Last modified: 2012-08-24
Abstract
Between the devil and the deep sea:
Cost and Differentiation Strategies for Brazilian Coffee Producers
Keywords: differentiation; resources; competitive advantages; coffee; Industrial Organization; Resource Based View
Justification/Motivation
The paper examines when strategy of differentiation is feasible for producers (focusing on Brazilian producers of specialty coffee). Our main hypothesis is that sustainable competitive advantages (SCA) should emerge specifically when an integration of both Industrial Organization (IO) and Resource Based View (RBV) perspectives is considered by the firm.
Research Problem
The issue of income distribution in economic global value chains has been extensively discussed in recent decades. Several studies have shown a decline in the rural segment’s share of the total income generated in the production chain (Cankorel, 2000; Daviron, Ponte, 2005; Leibtag et al., 2007; Morisset, 1997; Talbot, 1997). From this perspective, rural development theorists have suggested ‘decommoditization’ (differentiation) as the solution to the crisis in the countryside, and to achieve sustainable competitive advantages (SCA) (Fitter, Kaplinsky, 2001; Kaplinsky, 2000, 2004; Valceschini and Nicolas, 1995).
The discussion is supported by the Industrial Organization (IO) approach, which argues that firms obtain SCA by creating entry barriers against competitors and opening potential consumer markets (Foss, 2005; Porter, 1985). It is therefore assumed that – considering the structure of the industry, managers should explore market characteristics in order to create market power (by erecting barriers to competitors), preventing (or mitigating) the action of competitive forces. In that view, competitive strategy is the search for a favorable competitive position in an industry, referring a fundamental arena in which competition occurs.
This approach, however, has been the target of criticism, both theoretical and empirical. On the empirical level, studies of several Brazilian production chains challenge this proposition, such as coffee, soybean, corn, rice, and cotton (Farina and Zylbersztajn, 1998). Product differentiation has not prevented the exclusion of producers who can not keep pace with increased productivity and lower prices. On the theoretical side, studies on strategy field have criticized this approach as well (Barney, 1991; Conner, 1991; Peteraf, 1993; Rumelt, 1984; Wernerfelt, 1984). These scholars have admitted that although firm’s competitiveness in the short term may be linked to the price/performance ratio, in the long run there will be a convergence of similar patterns of product cost and quality, implying that increased barriers to competition would be less important as sources of differentiated advantages.
Such critics argue that over the long term, competitiveness arises from the ability of firms to build, with lower costs and greater speed than the competition, the core capability to produce products or services that cannot be foreseen by competitors (Barney, 1991; Peteraf, 1993; Rumelt, 1984; Wernerfelt, 1984). From this approach, Resource Based View (RBV), scholars argue firm’s possession of strategic resources as the main source of its competitive advantage. RBV approach, however, is also subject to criticism, because unlike the IO (discussed above) it focuses on the factor market as if the consumer had no importance in the strategy.
From that review, it is quite natural to assume both theoretical approaches, RBV and IO, as complementary in developing firm strategies, since each addresses only one side of SCA. This is our main statement in the paper; proposing SCA emerges specifically when an integration of both perspectives (IO and RBV) is take into consideration by the firm.
It is worth noting that this proposition is not new, nevertheless. In fact, some authors, such as Montgomery and Porter (1998, p. xviii) have previously suggested the integration of these views by admitting “sustainable competitive positions (a) reflect certain economic regularities; (b) are often the product of non-inherited advantages created; and (c) are built around sets of unusual capabilities that are difficult to be imitated by competitors.”
However few are the studies that test this theoretical issue. This is our main goal in the paper.
Based on a survey undertaken with 119 Brazilian producers of specialty coffee, we go further the literature testing the relationship between differentiation and resources in one agricultural sector.
The choice of specialty coffee chain was deliberate, driven by two main factors. First, coffee is one of the world’s most valuable agricultural commodities (Kaplinsky, 2004; Vorley, 2003), and Brazil is a leader in global coffee production. Second, coffee is an emblematic case concerning ‘decommoditization’ possibilities for agricultural products as a result of consumer trends.
The paper is structured in four parts in add to this introduction. Next section contextualizes when differentiation and resource based strategy are not enough in the increment of producer income. Third section applies an empirical analysis to the case of coffee farmers in Brazil and presents the methodology used in the paper. Fourth section presents a discussion of the results. Finally, the fifth section ends the paper by offering final remarks.
Methods
The data used in the estimate was obtained from interviews with 119 producers of specialty coffees in the states of São Paulo and Minas Gerais, conducted by telephone between July and November 2007.
The dependent variable [DPRICE] is the percentage increase in the price of specialty coffee in relation to commodity coffee, as perceived by the producer. In the questionnaire, we asked for an “estimate of the price differential between the specialty coffee sold by your property and that sold in the non-specialty market (as a percentage).”
The independent variables are:
[DCOST], referring to the increase in percentage of the cost of specialty coffee compared to the cost of the commodity, as perceived by the producer. In the interview, we asked producers to “[estimate] the cost differential between the specialty coffee produced on the farm and non-specialty sold on the market (as a percentage)”.
[CERTIF], a dummy variable for coffee certification;
[EDU], related to the level of education of rural producer.
[ASSOC], a dummy variable for the producer’s membership of an association or grower cooperative.
[CONTRA], referring to the percentage of production sold through contracts.
[ROAST], concerning with the percentage of production retained for their own roasting (forward vertical integration).
[NBUYER], the number of buyers of coffee.
[TIME], related to the length of business relationships with key buyers.
It is important to note that the variables related to increases in prices and costs associated with the production of specialty coffees reflect an impressionistic assessment by the interviewees, and are not necessarily anchored in accounting procedures or objective market information.
The equation was estimated using the ordinary least-squares method, where the subscript indicates the i-th producer of specialty coffees and β are the parameters to be estimated:
DPRICEi = β 0 + β 1DCOSTi + β 2CERTIFi + β 3EDUi + β 4ASSOCIi + β 5CONTRAi + β 6ROASTERi + β 7NBUYERi + β 8TIMEi + ei
The estimates were obtained using PASW (Predictive Analytics Software) Statistics 17.0, produced by the IBM Corporation.
Results and discussion
Estimates of equation parameters and variance analysis are presented in Tables 1 and 2. The estimate of the constant indicates that the specialty coffee producers interviewed assessed the premium for specialty coffee to be 15% over the price of commodity coffee. If the coffee is certified, it is perceived to obtain a price nearly 21% higher than the price of the commodity.
The increase in the perceived cost is positively associated with the perceived increase in prices: on average, producers estimate that a 1% increase in costs required for the production of specialty coffee is associated with an approximately 0.65% increase in the price received.
Producers of certified coffees estimate their costs to be on average 16.7% (standard deviation of 11.54%) higher than the costs of producing commodity coffee, while establishments that do not certify specialty coffee estimate the increase in costs at 15% (standard deviation of 9.26%).
Table 1. Coefficients.
Model
Non-standardized Coefficients
Standardized Coefficients
t
Sig.
B
Std. Error
Beta
(Constant)
15.026
4.644
3.236
0.002
ASSOCI
-0.123
1.187
-0.008
-0.103
0.918
CERTIF
5.963
2.539
0.2
2.349
0.021
CONTRA
-0.01
0.013
-0.065
-0.794
0.429
DCOST
0.648
0.119
0.459
5.443
0
EDU
-1.527
1.079
-0.118
-1.416
0.16
NBUYER
-0.151
0.78
-0.016
-0.193
0.847
TIME
0.087
0.132
0.056
0.662
0.509
ROASTER
0.211
0.558
0.031
0.378
0.706
Note: Dependent Variable (DPRICE).
Table 2. ANOVA.
Model
Sum of Squares
DF
Mean Square
F
Sig.
1
Regression
7,114.413
8
889.302
5.482
0.000a
Residual
18,007.585
111
162.230
Total
25,121.998
119
The dispersion of the increase in cost may be associated with the presence of fixed costs in certification, so that for smaller establishments the incremental cost of certification is relatively higher. Another factor that may explain the dispersion of the perceived increase in costs is the existence of entry costs for certification systems, making the cost for new entrants relatively higher than for establishments that participate in certification systems for a longer period.
The estimates of the coefficients of the variables related to the governance of transactions (CONTRA and ROASTER) were not statistically significant. Apparently, the governance structure does not influence the price obtained as perceived by the producers of specialty coffees. Likewise, the variables designed to capture aspects of the transactions related to the availability and ability to process information (EDU, ASSOC, TIME, and NBUYER) did not prove significant.
The coefficient of the variable level of education had a negative sign, with a significance level of 16%. One hypothesis to explain this phenomenon is that the level of education may be associated with conservative estimates of the quality premium for specialty coffee.
The number of buyers and the length of relationship with key buyers did not affect the quality premium for specialty coffee. It is likely that the specialty coffee market is transparent, so that even those who sell for few and/or traditional clients have adequate information on market conditions.
The variable DCOST, the incremental costs required for the production of specialty coffees, captures aspects of the technology used for the production of specialty coffees, and at the same time reflects specific characteristics of each agricultural establishment. Different businesses require different increments of cost to produce specialty coffee, depending on the ownership of resources: natural conditions (climate, topography, soil) and the provision of physical and human capital that existed prior to the decision to produce specialty coffees. Incentives to produce specialty coffees are therefore different for different facilities, as predicted by the RBV.
The basic premium for specialty coffee (regression intercept) is very close to the average increase in cost borne by the producers of specialty coffees: some 15%. The proximity of the price increase and cost increase, valid for the average producer, suggests strong competition among producers of specialty coffees. Performance starts to depend on the position that each producer occupies on the axis of quality versus cost.
Being a good producer is not enough; it is necessary to be better than others with regard to the combination of quality versus cost. The choice of the optimal strategy is conditional on the firm’s resources.
With full information, producers who were below the line DPRICE = DCOST would not have produced specialty coffee, because they had no advantages arising from private resources – the quality differential is small and/or the cost is too great.
Perhaps the apparent irrationality of these producers ‘below the line’ is explained by lack of information (bias of overestimation of returns) or by a dynamic effect: the return on a strategy is conditioned by the number of producers who adopt the same strategy. For the pioneers, it could be profitable to produce specialty coffees, even with large increases in cost. Over time, the premium falls because of the entry of new producers of specialty coffees. In this interpretation, it is the ‘line’ (of the feasibility of the strategy) that is lowered.
Figure 1 (payoff matrix) shows certifying decision as a result of resources that enable the generation of income differentiation.
RESSOURCES
ADEQUATE
INADEQUATE
CERTIF
YES
++++
-
NO
+++
++
Figure 1. Payoff matrix.
Indeed, Figure 2 shows that there are incentives for some firms (with a low incremental cost) to produce specialty coffees; for other establishments, incentives will be positive only if the special coffee receives some type of certification. For a third group of producers, with high incremental costs, incentives are insufficient for the production of specialty coffees. The vertical axis (Incremental Price) represents, as a percentage, the average premium over commodity coffee that producers believe they receive for quality coffee, positively associated with product differentiation. The horizontal axis (Incremental Cost) reflects the allocation of firms’ resources: firms less (more) gifted exhibit higher (lower) incremental costs to produce specialty coffee. It should be noted that the chart reflects both IO and RBV approaches. Vertical axis shows the influence of differentiation, while horizontal axis discloses the gains based on resources.
Figure 2. Incentives for the production of specialty coffees, according to the firm’s resources and capacity for product differentiation.
Figure 3 shows the plot of real data, taking into account the incremental gain variables arising from the differentiation and costs resulting from the ownership of resources.
Figure 3. Plot of observations in the plane incremental cost X incremental price
Note: l without certification; p with certification.
Figure 2 shows, in greater detail, the position of firms in the plane incremental cost X incremental price: 17 observations are in the region of weak or nonexistent incentives for the production of specialty coffees, three of which correspond to certified coffee and 14 without certification. Most producers are in the area where incentives would be sufficient to produce specialty coffee with or without certification.
From the results presented above, a question remains opened nevertheless: the relevant resources in the certification decision. It’s important to notice that risk aversion may influence the decision toward certification process, even in the presence of adequate resources (see Figure 1).
We estimated the conditional probability of farmer does not certify their coffee, by using Probit model. Main finds are showed on Tables 3 and 4.
Results suggest probability of certifying is positively related to marketing channel practiced by coffee producers. The relationship with roaster industry and exporter increases the probability of certifying; while the supply for industry and coffee trader is related to a lower probability of certification.
In add, the probability of certifying decreases with the participation in associations / cooperatives. However, the volume of coffee is positively related with certification decision (probably due to fixed costs of certification). In add, members of associations / cooperatives can produce specialty coffees with lower increments of costs comparatively to producers not associated. In turn, non-members who certify get larger price increments. Risk aversion may be a component of the decision to join a cooperative or association (lower incremental cost).
The relevance of association as an apparent substitute for certification led to investigate the differences between groups of producers (Tables 5 and 6).
For most, association and certification function as substitutes (Groups 2 and 3, related to 90 observations): certification increases price increment, whereas association reduces cost increment. For a small group (Group 1, with 17 observations) however, certification and association seem to behave as complements. This strategy apparently requires high volume production.
Tables 3. Parameter estimates.
Tables 4.
Tables 5.
Table 6.
CERTIF
NÃO CERTIF
1 ASSOCI
(N=17)
2 NÃO ASSOCI
(N = 41)
3 ASSOCI
(N = 49)
4 NÃO ASSOCI
(N = 51)
INSTR
3,706
0,772
2,780
1,275
3,347
0,948
3,077
1,045
RENDACAFE
54,706
31,941
61,268
33,128
52,083
32,811
61,040
34,773
NCOMP
3,188
1,223
2,583
2,062
2,261
1,769
2,083
1,028
COOP
4447,706
6811,086
917,317
3250,179
1079,449
1716,231
918,846
1875,062
EXPORT
5643,118
9107,406
3114,488
5250,704
1225,714
5740,640
1257,058
4321,628
TORREF
463,824
851,597
679,268
2719,949
114,082
380,120
75,519
277,409
SOLUVEL
0,000
0,000
82,927
471,117
2,041
14,286
50,962
367,489
TRADER
0,000
0,000
53,659
240,933
20,408
142,857
2,500
18,028
TEMPOREL
11,588
6,548
9,475
7,900
12,723
8,922
12,694
9,894
VOLUME
7412,941
7098,598
2271,667
3298,234
1864,362
2859,327
1592,245
2431,539
CONTRATINF
5,294
21,828
2,707
11,833
1,939
6,833
4,423
12,274
CONTRSAFRA
1,765
7,276
1,463
7,925
1,633
10,072
0,192
1,387
CONTRLP
0,000
0,000
0,366
2,343
21,429
142,887
1,154
8,321
CONTRCASA
1,765
4,982
0,561
2,098
1,673
5,035
1,442
5,454
CONTRCPR
4,588
8,931
3,171
10,826
0,918
3,174
0,673
3,136
CONTRTODOS
13,412
25,194
8,268
18,277
27,592
142,878
7,885
17,611
INTVERT
0,588
2,425
0,366
2,343
0,306
2,143
0,000
0,000
SPOT
82,176
28,050
86,268
27,344
83,429
29,982
89,808
22,073
OUTROS
5,000
11,456
4,146
12,293
3,163
10,738
3,077
15,408
DCUSTO
15,000
12,982
17,415
11,025
13,568
9,424
16,900
8,924
DPREÇO
24,706
10,073
28,488
15,097
20,413
14,148
21,784
13,000
Final remarks
The paper shows differentiation is not for all. The empirical results focus on coffee segment are not restrict to it. In fact, our results suggest differentiation strategy is no guarantee of success in achieving monopolistic gains doesn’t matter the segment, as the costs involved in its implementation can erode the gain in market share of the product. For this reason, a strategy of differentiation should be aligned with a resource strategy that enables the generation of economic rents in creating entry barriers to the factor market.
The attempts to unify different views on the study of strategy show us the long path to be follow nevertheless. Though the above-mentioned analysis point to the limits of dealing with the issue of value creation and capture in an isolated manner.
Having this in mind, the question of this paper (When differentiation and resource based strategy are not enough) must be interpreted in light of an integration view considering both: IO and RBV perspectives. RBV view explains why successful strategies may not be copied by all producers. From the perspective of OI, the different resource endowments, especially the non-reproducible, act as barriers to entry in a differentiated segment. Each firm has a unique combination of resources.
According Brazilian specialty coffee producers, the incentives to adopt the strategy of differentiation vary significantly between firms. Some hope to achieve substantial improvements in product quality, translated into higher prices received, with relatively modest increases in costs. Other producers believe that prices can only improve with investments that result in high costs. The differences between firms are due, at least in part, to irreproducible factors peculiar to the firm. The analysis of strategic choice of firms cannot be detached from the identification of resources available to competitors.