Potential for Economic Benefits to the Producer from Altering the Composition of Milk

Potential for economic benefits to the producer from altering the composition of milk

D. J. Garrick and N. Lopez-Villalobos

Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Palmerston North, New Zealand

Abstract

Cost-price models were developed to describe milk collection, manufacture and marketing of standardised fluid milk, butter, cheese, casein, and milk powders. Market constraints were modelled by fixing fluid milk demand to 10% or 70% of milk production. Milks representative of New Zealand Holstein-Friesian (HF) and Jersey (J) breeds, and novel technologies were considered. The true value of each milk was assessed from its own processing performance on the basis of fat, protein, lactose and volume considerations. Average milk was worth £0.193/kg when a significant fluid market exists, reducing to £0.112/kg when most milk was manufactured into concentrated dairy products for sale on the world market. Milk from different breeds varied in true value. On a per kilogram basis, HF milk was less valuable than J milk.

Single and multiple component payment systems were quantified for various subsets of milk components and used to obtain predicted values of a range of milks for comparison to their true values. Values of milks that differ in composition from average milk tend to have predicted values that deviate from their true value. The extent of such bias varies depending upon the payment system considered. For example, volume-based payment over-valued HF milk and penalised J milk. Other payment systems undervalued HF milk and over-valued J milk.

Payment systems should be fair, discourage unfavourable changes in composition and provide opportunities for shifts towards the production of more valuable milk. The marketing mix and the choice of payment system have major impact on the potential for economic benefits to the producer from modifying the composition of milk. Payment systems need careful, thorough investigation in concert with market research and studies into breeding and other management opportunities for modifying milk composition.

A value-based payment system can encourage producers to alter the composition of their milk in order to increase revenue.

Keywords. dairy cattle, economic, payment systems

Introduction

The erosion of prices for commodity dairy products over several decades has led to continuous reduction in milk payment in many countries (International Dairy Federation, 1998). Environmental concerns and other legislative factors that have developed over this time frame have prevented reductions in farm costs, despite some technological developments. Producers have responded to such a cost-price squeeze in several ways. These include leaving the industry, accepting a reduced standard of living or making a conscious effort to improve their farm business. Options available for producers to improve their business varies with particular circumstances, but may include: increasing the number of cows; increasing the productivity per cow; decreasing costs per unit of milk; or increasing the value of the milk.

The value of milk to the industry depends upon its composition and the aggregate profit of the product mix manufactured from the milk. The value of the milk to the producer depends upon the value of the milk to the industry, and the payment system used to reward producers. A value-based payment system can encourage producers to alter the composition of their milk in order to increase revenue. This paper investigates the value of milk, explores some milk payment systems and considers possible economic benefits to the producer for altering milk composition.

The composition of milk and yield of dairy products

A model was developed to calculate the yields of dairy products that could be obtained from milk of given composition. Milk was described in terms of its concentrations of fat, protein, lactose and minerals. In this paper, concentrations of casein, lactose and minerals were considered constant at 77 g/100 g of protein, 4.7 g/100 g of milk and 0.7 g/100 g of milk respectively. New Zealand average milk was processed using this model, as were milks representative of particular breeds or breed crosses (Livestock Improvement, 1998) and some futuristic novel milks, high in solids or protein. The fat and protein concentrations of these milks are shown in Table 1.

Table 1 Concentrations (g/100 g) of fat and protein in milks evaluated†

Component
Milk / Fat / Protein
Average New Zealand / 4.68 / 3.53
Holstein / 3.50 / 3.20
Holstein-Friesian / 4.39 / 3.41
Jersey / 5.77 / 4.01
Ayrshire / 4.36 / 3.50
Crossbred Holstein-Friesian  Jersey / 4.79 / 3.59
High solids / 7.02 / 5.30
High protein / 4.68 / 5.30

† The concentrations of casein, lactose and minerals were 77 g/100 g of protein, 4.7 g/100 g of milk and 0.7 g /100 g of milk, respectively.

The processing of the milk was considered to start with cooling and storing of the milk on the farm. The milk was transported a given distance to be received at the dairy factory. The milk was separated into cream and skim milk. The manufacture of 1000 kg of average milk resulted in 117 kg cream and 883 kg skim milk. The next processing step varied according to the product being manufactured, but typically involved ultrafiltration of skim milk into lactose-rich permeate and protein-rich retentate. Skim milk obtained from national average milk produced 652 kg permeate and 231 kg retentate. The three processing intermediaries (cream, permeate and retentate) were then recombined in varying proportions to ensure that the compositions of the final dairy products were standardised in terms of fat and protein. The products manufactured included standardised fluid milk, butter, cheese, casein, and whole milk powder (WMP), skim milk powder (SMP), whey powder (WP) and butter milk powder (BMP). The standard compositions of the most important of these products are in Table 2.

Table 2 Standard composition (g/100 g) of dairy products†

Fluid milk / WMP / SMP / Cheese‡ / Butter / Casein
Fat / 3.50 / 26.50 / 0 / 35.00 / 84.00 / 0
Protein / 3.20 / 25.10 / 3.71 / 24.50 / 0.59 / 89.00
Lactose / 4.78 / 39.80 / 4.93 / 1.39 / 0.79 / 0.56
Minerals / 0.71 / 5.90 / 0.73 / 3.85 / 0.12 / 0.08
Water / 87.82 / 2.70 / 90.63 / 35.26 / 14.50 / 10.35

† WMP = whole milk powder and SMP = skim milk powder.

‡ Includes 1.7 g of salt added.

In the manufacturing process, mass balances were maintained such that all milk components were represented in the final products or, in the case of water, lost during processing. Five discrete channels were available for processing milk. These include standardised fluid milk, WMP, SMP, cheese and casein/butter. The processing of standardised fluid milk typically resulted in surplus fat and protein. Accordingly, butter, butter milk powder, casein and whey powder were by-products of fluid milk manufacture. The processing of whole milk and skim milk powders resulted in surplus fat and protein manufactured into butter, butter milk powder, casein and whey powder. Cheese processing was typically limited by availability of protein and lead to surplus fat and whey, that was manufactured into butter, butter milk powder and whey powder. The processing of casein utilised all the protein, with surplus fat and whey manufactured into butter, butter milk powder and whey powder. Some example product mixes that can be manufactured in a single channel from 1000 kg national average milk are in Table 3. The model allowed for milk to be apportioned in a mix among these five channels if desired.

Table 3 Yields of dairy products (kg) in mass balance from 1000 kg average New Zealand milk in five different processing channels

Processing Channel
Product† / Fluid Milk / WMP / SMP / Cheese / Casein/butter
Fluid Milk / 927.0 / - / - / - / -
WMP / - / 108.9 / - / - / -
SMP / - / - / 76.0 / - / -
Cheese / - / - / - / 111.1 / -
WP / 3.1 / 4.4 / 4.9 / 62.1 / 59.0
Casein / 4.2 / 6.1 / 6.7 / - / 28.3
Butter / 17.1 / 21.3 / 54.9 / 5.9 / 55.7
BMP / 1.8 / 2.3 / 5.8 / 0.6 / 5.9

† WMP = whole milk powder, SMP = skim milk powder, WP = whey powder and BMP = butter milk powder.

The value of milk

A model of the marketing of dairy products was integrated with the processing model described previously. In concert, the models allowed determination of the value of milk and derivation of milk payment systems. Fluid milk was assumed sold on the local market, with other products sold at world market prices, shown in Table 4. Costs were included in the model for the storage, collection, reception and standardisation of milk, the processing of milk for the manufacturing steps in each channel and for packaging, storage, transport and marketing of the dairy products. The model recognised three classes of costs: volume-related costs, composition-related costs and costs determined by the quantities of products manufactured. Volume-related costs include those required for on-farm chilling and storage, for collection and transportation, and reception of milk at the processing plant. Composition-related costs include, for example in processing of milk powders, costs incurred after milk is received at the factory up until the completion of evaporation. Costs related to the quantity of produce include costs for product packaging, product storage, transportation to foreign markets and marketing. The partitioning of costs between the three classes varied with the product mix and with the composition of the milk. Some details were given by Lopez-Villalobos et al. (1999a) for whole milk powder, butter and casein.

Table 4 Market prices of dairy products (free on board)

Dairy product / £/kg
Fluid milk / 0.450
Whole milk powder / 1.080
Skim milk powder / 0.957
Cheese / 1.312
Casein / 2.469
Butter / 0.998
Butter milk powder / 0.957
Whey powder / 0.700

The net return from milk was defined as the aggregate market income from dairy products, less all processing and marketing costs. A processor margin was included in the manufacturing cost such that the net return represented the value of the milk to producers.

Market scenarios were modelled for the fluid milk processing channel by fixing fluid milk demand to one of two extremes. A 10% fluid milk market with an export-oriented processing scenario (e.g. New Zealand) was considered as one extreme. The residual milk after meeting the small fluid demand was all used in the cheese channel or for a channel mix comprising whole milk powder 30%, skim milk powder 25%, cheese 22% and casein/butter 23%, resulting in a typical New Zealand product portfolio. A 70% fluid milk market with high value processing scenario (e.g. Japan) represented the other extreme scenario. Residual milk was used either for cheese or equally split between cheese and WMP.

The values of milks according to these market conditions are shown in Table 5. The value of milk increases markedly with greater fluid milk demand. The change in milk value with variation in composition (for example comparing Jersey and Holstein-Friesian milks) is proportionately greater at low fluid as compared to high fluid milk demand. This would suggest that the relative advantage of breeds with higher value milks may be greatest in exporting countries such as New Zealand and least in countries that consume most of their milk in fluid form. The novel milks with high solids or high protein did not have sufficient carrier to allow production of high quantities of fluid milk without addition of water.

Norman et al. (1991) and Schmidt and Pritchard (1988) showed that Jersey milk was more valuable than Holstein milk when processed into cheese.

Table 5 The true values (£/kg) of various milks in two fluid scenarios and with different mixes of dairy products

Milk
Fluid
Milk / Residual / Average / Holstein / Holstein-Friesian / Jersey / Ayrshire / Crossbred
HF  J† / High solids / High protein
10% / Cheese / 0.124 / 0.107 / 0.119 / 0.143 / 0.120 / 0.126 / 0.180 / 0.159
10% / Channel mix║ / 0.112 / 0.097 / 0.108 / 0.130 / 0.109 / 0.114 / 0.162 / 0.144
70% / Cheese / 0.196 / 0.181 / 0.191 / 0.211 / 0.193 / 0.197 / N/A§ / N/A
70% / WMP‡-Cheese / 0.193 / 0.178 / 0.189 / 0.210 / 0.190 / 0.195 / N/A / N/A

† HF = Holstein-Friesian and J = Jersey.

‡ WMP = whole milk powder.

§ N/A = not applicable.

║ Channel mix was whole milk powder 30%, skim milk powder 25%, cheese 22% and casein/butter 23%.

Payment systems for milk

Payment systems were constructed as a method of approximating the true value of milk on the basis of one or more of its components. It was envisaged that such payment systems might be applied to milk supplied by individual producers for the purposes of calculating their reward. Single component payment systems were based on volume of milk only (VO), or on yield of fat (FO), on yield of protein (PO), or the sum of the yields of fat and protein (FPO). The payment systems were quantified by dividing the value of the milk by the quantity of the particular component. Multiple component payment systems took the form of linear equations that included the yields of fat and protein with a penalty for volume of milk (FPV) or yields of fat, protein and lactose with a penalty for volume of milk (FPLV). Volume charges were calculated by summing up all costs unrelated to the quantity of product manufactured and dividing by the volume of milk processed.

Economic values in the multiple component payment systems were obtained by distributing the industry revenue plus any volume charges according to the ratio(s) of the marginal revenues for each component. Marginal revenues for milk components were calculated by comparing the true values of two milks, the milks being identical except for a small change in one component (e.g. fat) and corresponding modification to the water content such that the milk volume remained constant. This approach is similar to that suggested by Ladd and Dunn (1979) for estimating the value of milk components to a dairy manufacturer.

Payment systems assessed separately for milk from each breed are shown for the two market scenarios in Table 6 (10% fluid milk) and Table 7 (70% fluid milk). These payment systems, applied to milk of the composition from which the payment system was derived, all gave the same value of the milk, equal to the true values as shown for their relevant channel mixes in Table 5.

Table 6 Payment systems† (£/kg) for milks from various breeds with 10% demand for fluid milk‡

Single components / FPV / FPLV
Milk / VO / FO / PO / FPO / F / P / V / F / P / L / V
New Zealand average / 0.112 / 2.405 / 3.182 / 1.370 / 1.012 / 2.285 / –0.016 / 0.925 / 2.088 / 0.235 / –0.016
Holstein / 0.097 / 2.767 / 3.027 / 1.446 / 1.101 / 2.317 / –0.016 / 0.998 / 2.101 / 0.224 / –0.016
Holstein-Friesian (HF) / 0.108 / 2.458 / 3.165 / 1.384 / 1.023 / 2.310 / –0.016 / 0.933 / 2.106 / 0.231 / –0.016
Jersey (J) / 0.130 / 2.248 / 3.234 / 1.326 / 0.978 / 2.209 / –0.015 / 0.898 / 2.030 / 0.250 / –0.015
Ayrshire / 0.109 / 2.509 / 3.123 / 1.391 / 1.018 / 2.305 / –0.016 / 0.929 / 2.103 / 0.233 / –0.016
Crossbred HF  J / 0.114 / 2.385 / 3.188 / 1.364 / 1.008 / 2.276 / –0.016 / 0.921 / 2.081 / 0.237 / –0.016
High solids / 0.162 / 2.315 / 3.063 / 1.318 / 0.921 / 2.122 / –0.015 / 0.852 / 1.963 / 0.282 / –0.015
High protein / 0.144 / 3.078 / 2.715 / 1.443 / 1.016 / 2.097 / –0.015 / 0.936 / 1.932 / 0.267 / –0.015

† F = yield of fat, P= yield of protein, L = yield of lactose, V = volume of milk, VO = volume of milk only, FO = yield of fat only, PO = yield of protein only, FPO = yield of fat plus protein only, FPV = yields of fat and protein with a penalty for volume of milk, FPLV = as FPV with yield of lactose included.

‡ Residual milk, after meeting fluid milk demand, was used for whole milk powder 30%, skim milk powder 25%, cheese 22% and casein/butter 23%.

Table 7 Payment systems† (£/kg) for milks from various breeds with 70% demand for fluid milk‡

Single components / FPV / FPLV
Milk / VO / FO / PO / FPO / F / P / V / F / P / L / V
New Zealand average / 0.193 / 4.137 / 5.474 / 2.356 / 2.105 / 2.951 / –0.009 / 1.921 / 2.693 / 0.377 / –0.009
Holstein / 0.178 / 5.099 / 5.577 / 2.664 / 2.378 / 3.285 / –0.010 / 2.167 / 2.994 / 0.356 / –0.010
Holstein-Friesian (HF) / 0.189 / 4.310 / 5.549 / 2.426 / 2.133 / 3.082 / –0.010 / 1.948 / 2.813 / 0.368 / –0.010
Jersey (J) / 0.210 / 3.635 / 5.230 / 2.145 / 2.056 / 2.483 / –0.008 / 1.869 / 2.257 / 0.422 / –0.008
Ayrshire / 0.190 / 4.371 / 5.440 / 2.424 / 2.136 / 3.051 / –0.009 / 1.948 / 2.784 / 0.373 / –0.009
Crossbred HF  J / 0.195 / 4.074 / 5.445 / 2.330 / 2.096 / 2.900 / –0.009 / 1.912 / 2.646 / 0.381 / –0.009

† F = yield of fat, P= yield of protein, L = yield of lactose, V = volume of milk, V = volume of milk only, FO = yield of fat only, PO = yield of protein only, FPO = yield of fat plus protein only, FPV = yields of fat and protein with a penalty for volume of milk, FPLV = as FPV with yield of lactose included.

‡ Residual milk, after meeting fluid milk demand, was used for whole milk powder 50% and cheese 50%.

The volume penalties were relatively insensitive to the composition of the milk being processed. In all milks investigated, protein had between two and three times the value of fat in a low fluid market and one to two times the value of fat in a high fluid market. The fat and protein payment relativity varied with the composition of the milk because the component that limited the production of valuable products varied among milks. For example, using Jersey milk, a marginal increase in protein content allowed increased cheese yield at the expense of butter. In the case of Holstein milk for the channel mix in Table 7, a marginal increase in protein could not be utilised in cheese as fat was limiting.

Lactose had positive values in the payment system in all the milks investigated. The value of lactose was between £0.224/kg and £0.422/kg, about one-ninth of the value of protein in a low fluid market and one-seventh the value of protein in a high fluid market. The value of lactose tended to increase in more concentrated milks as it was a limiting component for WMP and SMP production. Additional lactose allowed reduction in butter manufacture with concomitant increase in more valuable WMP. The positive value for lactose resulted in erosion of the component values for fat and protein, relative to the FPV payment system that ignored the value of lactose.

In most industries, one payment system would be applied to milks of varying composition sourced from many producers. Accordingly, it is worthwhile considering the impact of a payment system sourced from average milk, when applied to individual milks that contribute to the average.

The true values of milks from various breeds in an industry with low fluid demand are shown in Table 8, along with the estimated values of the same milks from payments systems derived from the analysis of average milk. Various single and multiple component payment systems from Table 6 were considered. In almost all cases, the estimated values of the milks differed from the true value of the milk. Some milks were under-evaluated and other milks were over-evaluated. The extent of upward and downwards bias must be equal, when weighted according to the relative contribution of each breed to the national average milk. Volume-based payment over-evaluated the Holstein-Friesian milk and penalised Jersey milk. Single-component payment based on protein (PO) predicted milk values closer to the true values than the other single-component systems, because protein constituted the most valuable component of the milk. Multiple component payment systems estimated milk value more closely than single component payment systems. The inclusion of lactose (FPLV) slightly improved the prediction relative to FPV, the current payment system used in New Zealand. These results showing multiple component payment systems are fairer than single component payment systems are in agreement with Paul (1985) and Emmons et al. (1990).