630.7 I (fib no. 681 cop. 8 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN AGRICULTURE 5, Pi AGRICULIUUE SELECTION OF FRESHE ATES THAT WILL MAXIMIZE DAIRY PROFITS An Application of Linear Programming Using DHIA Data From Northern Illinois K. R. Tefertiller and C. B. Baker Bulletin 681 UNIVERSITY OF ILLINOIS AGRICULTURAL EXPERIMENT STATION CONTENTS Review of Literature 3 Production Conditions 4 Market Conditions 12 Plans for Maximum Profits 16 Varied Labor Supply 16 Varied Seasonal Price Movement 20 Optimum and Actual Freshening Dates 22 Conclusions 23 Literature Cited 24 Appendix A: Roughage Requirements of Dairy Cows 25 Appendix B: A Linear Programming Model for a Two-Price Milk Market . . .30 TO MILES ECONOMIC AREA 2 ECONOMIC AREA C COUNTIES OF OHIA RECORDS (Left) location of Illinois economic areas 2 and C and of DHIA records from which lactation curves were derived; (right) Chicago milk market. (Fig. 1) K. R. TEFERTILLER, formerly a Fellow in the Department of Agricultural Economics and now Assistant Professor of Agricultural Economics, Texas Agricultural and Mechanical College; and C. B. BAKER, Professor of Agricultural Economics, University of Illinois. Urbana, Illinois January, 1962 Publications in the Bulletin series report the results of investigations made or sponsored by the Experiment Station DAIRY HERD MANAGEMENT is affected by numerous seasonal variations quantity and quality of milk per cow, quantity and quality of feed supplied by the farm, labor available for the farm as a whole and for the dairy enterprise, and prices for milk and milk prod- ucts, as well as resources bought for the dairy enterprise. The results of a study to select freshening dates that maximize annual income from a dairy herd are reported here. 1 The results apply most specifically to a herd in the Chicago milkshed (Fig. 1). They would differ little elsewhere, however, in the presence of similar sea- sonal price movements, provided that production conditions were not too different from those in the Chicago milkshed. These conditions, with details of method and findings, are described later. REVIEW OF LITERATURE In an early study of 10,870 cows, over the period 1910-1920, Mc- Powell (7) J found that cows freshened in fall months produced the highest average annual milk output. Lowest production came from cows freshened in spring months. Neither breed nor age was signifi- cantly related to the influence of freshening dates on milk production. Similar results were found by Morrow, Keener, and Hall (8) in a study of dairy herd management in New Hampshire, and by Dow, Johnson, and Parsons (2), a regional group studying dairying in a 12-state area in the northeast, including New York. Production varia- tion by season of calving was ascribed generally by these investigators to climate, feeding rates, management practices, and other related fac- tors. This regional group found that older cows showed a greater out- put response to month of calving than did younger cows. From 15,442 records of cows in Dairy Herd Improvement Associa- tions, T. E. Woodward (9) found November fresheners to yield the highest total production; July fresheners, the lowest. Highest peaks of production were exhibited by April fresheners. High points on their lactation curves coincided with the lush pasture season of May and June. Cows calving in August produced the lowest peaks of production. These seasonal variations accord well with the findings of Baum, Mey, and Shaw (1) who fitted monthly lactation curves to milk- production data from 244 individual yearly records of cows in western 1 Selection was made with use of linear programming, a mathematical tech- nique that assures a maximum net income, given stated conditions. (For further discussion, see Appendix B, p. 30). 5 Numbers in parentheses refer to Literature Cited. 4 BULLETIN No. 681 [January, Washington. The Washington investigators, however, made no attempt to relate total volume of milk to freshening dates. Fat content of milk is affected by temperature and stage of lacta- tion. Woodward (9) found fat content to be lowest in the third month, and to increase thereafter with a more rapid rise toward the end of the lactation period. He concluded that fat content is related inversely to milk quantity except after peak production, when both decline for a short period. Fat content tends to be lowest during the hot months of summer and highest in the cool months of fall and winter. Baum, Mey, and Shaw ( 1 ) found that in western Washington cows freshened in August and through September were cheapest to feed. Cows freshened in January and through March tended to be the most expensive. These results originate from seasonal variation in feed prices and in farm supplies of feed, especially pasture. Cows in lush pasture months (freshened the preceding July and August) convert pasture feed into body maintenance and weight gain. On the other hand, pasture feed is converted into peak period milk production for cows freshened in late winter. However, grain prices are lowest in early fall months, benefiting the fall- freshened cows, then in peak production. Cows freshened in March and April, on the other hand, required the most grain when grain prices were highest. Similar results are given in studies by McPowell (7), Dow, Johnson, and Parsons (2), and by Morrow, Keener, and Hall (8). PRODUCTION CONDITIONS This study is based on 42,747 monthly records of cows in all- Holstein herds in Ogle, Whiteside, Will, and DeKalb counties, in northern Illinois. The records were taken in DHIA (Dairy Herd Im- provement Associations) between January, 1955, and August, 1957. The lactation period is assumed to continue for 305 days from the middle of the month shown as "month of freshening" (Table 1). The records accord with the findings of others in indicating that the length of lactation period is largely independent of month of calv- ing. The records indicate that December fresheners produced the high- est annual output of milk, an average of 11,905 pounds per cow. August fresheners produced the lowest annual output, 10,450 pounds. When monthly production is expressed as a percent of annual produc- tion, the highest peak, 15 percent, is yielded by May fresheners in June; the lowest peak, 12 percent, by January fresheners in April and May. The fact that the cows were in DHIA creates selectivity in the ob- servations with respect to numerous management factors. The annual production reported in Table 1 far exceeds the average for the state as 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS tf -H ^J< r)< t>- S i Tj Tf Ov fO O O O PO r^ r-i *- o oo > ro t^ ^ c 'M g ^ T* rj< 10 oo -H r^ 10 -i vO O OO ro t-. vO fS ro P^ f*> 10 1~- -> ) -c -H o OO r^ PO l-~ 10 ^ d DeKalb O C . 3 1-' 3 P 3 -if'5t^f v l^O>O'*5OO*~-^J < >O i CS Tj< IO ** PO CS <>! tS O (N t ro CS -" O O O\ t~ t*5 fr ^ ^ ro 1 u, uy iviumii ui ficsiic it Associations, Northel 00 u Q. JJ bib < ^x 3 t^ P> r>J t-~ -- Ov ^O O * *O f> fO VO * 10 oo rs rs <5 fs -< T* QV s O i O Ov Ov Ov t^ fO ^- * tS N OO 4 ,-H v-l ^H ^H O OO tS OO O ** ON O N iOt~ OO O CS Tj< OO CN l Tj< O\ -i OO f) OO O O O\ OO OO OO PO *O T< ts ^-i O ^J 1 *O Q ^J< " OO O "5 PO t PC O a T iai en i 'o j= 0) 3 t^ t>- i \r> o *^- PO ^ ^O ro t- c^ PO oo o\ ONOOX--PO OO >O O c3 '-' O ON PO rovement H H^ 3fe e o S >> rt E 00 Ov 'O O\ 10 ^H \O IO T}O O O\ oot^-rc oo vO o rr> ra O O\ O\ O B u a ffi 2b 5 a < oo PO o -" PO PO * ^t - 1 PO -< t~- oo P'i Tj< \O O> -l CS OO '-I ^ O> -H -l t^po t^ \o o ^ CN es o oo oo oo in Dairy freshening H - 4> 3 x X*- 3.S u rt "5 PO O f^ P> *O O O OO *-< -H ,-c 10 o\ N OO \O CN t^ O\ ^O vO ON ^t 1 ON O PO t-- 10 vO rj< CS -i-i O OO 1^ vO ^ *- fc =J3 II D J M U JK _Q 0) U< O'f ^-cs^Tioa>o c^-t TJ< o\ VO PT> O vO O OO >O -- tS O NO r^ O t-- O ON ^* NO ON VO >O PO ON t IO ON ON t~ t- U) Tf PO PO P>> O ON OO O CS \O E O O u u -S en 3 - C ra - . ^ ^ c txj > s s K cvjo. 8 CE< c -i e S'S > ? "- 0-2 __ > X C4S JC 4-> 1 * ', c QS s 1 e'J3l3^Src s * > ^Q.>y J 3&a3$$MSt$i& J 6 BULLETIN No. 681 [January, a whole and for that of the area included in the study. Yet much evi- dence suggests that the response to most management factors, such as improved feeding, is more in the level of annual production than in difference in seasonal distribution of milk production. However, selec- tivity, originating in the source of the data, must be kept in mind in considering the results of this study. Records in this study indicate variations in fat content similar to Woodward's findings (Table 2). They indicate a tendency toward a seasonal decline in fat content associated with the high temperatures of summer months. Indeed, this effect seems strong enough partly to offset the influence of stage of lacation on fat content. The study relates to an owner-operated farm within a 55- to 70-mile radius of the Chicago metropolitan milk market, in Illinois Eco- nomic Areas 2 and C, the latter being the Chicago metropolitan area (Fig. 1). Dairy farms are the most frequent type of commercial farms in the areas, most farmers selling Grade A milk to the Chicago market. To represent a dairy farm typical of the area, 168 acres of farm- land 1 were assumed used to produce 40 acres of corn for grain, 10 acres of corn for silage, 31 acres of oats, 32 acres of alfalfa hay, and 39 acres of pasture. Crop yields are based on averages for dairy farms in the areas. Since dairying is the only livestock enterprise, the entire crop output is available to support it. Based on Census estimates for 1954, 2 the farm family was assumed to supply a total of 350 hours of man-equivalent labor per month from October through May; 450 hours from June through August; and 400 in September. 3 Deducting the labor required for nondairy enterprises left for dairying the labor supplies reported in Table 3 (p. 8). Total labor required per cow was fixed at about 110 hours per year, distributed differently over the year, according to month of freshening. The largest use of labor occurs in the month of freshening and during the winter months. Thus, cows freshened from November through March require the largest amounts of labor in the month of freshening. It seems likely that the total, as well as distribution, of the annual labor required would vary by month of freshening, but data available did not provide a basis for making any distinctions. (For monthly labor re- quired by month of freshening, see Table 4, p. 9.) 1 Arithmetic mean for farms in Illinois State Economic Areas 2 and C, as reported by the U. S. Census of Agriculture, 1954. "Source: U. S. Census of Agriculture, 1954. * A survey made in 1958 in essentially the same area revealed that when hired labor is added to family labor, the total available on similar farms ranges from 365 hours per month in November and January through March, and from between 453 to 456 from June through August. For further information, see M. R. Langham and C. B. Baker, Optimum Plans for Farms in Northeastern Illinois; 111. Agr. Exp. Sta. AERR-40. 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS V c 'v a I v z: (A W 4) r^ VM ^rt I s O '^ -M 13 .s 'o c V u M u 1 II "-" oo T*< cr> oo I if) -* \o -i O to fs rh O oo O oo to to to to to to to to to to to to 10 oo "* Ov to ir> r-J TJ< O O to ir> to to to to to to to fO to to to to to to fO to to to to to to to fO fO ^ p\ oo * i oo p >r to to to to to to to to to to fO to .-H ,-< r^ -H <-i O to to to to fO fO to to to to to to to Tj- t^ t^ t~ ro O\ Tf Ov 00 1 o Q > O O - 1 O ^ -~ *< t^ 00 VO t* *5 O\ % s o V o Z O - O O t~- t t- J~- \O IO <^> O Ov ^H vO >O o o -i oo W c 1 $ ti OO O Ot^-t^ vo Tf i OOO ~H O\ ^^,^,^H ^H ^H^ O vO VO vO CS t~ -H 10 O *3 >O -< vO OO 1 S fH tJC .S 'S 3 < >> OO OO - vOTfi -> 0000 -^ O O\ vO vO vO fS ^-i O O t^- PO O vO vo OO o O U O O O O vO Tf ^Ht OOO\O O Os vO vO vO vo O O ** I s - OO ir> VO vO OO o VM O U S 3 > > O O O OO rj< u-~ t- oo OO O O ON * - "" '-' '-' '-' O vO vo OO O*^^ tN fO >O vo vo 00 h & 1 S u.' OO ON t- -H*~I^ 00 00 00 O O Ov vO O ^C >O t* t~ vo vo OO '1 V K a < U.' O O\t^- POt^t^OOt^ OOOOOO Os 1 1 Ti 1 1 O O^*"O\t v lt~fSt^.t^.r'5>OvOvO OO .8 J > S JD OS 1-- r*5 O t^ OO * *>- OO OO O O OS -!- ^-H T-l O Tj< OS vO CS (S t> t- t^ PO >O vO O OO 2 S U fe c t-~ ro O O OO t^ t^. t~. OO OO O O\ OS OS vo vo t^ l- t~- t- t-~ rc if> O ** 00 <4H O S rt i i "5 O O O t~ l~- * t^- OO OO OS t OS -i "- "- 1 > O M ^_ C3 *- ) n U c* o .c c : : c C^ti'g.+J&W fl C -rr Cwn^ t "^*?5^ i TiOflJ ^ ^^^A^<^ozQ 10 BULLETIN No. 681 [January, I I HH O "w ' M O I 2 . I! . o in infMiCNoir>cso -- _i rn -! CS CO CO CO O\ fO rr> (O <*5 co O co Q O ^ ''Ov^CNlONOOOOONON ^H ^ ^H pqco CS co CS CS 5 (V| 1-1 O O OO ON t~ co co OO co ON 00 *> >O 00 00 ON ON NO t t^- * ^4 ^H i- CN CN CS CN CS CS CS CS SO 00 >O co '-H O NO ON ON NO CN CS n in ro oo NO t- 1~ TJ< -* cs ^ 10 ,_! ^H i-l CO CS CS CN CS CS CS CS CS OO Ov <*5 -^ NO CS ir) vO CS O ON ON CS ^ in CS 10 OO t- if) NO -I CO CO CO CS CS CS CS CS CS *-l *H O Be I O V ** ZG u H _. o ~ 3 5 o * iE be O2 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 1 1 I I .B W) % O H |.s II s > fa o 4) *H & ^3 s ^ O *4 O O M I 10 U t/3 O) S S 3-S e e n o 12 BULLETIN No. 681 [January, MARKET CONDITIONS To encourage milk production seasonally consistent with demand, a "base-excess" plan was introduced for milk in the Chicago market in 1954. 1 The months of September through November are termed "base periods"; March through June "excess periods." During the excess period, each producer receives each month two prices for his milk. He receives a "base price" for the quantity that does not exceed his average monthly production in the preceding base period. He re- ceives a "surplus price" for the quantity in excess of this base quantity. Historically, the surplus price has been less than the base price by 40 cents per 100 pounds. 2 Since monthly production is known for cows freshening in each calendar month (Table 1), quantities of milk to be sold at the "base" and "surplus" prices can easily be computed for any combination of freshening dates. For example, a cow freshening in May establishes a fall base equal to 9.56 percent of her annual production and 3 percent of the annual production was produced in March, none in April, 7.06 in May, and 14.39 in June. Hence, if all the herd had been freshened in May, 4.83 percent of the annual production would be sold at "sur- plus price" during June as the "fall base" is less than the monthly production in June. But the entire amount of milk produced in March, April, and May would be sold at the "base price" as the "fall base" is greater than monthly production during these months. The price paid for milk is modified further by three additional factors. 3 The first is a price premium negotiated for Class I milk sold each calendar month. The premium applies only to nonexcess milk. Hence, the premium tends to accentuate the difference between base and surplus prices. The second factor is locational. Two cents per 100 pounds are deducted from the uniform price of milk for each 15-mile- wide zone outside a 55-mile zone. Within the 55-mile zone, producers receive a small premium above the uniform price. The third factor is a butterfat differential. Between 1955 and 1958, the butterfat differ- ential varied from 6.9 and 7.2 cents for each tenth of a pound above or below 3.5 pounds per 100 pounds of milk. The "base price," adjusted for butterfat test, is given for a cow freshened for each month (Table 7). The quantity-weighted mean 1 Federal Milk Order No. 41 Reporter, "Grade A uniform price for base milk," January, 1959. 1 Had the negotiated price premium been taken into account, the difference between "base" and "excess" price would have been 65 cents per 100 pounds. When this differential was used in solving for optimal freshening dates, no differ- ence of substance was found in the freshening pattern. 1 A fourth factor is a premium paid for bulk-handled milk. However, this factor in no way influences seasonal variation of price. 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 13 t: $ 3 oo 4> U1 3 O\ ~ r iaS ^H ll 3 ** , o 3 c3 su H G I- to w IS S2 Ml |a 8VM o V p S'o "1 vo O * Ov 1-1 OX Ov Ov O CO *O vO VO O ON OO OO vO OO O Ox t u^cocovOt^OOONOO CO CO CO CO CO PO CO CO PO PO CO t^ TJ< 00 00 O CO * vo CS vo -J< O^t-u^-^Tt^vot^OxOOv ro CO co ro ro co CO co CO ^ fO *^< CO CO CO CO CO CO CO CO CO ^* o e V *** fctx J-QIO o S x u O u OE its f-c ~ r 7 ' " 14 BULLETIN No. 681 [January, base price for the lactation output of the cow is shown across the bot- tom of the table. The price so reported for March through June would apply only to the quantity of milk, each month, equal to or less than average production in the previous months of September through November. Quantities above these figures are actually priced at 40 cents per 100 pounds below the figures shown in the table. 1 Farmers actually receive for milk a "blend" price, a price averaged for all classes of milk, for the eight months excluding March through June. The prices reported in Table 7 are comprised only of those ele- ments of price that vary by month of sale. Thus the income levels resulting from problem solutions reported below are different from actual experience. However, the seasonal distribution of freshening dates are in no way affected by raising or lowering milk prices. Hence, these "partial" prices were used for computational simplicity. Besides milk, the dairy cow produces meat in the form of veal calf and cull-cow beef on herd replacement. Compared with the value of milk, these by-products add little to the net value of output. Annual production per cow is taken as 342 pounds, 2 60 pounds, 3 120 pounds 4 per cow, respectively, for cull-cow beef, veal beef, and replacement heifer, regardless of month of freshening. There appears to be little reason for assuming that the price varies with respect to month of freshening for herd replacement or for veal-calf beef. Utility grade heifers averaged $13 per hundred pounds live weight (4) and veal calves, $13. 50. 5 Cull-cow beef is valued at prices two months prior to freshening month, averaged over the period, 1953-1957 (4). A summary of prices used for products of dairy cows is given in Table 8. For each product, the price is a quantity-weighted estimate for cows freshened in the month indicated. In the last row, the value of total cow production per 100 pounds of milk is shown. In the actual computation, however, these prices varied according to the amount of milk sold at excess prices. This quantity is influenced by the freshening pattern of the entire herd. 1 Federal Milk Order No. 41 Reporter, "Grade A Uniform Price for Base Milk"; January, 1959. 2 Average weight of a Holstein cow at replacement age was 1,489 pounds (6). Replacement was estimated at 23 percent per year (based on unpublished data of R. W. Touchberry, Department of Dairy Science). ' Bull calves are ordinarily sold for veal from dairy farms in northern Illi- nois (based on observations from Farm Bureau Farm Management Service Records, 1957-1958). 4 About half the replacement heifers are sold as culls before first freshening. The estimate includes total pounds of cull heifer sold from birth to freshening date. ' Farm Bureau Farm Management Service Records, 1957-58. J962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 15 I bfl 1 S ta **-! O c ^3 sl 13 T3 *O CA 0) 00 JO 2 sS u, oooo o <#> ^ -HQ O t >o O fJ er> O <^5 f> > I. o o w_^5 Os s "a = a s . s -Sl'5 3 . w "^ o g ^ ee-a ^-. O |! |2 S c (0 o 2 B a .2 n S ^| ^Sg> 13 g.S i g m ; s s i-i ?^ ,rt rt ^^li^^s " *^ E *^* P ' 3 .--i o i> *-. s/ o.rt o u ^ rt *- vt r *e 16 BULLETIN No. 681 [January, PLANS FOR MAXIMUM PROFITS All or any part of the herd can be freshened in any month. The number of cows milked and the distribution of freshening dates are assumed limited only by the feed and labor supply as already described. Any hay beyond requirements of the dairy herd is assumed to be sold at $21.20 per ton; any grain at $2.35 per 100 pounds of corn equivalent. The method used to select a profit-maximizing plan yields estimates on value of adding a unit of any resource that limits the number of cows to be freshened. The usefulness of this added information will be demonstrated later. Varied Labor Supply Freshening dates that maximize returns were selected under two conditions: (1) that the operator hires no labor to supplement the supply furnished by the farm family; and (2) that labor is hired to the extent found profitable. Labor was priced alternately at two wage rates, 80 cents and $1.70 per hour. 1 Results of the three sets of conditions are given in Tables 9 and 10 under headings of Plan I (no labor hired), Plan II (labor hired at 80 cents an hour), and Plan III (labor hired at $1.70 an hour). 'Other plans were estimated for higher wage rates. Until wages reached $3.60 an hour, however, no substantial change was observed in freshening patterns that maximize returns. PLAN I MILK PRODUCED (THOUSANDS OF POUNDS) NUMBER OF COWS FRESHENED I I I I I Optimum distribution of milk production; no hired labor. (Fig. 2) M J J A MONTH OF PRODUCTION 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 17 Plan I With no supplemental labor, the maximum-profit plan is domi- nantly one of fall freshening, 27 out of 30 cows (Fig. 2). A substan- tial volume of annual milk production, over 10 percent, is sold at surplus price. Hay produced on the farm is practically used up in this plan, though some pasture is in excess during June and July (enough perhaps for three more cows in these months). An appreciable amount of grain, about 1,000 bushels of corn or corn equivalent, is available for sale. Slightly less than 20 percent of family labor available for dairying is unused for dairying in June; slightly more than 40 percent in August. The plans differ little in income generated from the dairy enterprise. Income is based on milk prices that exclude all but seasonally variable components. Plan II When labor is hired at 80 cents an hour, the freshening pattern (Fig. 3) shifts from a dominantly fall freshening program to one of spring freshening. The consequence is to reduce milk sold at surplus to about &i/2 percent of the annual total. Little change occurs in the use or nonuse of family labor. The equivalent of about 3 man days are hired in July and about 4 in September and October. As additional labor becomes available during the summer months, the optimal- freshening program shifts to spring fresheners that use pasture as a Optimum distribution of milk production; labor hired at 80 cents per hour. (Fig. 3) M J J * MONTH OF PRODUCTION 18 BULLETIN No. 681 [January, Table 9. Profit-Maximizing Plans for Freshening Cows in All-Holstein Herds: 55- to 70-Mile Zone From Chicago Milk Market, at Selected Wage Rates for Hired Labor Plan I, Plan II, Plan III, Item family labor labor hired, 80 cents labor hired, 31.70 only per hour* per hour' Number of cows freshened in b : Aor. . . 3 28 26 T F June , 2 1 Aue. . , 6 Oct 8 Nov 13 Dec... 3 Products sold Milk at base price, pounds 304,763 Milk at surplus price, pounds 35,179 Grain, 100 Ib. of corn equivalent 638 Hay, ton 2. Unused resources Pasture, June and July, TON pounds. . . . 3,836 Family labor, man hours May June 85 Aug 168 Labor hired, man hours July Sept Oct Income, dollars' 13,849 318,774 29,711 694 11 72 179 27 42 37 14,445 316,141 33,816 685 8 74 179 27 43 14,379 m To limit found profitable. b Rounded to nearest unit. The solutions can and do indicate fractions of cows. For this reason the following milk production is not exactly the same as would be estimated by sum- ming products of cow numbers and the annual production indicated in Table 1. c Total annual income from the dairy enterprise with no deduction for the cost of re- sources with quantities fixed as shown in Tables 3 and 4. Table 10. Values of Added Units of Resources Found to Limit the Number of Cows Freshened in All-Holstein Herds : 55- to 70-Mile Zone From Chicago Milk Market, at Selected Wage Rates for Hired Labor Limiting resource Value of added unit under Plan I For description of plans, see Table 9. b To limit found profitable. Plan II b Plan III b Silage, dollars per ton .60 11.60 10.00 Hay, dollars per ton 34.00 32.00 Pasture, cents per pound, TDN May June .047 .057 July Sept. . .04 .065 .045 Family labor, dollars per man hour May 1 56 July . . 11 21 80 1.70 Sept.. 12 33 80 1 70 Oct 4 17 80 1 59 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 19 substantial part of their feed requirements during months of high milk production. Pasture unused in Plan I was used by the spring fresh- eners in Plan II. The value of increased corn sales in Plan II was greater than the loss in income from the decrease in hay sales. In sum- mary, additional labor in certain summer months allows spring fresh- eners to make up the optimal- freshening program and these fresheners use cheaper feeds than did those in Plan I. Plan III With a wage rate of $1.70 an hour, the freshening pattern becomes slightly less extreme than that in Plan II, though it is still dominantly a spring- freshening program. There is also virtually no change in total milk production (Fig. 4), though a slight increase did occur in the percentage sold at surplus. As between Plans II and III, the shift is primarily from fewer April fresheners to more December fresheners. Cows freshening in December produced the largest amounts of milk annually. December fresheners have a "second freshening" from lush spring pastures. These pastures are a cheap source of feed during months of relatively high milk production. When the price of labor was increased to $1.70 an hour, it became more profitable to shift to the high-producing December fresheners even though a larger quantity of milk had to be sold at surplus price. No labor is hired in October. The amount of labor hired in July and September remains at about 3 and 4 man days, respectively. Optimum distribution of milk production; labor hired at $1.70 per hour. (Fig. 4) M J J A MONTH OF PRODUCTION 20 BULLETIN No. 681 [January, Effect of adding one unit of resource For each plan, Table 10 shows: (1) resources that limit further expansion of the dairy enterprise, and (2) the value, in terms of in- come, of adding one more unit of the resource that thus limits the plan. In Plan I, although silage was limiting, the value of an added pound (60 cents per ton) was very little. But, an added pound of total digestible nutrients in the form of pasture during July through Sep- tember would have added 4 cents to income. An hour of labor added by the family would have added $1.56 to income. Values were still higher for July, $11.21; September, $12.33; and October, $4.17. The size of these values indicates that labor was the most restrictive re- source in Plan I. In reality, farmers add to their labor supply in seasons of peak requirements by working longer hours and by exchang- ing work with neighbors. The value of such measures is clearly indi- cated by these estimates of increases created in income. These figures must be used with considerable caution. The method of computation provides no basis for estimating the quantity of the resources for which these estimates would be valid. For labor, we know from the results of hiring it in Plans II and III that the quantity that would produce such high returns is quite small. In Plans II and III, hiring labor removes labor as a limiting resource and adds as limiting resources hay and pasture in May and June. In Plan II, an added ton of silage would be worth $34.00 and hay $11.60; an added pound of total digestible nutrient would be worth 4.7 cents in May and June and 6.5 cents in July and September. The value of an added man hour of labor from the family during July, September, and October is, of course, equal to the rate for hired labor. Since surplus family labor is found in May and June, a man hour of labor added in these months would have no value. In Plan III, the various types of roughage are again the only re- strictive resources, as labor may be hired for $1.70 per hour. As in Plan II, surplus family labor is found in May and June. But labor was hired only in July and September and the value of an added man hour of labor was equal to the rate for hired labor ($1.70). The value of an added man hour of labor in October was $1.59. Since the cost of an added man hour of labor was $1.70, no labor was hired in Oc- tober even though there was a labor shortage. Varied Seasonal Price Movement The seasonal price pattern for dairy-cow products can change for many reasons. The consequence of such a change was examined by arbitrarily assuming that cull-cow beef is priced at a constant over the 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 21 year at the average of prices given in Table 8, $10.47 per 100 pounds. The effect on total value of production per 100 pounds of milk is shown in Table 11. In general, prices are reduced for cows freshened in late winter, spring, and early summer and are increased for fall fresheners. Table 11. Value of Dairy Products per 100 Pounds of Milk Produced; by Month of Freshening, Under Alternate Assumptions on Price of Cull-Cow Beef Month of Price of cull cow beef freshening Varied Constant Change Tan. $3.66 #3.66 $ 00 Feb. 3.71 3.70 - 01 Mar. . ... 3 73 3 70 - 03 Aor. ... 3 73 3 70 - 03 May 3 . 74 3.71 -.03 June 3.72 3.70 - 02 July.. 3.73 3.73 - 00 Aug. . . 3.72 3.74 .02 Sept.. 3.71 3.74 03 Oct 3 . 66 3.68 02 Nov 3 63 3 67 04 Dec... 3 65 3 67 02 With the changed pattern of monthly prices, new solutions were found under conditions of: (1) no labor added to family-supplied labor; and (2) labor hired as long as profitable at 80 cents an hour. Under the first condition, the solution differed in no significant way from Plan I (Table 9). Operator's returns were increased slightly, but the freshening dates were unchanged as were the milk quantities Table 12. Profit-Maximizing Freshening Dates for Cows in All-Holstein Herds: 55- to 70-Mile Zone From Chicago Milk Market; Labor Hired at 80 Cents per Hour and Cull-Cow Beef Constant Over the Year Item Amount Item Amount Number of cows freshened in Unused family labor, Apr 23 man hours Dec 7 May 3 June. . 78 Products sold ^ j7 8 Milk at base price, pounds. . . . 312,459 Oct 5 Milk at surplus price, pounds. . 39,577 Grain, per 100 pounds of corn Labor hired, man hours equivalent 674 July 26 Sept 43 See footnote b in Table 9. 22 BULLETIN No. 681 [January, sold at base and surplus prices. Therefore the results are not reported in detail. Under condition 2 when labor is hired, however, the results (Table 12) are comparable with Plan II (Table 9). Five fewer cows are freshened in April. Instead of two cows freshened in June, seven are freshened in December. Milk sold at surplus price increased by about one-third. The grain surplus is about the same, as is the surplus of family labor. No hired labor is required in October and operator's returns are reduced only slightly. Optimum and Actual Freshening Dates The computed optima for freshening dates rest on a specifically defined production situation, as previously described. The sensitivity of the results to changes in labor supply suggests that the dates would differ substantially for differences in nondairy enterprises. For these and other reasons, it would be surprising to find much accord among dairy farmers generally in the area and especially over larger or dif- ferent areas. This would be true even though all dairy farmers so bred their cows as to attain profit-maximizing dates in their own specific production situations. Nevertheless it is of interest to compare results of this study with average freshening dates in somewhat similar herds in the area. Such comparison is possible by reference to the first five columns of Table 13. In the first four columns data are summarized from solutions pre- viously reported. In column five the percentages of cows freshened each month are given for DHIA herds of 30 cows or more in northern Illinois. Plan I, using nonsupplemented family labor, most closely approximates the actual situation. Any single plan could be expected to be far more extreme than an aggregate of plans for many herds. In Plan I, no cows are freshened in months reported low in DHIA herds, either all or only those of 30 cows or more. The fall freshenings of Plan I coincide roughly with the months of heavy freshening in all DHIA herds in Illinois. The same relation seems generally to hold for herds reported for Wisconsin, Michigan, and New York. The plans using hired labor, however, seem distinctly at odds with the freshening pattern in either northern Illinois or in the three-state area. In recent years, the seasonal variation in the Chicago area seems to have been reduced materially. Such an observation is consistent with findings in this study. Little response was found to the small change introduced in the seasonal price pattern for output of the dairy enter- prise. However, from varied labor conditions, wide differences were noted in the profit-maximizing freshening pattern, though with rela- tively little effect on income generated by the dairy enterprise. Thus 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 23 Table 13. Optimum Freshening Dates Compared With Dates Reported for Dairy Herds in Northern Illinois, Wisconsin, Michigan, and New York Percent of cows freshened Optimum plans Plan I Plan II, labor hired, Plan 1 1 A, labor hired, Of) ront < Plan III, labor hired, 1 70 rwr DHIA northern herds, Illinois' Wiscon- sin, family labor only 80 cents per hour; price of cull-cow beef constant per hour; price of cull-cow beef varied monthly hour; price of cull-cow beef varied monthly 30 cows or more All herds Michigan, and New York b Tan. . , 9 6 9 3 9 Feb 8.8 7 6 7 9 Mar 7.4 8 7 7 9 Apr 10.0 93.1 76.7 86 2 7.0 7 9 8 May 6 2 6 8 6 2 June 6.9 3 5 6 7 6 4 5 July.. 6 8 5 4 Aue. . . 20.0 8 4 7 5 6 1 Sept.. 10 4 11 9 1 Oct 26.7 10 4 10 3 10 6 Nov 43.3 9 6 11 10 9 Dec 23.3 10.3 8.7 8 9 9 7 Unpublished data from DHIA records for 12-month period ended Tune, 1959. "U. S. Dept. of Apr., Milk Production, March, 1960, for 1959; DHIA herds in Wiscon- sin, crop reporter data in Michigan, and a combination of DHIA and crop reporter data in New York. The authors are indebted for this reference to J. Russell Ives, Director, Depart- ment of Marketing, American Meat Institute, Chicago, Illinois. over a range of farms, the study results would lead one to expect to find widely varied freshening patterns. On each farm, moreover, the total supply of labor and the nondairy demands for its use appear to be important determinants of the freshening pattern. CONCLUSIONS When labor for dairying is limited from May to October, a fall- freshening program seems most profitable. Such a limitation might be found on farms with large labor demands from crop enterprises. Yet with sufficient labor or with labor hired up to $1.70 an hour, income can be increased by changing to a spring- freshening program. The relatively small change in income suggests that the freshening program might be better adjusted to the seasonal variation in labor supply rather than the reverse on farms where dairying is not the sole enterprise. These conclusions depend heavily on the production situation as- sumed for the dairy herd. For example, fall fresheners require most 24 BULLETIN No. 681 [January, feed during months of low feed prices. Hence, if all feed is purchased, fall fresheners would be favored relative to nonfall fresheners. In the study situation, however, all but protein supplement is supplied from the farm. Thus, no seasonal pricing was used for feeds and the feed supplies are expressed, except for pasture, as annual supplies. These conclusions are especially dependent on the amount of pasture avail- able, as no cost was assigned to pasture. Hence, a farmer with a large supply of pasture might differ from a farmer with a rather limited supply of pasture in a profit-maximizing seasonal freshening pattern. Milk prices so vary seasonally as to favor fall fresheners, produc- ing a relatively large proportion of annual milk output in the high-price fall and winter months. Yet after taking into account the seasonal variation in price of cull cows, inversely related to milk price, and adjustments for butterfat content, the actual seasonal variation in price of dairy-cow products was substantially reduced. In other production situations, especially in different culling programs, the results might have been different. The results seem to be fairly stable in the presence of changes, at least small changes, in the seasonal pattern of prices for products of dairy cows unless the labor supply also is changed. In this event, the profit-maximizing freshening dates can change markedly. LITERATURE CITED 1. BAUM, E. L., MEY, C. W., and SHAW, A. O. Relationships among fresh- ening date, feeding practices, production level, and profits in milk production. Wash. Agr. Exp. Sta. Tech. Bui. 4. 1952. 2. Dow, G. F., JOHNSON, S., and PARSONS, M. S. Factors affecting seasonal milk production and their effect on producers' cost and returns. Me. Agr. Exp. Sta. Bui. 459 (issued in cooperation with Northeast Experiment Stations). 1948. 3. ECKLES, C. H. Dairy cattle and milk production. 5th ed., pp. 378-379. Macmillan Co. New York. 1956. 4. Livestock and Meat Statistics. U. S. Dept. Agr., Agr. Mktg. Serv., [Dept. Agr.] Stat. Bui. 230. p. 227. 1958. 5. LOOSLI, J. K., HUFFMAN, C. F., PETERSEN, W. E., and PHILLIPS, P. H. Nutrient requirements of domestic animals: No. Ill, nutrient re- quirements of dairy cattle. National Research Council, Pub. 464 (rev.). 1956. 6. MATTHEWS, C. A., and FOHRMAN, M. H. Beltsville growth standards for Holstein cattle. U. S. Dept. Agr. Tech. Bui. 1099. 1954. 7. McPowELL, J. D. Influence of season of freshening on production and income from dairy cows. U. S. Dept. Agr. Bui. 1071. 1922. 8. MORROW, K. S., KEENER, H. A., and HALL, C. N. Analysis of certain factors involved in dairy herd management in New Hampshire. N. H. Agr. Exp. Sta. Tech. Bui. 86. 1945. 9. WOODWARD, T. E. Some studies of lactation records. Jour. Dairy Sci. 28:209-218. 1945. 1962] SELECTION OF FRESHENING DATES THAT WILL MAXIMIZE DAIRY PROFITS 25 APPENDIX A: ROUGHAGE REQUIREMENTS OF DAIRY COWS To estimate roughage required monthly for cows freshened in each month, the following data were available : ( 1 ) milk produced per cow during each of the first 10 months after freshening; (2) average body weight of cows; and (3) grain consumed during each of the first 12 months after freshening. From (1) and (2) total-digestible-nutrient requirements were estimated. An example of this computation is given in Table 14 for a cow freshened on February 15. The DHIA records furnished evidence on grain consumed per cow per day. Mean estimates and standard deviations about means are given in Table 15. From these data, it is a simple extension to compute the monthly grain consumed per cow. The amount of feed required from nongrain sources is the differ- ence between total feed required and the amount supplied by grain. As an example of this computation, the roughage required for cows freshened on February 15 is shown in Table 16. The nongrain feed requirement is supplied by roughage pasture, hay, silage, or hay and silage. Pasture was assumed to be used in season to the limit supplied by the farm or required by the herd, which- ever limit is reached first. The total digestible nutrients furnished daily per cow from pasture are given in Table 17. The remaining roughage requirements must come from silage or hay, or both. Estimates of hay and silage required to balance the ration of a cow freshened on February 15 are given in Table 18. Require- ments of hay and silage for cows freshened in other months were esti- mated similarly and are reported in Tables 19 and 20, respectively. To accord more closely with farm- feeding practice, both hay and silage were allowed to replace pasture as needed, if the value of either exceeded in feeding its value if sold ($21.20 per ton for hay; $8.50 per ton for silage). Unused pasture was assumed to have a value of zero. In the programming model, the system of transfers necessary to effect the required substitutions was as follows: Value Initial if sold, Substitute for pasture, pounds supply, cents per Resource pounds pound Hay Hay and silage Hay 160,000 .0106 1 1 Silage 336,000 .0043 Pasture May-June 24,500 July-September 21,000 When one pound of hay was substituted for pasture, the supply of 26 BULLETIN No. 681 [January, Table 14. Pounds of Total Digestible Nutrients Required per Cow per Day Not Furnished by Concentrates for Cows Freshening on February 15; 305-Day Lactation Period Average Average daily Average Daily 1 requirec PDN 1 for 8 Total Month of feeding body weight" milk produc- tion per cow 1 ' butter- fat test b Milk i Body 1 main- tenance o - o\ o -H o Ov oo T}< CN oo "^ "-i 1-1 o o - ts * > oo -