most to the variation in prices, f0, ber of animal units on feed. Th partial determination (r2yL2345) for the production of sorghum 5g zero for all other variables. =3 reasonable since sorghum grai‘ period, was not considered a maj planted primarily as a “catch” c principally as an on-the-farm feed. fore, a prior year’s supply and ( animal units on feed were the pl in accounting for more than 50' variation in sorghum grain pri, period. . Texas Data, 1946-57 g- A When the same equation W: _ Texas data during 1946-57, (Eq results differed considerably fro pre-World War II period. ' EQUATION x _ Y = -8.1085 + .000029x. + 0039K. _ (00081) (.0029) ;. .0040x. + 1.589x. -' (.0010) (18084) One of the primary reasons fered was because sorghum g 1 government support price progra 57, and subjected to free market mg 1924-40. Consequently, a privy or production was of no conseqj during 1946-57. All five variabl 44_ percent of the variation in prices during this 1946-57 pe, important variable during this L number of animal units on fj deleted from the equation, re utable variation in sorghum gr is verified by a coefficient of p‘ tion (r2y4_1235) of 44 percent. "i efficient of partial determinatio variables was zero with the ex l and wheat acreage combined l: (r2y2_1345) accounted for about 1 5 unexplained variation accounted? iables other than X4, the number "i on feed. ~ U. S. Data. 1924-40 The price model was fitted during 1924-40 (Equation XI) 1nd the results were considerably e national basis. _ EQUATION x1 i.00014X1 ~ .00029X2 —~ .000017X3 * ,j(.00026) (.00044) (.0000071) 33X. .1772) "ables accounted for 71 percent Z ‘ation, with the total amount of s-corn, barley and oats—being _ nt variable on a national basis. grains were followed in impor- tio of corn prices to sorghum verified by r2y3_1245 and accounted f the unexplained variation. The rtial determination for the X5 ) accounted for about 19 percent. ‘ prices to sorghum grain prices independent variable to allow for f corn prices on sorghum grain test improvement in accounting n, in sorghum grain prices was f- the production of the sorghum 1X1) was deleted from the equa- c from Equation XI during 1924- tion of the composite national ihum grain relative to all other therefore corn prices Would con- the national basis to governing prices. _ 7 ‘1- EQUATION x11 0i.000028X1 -* .00077X2 — .000029X3 —- .(.000045) (00045) (0000095) 7X. _ _f .2512) When Equation XII was fitted to U. S. data for 1946-57 (With the same variables that are in Equations IX through XI), the picture was com- pletely changed, in that variable X2, X3, X4 and X5 all contributed to the unexplained variation in the price. The total feed grain variable, X3, accounted for most of the unexplained variation. Total feed grains were followed in importance by the ratio of corn prices to sorghum prices, combined cotton and wheat acreage and the num- ber of animal units on feed. Sorghum grain pro- duction (X1) accounted for none of the unex- plained variation. All five variables accounted for 77 percent of the variation in sorghum grain prices in Equation XII, but with production left out of the equation the amount of the variation accounted for Was improved slightly and amounted to 79 percent. These results were reasonable expectations in the light of a rapidly expanding postwar econ- omy, guaranteed support prices (eliminating the significance of prior years’ supplies), technolog- ical innovations, relatively high free market prices and the shift and reallocation of alter- native land resources. ACKNOWLEDGMENTS The present study was made under the Texas Agricultural Experiment Station’s State con- tributing project to the Southern Regional Price Policy Project SM-14, “The Effect of Price Sup- ports, Acreage Adjustments and Surplus Re- moval Programs On the Marketing of Southern Agricultural Products.” l3 [Blank Page in Original Bulletin] [Blank Page in Original Bulletin] OPERATION i sum STATION Q nu sunsnrxous i "mas new LAIOILATOIIES A coorrurmo STATIONS Location oi field research units of the Texas Agricultural Experiment Station and cooperating agencies IN THE MAIN STATION, with headquarters at College Static “ matter departments, 2 service departments, 3 regulatory _ administrative staff. Located out in the major agricultural : " 21 substations and 9 field laboratories. In addition, there arv stations owned by other agencies. Cooperating agencies i, orest Service, Game and Fish Commission of Texas, Te U. S. Department of Agriculture, University of Texas, Te l College, Texas College of Arts and Industries and the Kin experiments are conducted on farms and ranches and in ru ORGANIZATION THE TEXAS STATION is conducting about 4-00 active research i in 25 programs, which include all phases of agriculture in: these are: Conservation and improvement of soil Beef cattle Conservation and use of water Grasses and legumes Grain crops Cotton and other fiber crops Vegetable crops Citrus and other subtropical fruits Fruits and nuts Oil seed crops Ornamental plants Brush and weeds Insects Two additional programs are maintenance and upkeep, an Research results are carried to Texas farmers, ranchmen and homemakers by county agents and specialists of the Texas Agricultural Ex- tension Service joclay ,5 Wedearcé .96 jomorrowb r0 State-wide Rese» ~15 ‘k The Texas Agricultural Experime is the public agricultural resear l oi the State of Texas, and is parts oi the A&M College of Tex Dairy cattle Sheep and goataf Swine l Chickens and t t Animal diseases “ Fish and game _ Farm and ranch Farm and ranch y Marketing agric Rural home ecoi Rural agricultur Plant diseases ' AGRICULTURAL RESEARCH seeks the A WHYS. the WHENS. the WHERES and - . hundreds of problems which coniront farms and ranches, and the many indul ‘ ing on or serving agriculture. Workers Station and the field units of the Texas n» Experiment Station seek diligently to find l these problems. j