The authors declare no conflicts of interest regarding the publication of this paper.
Social exchange theory (SET) explains why buyer-supplier relationships can generate cooperation beyond formal contractual obligations. Research shows that reciprocity can influence information exchange through credibility and benevolence trust, with information exchange then linked to calculative and affective commitment. Yet reciprocity is often represented as a general positive force, leaving unclear how suppliers interpret a buyer’s action, compare the relationship with alternatives, and decide which benefits to return, withhold, or redirect. This conceptual paper develops a preference-sensitive extension of SET for buyer-supplier relationships. It identifies four gaps: construct ambiguity in reciprocity; the assumption that exchange should be equivalent; the under-specification of supplier preference formation; and the possibility that relational investment produces opportunism as well as cooperation. The proposed model treats reciprocity as an interpreted and time-dependent mechanism through which buyer behaviour changes supplier’s preference and the allocation of scarce relational benefits. It integrates supplier attractiveness, satisfaction, expected future value, learning value, alternative customers, dependence, relationship-specific investment, perceived fairness, and the preferred rule of exchange. Six propositions specify how heteromorphic returns, long-term orientation, future dependence, acceptable asymmetry, goal congruence, and portfolio context condition reciprocal responses. The paper contributes a clearer micro foundation for supplier preference and a testable research agenda centered on longitudinal, dyadic, and portfolio-level designs.
Buyer-supplier relationships are governed by more than price and contract. They involve information sharing, technical assistance, early supplier involvement, problem solving, flexibility, investment, and expectations about future business. These exchanges are often incomplete, delayed, and difficult to value in a common unit. Social Exchange Theory (SET) is useful because it explains why firms provide benefits whose returns are uncertain, unspecified, or deferred. A buyer that shares forecasts or invests in supplier development may expect reliability, knowledge, innovation, or commitment in return. Research that incorporates reciprocity into an interfirm exchange model places it upstream of trust and information exchange rather than treating it as a simple one-for-one transaction ([
The central problem is that “reciprocity” compresses several distinct theoretical claims. It may describe a sequence of contributions, an expectation that benefits ought to be returned, or a norm that creates an obligation to respond. [
A supplier-preference perspective makes the missing decision process visible. A supplier evaluates the buyer’s current and anticipated value, compares that relationship with alternatives, interprets the buyer’s motives and treatment, and allocates scarce capacity, knowledge, attention, and innovation effort across customers. Evidence from supplier knowledge-sharing research shows that anticipated future dependence increases suppliers’ motivation to share knowledge through relational and learning motives, while current embeddedness changes the effects of economic and learning motives ([
This paper argues that supplier-preference models do not replace SET. They specify the actor, comparison set, and allocation decision that SET often leaves implicit. The resulting theoretical position is that reciprocity is not a constant direct effect. It is a preference-mediated process whose form and consequences vary across suppliers, exchanges, power conditions, time horizons, and relationship portfolios.
The paper makes three contributions. First, it separates the action, interpretation, exchange rules, preference change, and observed response that are often collapsed into a single reciprocity construct. Second, it explains how asymmetry and delayed returns can be compatible with effective exchange. Third, it develops propositions and an empirical agenda for testing reciprocity as a dynamic allocation mechanism rather than as a universal source of cooperation.
The present framework extends both Social Exchange Theory and preferred-customer research by generating predictions that existing models do not provide. Specifically, it predicts that identical buyer investments may produce different supplier responses when suppliers interpret those investments differently, when alternative customers are available, or when preference has not changed despite favourable exchanges. The model further predicts delayed, asymmetric, and non-reciprocal responses that cannot be explained by traditional assumptions of direct reciprocity. By identifying supplier preference as a mediating mechanism between exchange actions and allocation outcomes, the framework explains why similar buyer behaviours often produce divergent supplier responses across contexts.
SET explains relational behaviour that a narrow economic self-interest model cannot fully capture. Social preferences can include concern for the other party’s welfare and a disposition to reciprocate a history of positive interaction. In supply-chain experiments, relationship preference promoted cooperation, individual performance, and system efficiency over time, whereas status preference produced tougher actions and lower efficiency ([
The interfirm reciprocity model developed by Voss et al. specifies one important pathway: reciprocity affects information exchange indirectly through credibility and benevolence trust, and information exchange then mediates effects on calculative and affective commitment ([
That question concerns the supplier’s reference point, comparison set, and allocation problem. The supplier must evaluate the buyer’s treatment relative to expectations, alternative customers, dependence, relationship-specific exposure, and the anticipated future of the relationship. It must also decide which response is feasible given production capacity, knowledge, attention, and bargaining position. Without these elements, reciprocity can become a residual explanation: cooperation is observed and called reciprocal, while deterioration is interpreted as failed reciprocity.
The proposed extension treats supplier preference as a state that links exchange to response. Supplier preference does not mean an unchanging attitude or a purely affective liking. It means the supplier’s relative willingness to allocate scarce relational benefits to a buyer at a particular time. Preferred treatment is therefore an outcome of evaluation and comparison, not a mechanical return for buyer generosity. This framing is consistent with research showing that supplier relationship portfolios contain distinct sets of relationships along a closeness continuum and that supply-management practices affect both portfolio composition and supplier-provided benefits ([
A reciprocal transaction, a reciprocity belief, and a reciprocity norm are related but not interchangeable. A transaction concerns the sequence and structure of interdependence. Belief concerns what an actor expects. A norm concerns what an actor feels obliged to do. SET’s unresolved conceptual ambiguities include the rules governing exchange and the nature of the resources exchanged ([
The distinction matters in buyer-supplier research. If a buyer provides engineering support, a supplier might respond by sharing process knowledge, prioritizing production capacity, improving flexibility, reducing opportunistic behaviour, or merely maintaining the existing relationship. These responses differ in form, timing, cost, and strategic meaning. A binary measure of whether the supplier “reciprocated” cannot establish whether the response reflected obligation, calculation, goodwill, deterrence, or a change in the buyer’s position within the supplier’s customer portfolio.
The experimental evidence of Loch and Wu shows why reciprocity should not be treated as the whole of a supplier’s utility. Their study distinguishes relationship preference from status preference and finds that the former supports cooperation and efficiency, whereas the latter induces tougher actions and reduces performance ([
Recent supplier studies also point toward bounded rather than unconditional reciprocity. Prajogo, Cooper, Donohue, and Nair find that suppliers’ intention to continue relationships after customers’ unfair or socially irresponsible practices is shaped by supplier dependence and relationship benefits, a pattern they interpret as consistent with weak reciprocity rather than strong reciprocity ([
A second gap is the tendency to equate reciprocity with balanced exchange. Complex buyer-supplier relationships involve incomplete contracts, long time horizons, unspecified obligations, and non-contractual contributions. Tanskanen’s analysis of six buyer-supplier dyads show that exchange asymmetry can be explained by self-awareness of attractiveness, power-balancing mechanisms, and primacy ([
Equivalence is also difficult to define. A buyer’s investment in supplier development may be returned through quality or resilience, while a supplier’s investment in flexibility may be returned through future business. The contributions differ in form and may be separated in time. Hoppner and Griffith’s longitudinal study of 284 buyer-supplier relationships find that relational behaviour enhances financial performance when exchanged contributions need not be directly comparable or can be returned over a longer horizon. Performance declines when contributions must be comparable in form or returned quickly ([
This evidence challenges balance-sheet interpretation of reciprocity. The relevant question is not whether the parties contributed identical amounts, but whether the exchange is acceptable under the parties’ expectations and preferred exchange rule. A supplier may accept an unequal exchange when it signals goodwill, future value, or a credible commitment. The same asymmetry may be rejected when it increases relationship-specific exposure without a credible return. Asymmetry is therefore a boundary condition for reciprocity, not an automatic indicator of dysfunction.
Power, dependence, alternative customers, switching costs, and strategic value should be modelled as conditions that shape the meaning of unequal exchange. They influence both the supplier’s feasible responses and its interpretation of the buyer’s action. A preference-sensitive SET must consequently distinguish observed imbalance from perceived injustice and must measure whether the parties accept heteromorphic returns; returns that differ in form as legitimate.
The third gap concerns the supplier as an evaluator and allocator. Preferred treatment is not simply a mechanical return for buyer generosity. It reflects the supplier’s assessment of the buyer’s attractiveness, the value and quality of the relationship, the credibility of future business, and the costs and risks of allocating scarce resources to that buyer. In a study of 215 strategic buyer-supplier dyads, Moretto, Schoenherr, and Patrucco find that supplier development and involvement directly foster sustained supplier commitment, while information sharing affects commitment indirectly through buyer characteristics ([
Buyer investment can also carry an ambiguous signal. It may communicate commitment and support, but it may increase the supplier’s exposure to dependence, lock-in, or uncompensated relationship-specific assets. The supplier may therefore value the relationship and protect itself against the buyer at the same time. Treating investment as a uniformly positive predictor of cooperation obscures this dual interpretation.
Supplier knowledge sharing provides a second illustration. Chen, Zhao, and Lewis show that anticipated future dependence increases suppliers’ motivation to share knowledge through relational and learning motives, while current embeddedness moderates the economic and learning pathways ([
A supplier-preference model therefore contains three decision components. First, the supplier evaluates current and anticipated relational value. Second, it compares the buyer with alternative customers and competing demands. Third, it allocates scarce benefits such as capacity, information, flexibility, knowledge, and innovation effort according to its resulting preference ordering. Preferred treatment is thus an allocation decision embedded in a portfolio, not merely a reciprocal reaction inside an isolated dyad.
The fourth gap is the assumption that relational investment reliably produces cooperation. Supplier development is intended to improve capability and mutual performance, but its effects can diverge. Tran, Gorton, and Lemke’s structural-equation study of 204 buyers in Vietnam finds that supplier development improves buyer performance while also increasing supplier opportunism. The balance between positive and negative outcomes depends on goal congruence and long-term orientation ([
This finding complicates a linear investment → trust → cooperation sequence. A supplier may interpret investment as support, but it may also interpret it as evidence that the buyer is dependent, has limited alternatives, or will bear high costs if the relationship deteriorates. The supplier can then use the buyer’s commitment to improve its own bargaining position. Reciprocity in this case may be negative, defensive, or strategically selective rather than cooperative.
The implication is not that relational investment is ineffective. Its effect is conditional on interpretation and governance. Buyer support is more likely to produce constructive exchange when the parties share goals and orient the relationship toward the long term ([
The supplier-preference perspective shifts the unit of analysis from an isolated dyad to a portfolio. Suppliers serve multiple customers, and buyers manage multiple supplier relationships. A buyer’s treatment of one supplier is therefore interpreted against the supplier’s other customers; the buyer’s treatment of other suppliers also reveals how relational resources are distributed across its own portfolio.
Ellis, Oh, Henke, and Suresh combine case and survey evidence to identify three sets of supplier relationships distributed along a continuum of closeness. Their results show that supply-management practices affect the composition of these relationship sets and that supplier-provided benefits differ across them ([
A portfolio perspective also clarifies why equal treatment is not always the goal. A firm may allocate more resources to strategically important suppliers, while a supplier may allocate preferential treatment to customers offering the strongest combination of future value, relational quality, learning, and manageable risk. Reciprocity may operate within each dyad, but the meaning and opportunity cost of a response are shaped by the actor’s other relationships.
This perspective also yields a sharper definition of supplier preference. Preference is not simply the absolute quality of one dyad. It is the supplier’s relative willingness to allocate a scarce benefit to one buyer rather than to an alternative customer at a given time. A buyer can therefore improve its treatment of a supplier and still fail to receive preferential treatment if competing relationships become more attractive or less risky.
Supplier preference is the central construct in the model and represents the supplier’s relative willingness to allocate scarce resources to a focal buyer compared with other customers. Customer attractiveness and supplier satisfaction are antecedent evaluative constructs that influence but do not constitute supplier preference. Customer attractiveness reflects expected future value and strategic potential, whereas supplier satisfaction reflects evaluations of past and current exchange outcomes. Preferred customer status represents the recognition that a buyer occupies a superior position in the supplier’s portfolio due to accumulated preference. Preferred treatment is a behavioural outcome of supplier preference and preferred customer status and may be observed through enhanced resource allocation, innovation support, information sharing, flexibility, responsiveness, or preferential access to scarce resources.
A stronger SET model can retain reciprocity while making its conditions explicit. Let supplier preference for buyer “I” at time “t” be represented conceptually as:
where
The equation is a conceptual representation, not a validated measurement model. Its purpose is to distinguish antecedents that are often collapsed into a single reciprocity construct. Reciprocity is the process through which an observed buyer action changes INT, SP, or both. The supplier may then increase preferred treatment, maintain the existing level, withhold benefits, or respond negatively. The action does not automatically cause an equivalent return because its effect is filtered through interpretation, comparison, power, dependence, time, and the accepted form of exchange. Because supplier decisions are allocation decisions, portfolio ranking is the primary mechanism through which alternative customers influence supplier preference.
Although the model is presented conceptually, future empirical testing can employ longitudinal dyadic data. Customer attractiveness, supplier satisfaction, trust, dependence, commitment, and supplier preference can be measured using established survey scales. Allocation outcomes may be assessed through observable indicators such as capacity commitments, innovation support, responsiveness, information sharing, or preferential resource allocation. The lagged structure of the model permits estimation using panel-data techniques, longitudinal structural equation modelling, latent growth modelling, or cross-lagged panel analysis. Such approaches would permit empirical examination of preference formation and changes in supplier allocation behaviour over time.
When buyers provide supplier development support, information sharing, collaborative problem solving, or other relational investments that permit heteromorphic returns (buyer behavior), supplier interpretations that different forms of contribution are legitimate and mutually beneficial (mechanism) will increase the supplier’s allocation of knowledge sharing, innovation support, operational flexibility, and capacity allocation (observable supplier responses). This effect is more likely when exchange partners accept returns that differ in form rather than require direct equivalence ([
When buyers demonstrate long-term commitment through sustained investments, continued support, or deferred expectations of return (buyer behavior), supplier interpretations regarding relationship continuity and future relational value (mechanism) will increase the supplier’s responsiveness, capacity allocation, flexibility, and knowledge-sharing efforts (observable supplier responses). This effect is more likely when exchange partners accept delayed returns rather than require short-term settlement of contributions ([
When buyers provide credible signals of future business opportunities, learning opportunities, or long-term growth potential (buyer behavior), supplier interpretations regarding expected future value, learning value, and future dependence (mechanism) will increase the supplier’s knowledge sharing, innovation support, responsiveness, and preferential allocation of resources (observable supplier responses) ([
When buyers make relationship-specific investments, provide strategic support, or accept temporary imbalances in exchange (buyer behavior), supplier interpretations that exchange asymmetry reflects goodwill, credible commitment, or future relational value rather than exploitation (mechanism) will increase the supplier’s flexibility, capacity allocation, price concessions, and responsiveness (observable supplier responses). This effect is more likely when exchange asymmetry is perceived as acceptable and legitimate ([
When buyers engage in supplier development activities, capability-building investments, or resource commitments (buyer behavior), supplier interpretations regarding goal congruence, relational commitment, and long-term orientation (mechanism) will increase the supplier’s innovation support, knowledge sharing, flexibility, and responsiveness (observable supplier responses). However, when these investments are interpreted as increasing buyer dependence without shared goals, suppliers may reduce these allocations or exhibit opportunistic behavior ([
When buyers provide superior relational treatment through information sharing, collaborative engagement, supplier development, or equitable resource commitments (buyer behavior), supplier interpretations that the buyer occupies a preferred position relative to alternative customers within the supplier’s portfolio (mechanism) will increase the supplier’s capacity allocation, innovation support, knowledge sharing, responsiveness, and preferential access to scarce resources (observable supplier responses). This effect depends on the attractiveness of alternative customers and the supplier’s portfolio allocation priorities ([
These revised propositions align with the paper’s proposed causal logic:
Buyer Behaviour → Supplier Interpretation/Preference Mechanism → Supplier Allocation Response
and make the dependent variables directly observable and testable in future empirical studies.
These propositions require designs that can observe change, comparison, and response. First, longitudinal data is necessary because contributions and returns may be separated in time. Hoppner and Griffith show that a longitudinal interaction model provides a more accurate account than a cross-sectional interaction model when unobserved heterogeneity is a concern ([
This paper is a conceptual extension, not an empirical test. The proposed variables are not assumed to be independent, and the model does not claim that every supplier consciously computes a preference score. The framework is intended to discipline theorizing and measurement by identifying the evaluations and comparisons that must be distinguished if reciprocity is to be tested as a mechanism. It also does not imply that preference always produces cooperation. A supplier can prefer a buyer as a customer while withholding knowledge, demanding stronger safeguards, or using dependence to improve its bargaining position.
The preference-sensitive model proposed in this paper follows a four-stage causal sequence. First, a buyer action (e.g., investment, information sharing, flexibility, or support) is observed by the supplier. Second, the supplier interprets that action through existing exchange rules, expectations, prior experiences, and comparison processes. Third, that interpretation influences the supplier’s preference toward the buyer relative to alternative customers. Fourth, the resulting preference level affects supplier allocation decisions, including capacity allocation, innovation effort, knowledge sharing, flexibility, responsiveness, and other forms of preferred treatment. Reciprocity therefore operates through a preference-formation mechanism rather than as a direct action-response exchange.
SET remains useful for explaining why buyer-supplier relationships develop beyond formal contracts. Its limitation is not that reciprocity is irrelevant, but that reciprocity is often treated as too stable, dyadic, and unidirectional. The evidence reviewed indicates that relational behaviour can improve performance under heteromorphic and long-term exchange rules ([
The theoretical correction is to make supplier preference a central state variable. Reciprocity should be modelled as a conditional process through which buyer behaviour changes the supplier’s interpretation and relative evaluation of the relationship. That evaluation depends on attractiveness, satisfaction, alternatives, dependence, fairness, time horizon, investment exposure, and the form of return the supplier considers legitimate. Portfolio studies further show why the dyad cannot be isolated from the distribution of relational resources and benefits across multiple supplier relationships ([
The resulting claim is deliberately modest but consequential: the same buyer action can produce trust in one relationship, indifference in another, and opportunism in a third because the action is interpreted within different preference structures and comparison sets. Reciprocity, then, is not a constant. It is an interpreted response embedded in a changing supplier preference structure.
Atoapem Frimpong Barimah was the primary researcher. Dr. Vijayan Ramasamy was the advisor and reviewer of the initial draft, and the final manuscript that incorporated the peer reviewed suggestions for improvement.