In our questionnaire we specified thirteen clauses that address these issues. We verified these clauses in the exploratory case studies and the document analysis of the high-tech alliance contracts. Hence, we applied this categorisation of contractual clauses because it was tailor-made to our research context. We take the sum of the clauses included in the contract as a measure for the level of contractual detail.

We used two items to measure dedicated assets in terms of partner specific machinery and instruments. Spill-over risks are also measured by two items: one item measures these risks for the industry and one for the focal company in relation to a partner firm. We used five items to measure interorganizational trust. Our definition characterizes interorganizational trust as a multi-component construct based on three related components: forbearance from opportunism (measured by two items), care and concern (measured by two items) and lack of monitoring (measured by one item).

Two items were used to measure the size of the focal firm, i.e. the number of employees and annual turnover. The complexity and strategic importance of the project were measured by one item that directly relate to the particular construct. We used two items to measure risk avoidance by the focal firm: the inclination to use a detailed formal contract as well as to align with procedures and legal rules in an interfirm alliance.

4. EMPIRICAL RESULTS

4.1 Contract characteristics

The data from our sample show that companies aim for different levels of detail in the contract. About 10 per cent of the respondents used low detailed contracts with less than five clauses included in the arrangement; another 55 per cent used moderately detailed contracts with less than 10 clauses; and 35 per cent used highly detailed contracts with 11 or more clauses included in the contract, although only 13 per cent thereof reported to have all clauses covered in the formal contract. The latter aligns with a highly detailed contract.

Table 2 provides an overview of the relative importance of each contractual clause.

All contractual clauses were found in one of the contracts that we investigated. However, Table 2 shows that there is variation in the relative importance of the contractual clauses. The most important clauses concern those that address the investments of the alliance partners (present in 89.26 per cent of all contracts), the project plan (present in 86.45 per cent of all contracts), the duration (present in 85.17 per cent of all contracts) and the goal of the alliance (present in 84.91 per cent of all contracts). In less than half of all the contracts we find arrangements concerning patent rights, risk allocation, conflict resolution and licence agreements.

4.2 Contract functions

Before we present the regression results, we first determine whether 'contractual detail' is a unidimensional or a multidimensional construct. For this, we performed a two-stage factor analysis for the thirteen clauses that are included in our research (Joreskog & Sorbom 1993, 1996). Table 3 reports the results for the exploratory factor analysis and Table 4 for the confirmatory factor analysis.

The exploratory factor analysis produced three factors with eigenvalues greater than one, which together accounted for 54.6 percent of the variance in the data. The confirmatory factor analysis supports these findings: all factor loadings for a particular construct exceed the threshold value of 0.50 and are significant (with t-values > 2). Thus, our analysis provides three different factors or contract dimensions. Below, we offer an interpretation of these dimensions. The interpretation derives from carefully inspecting the content of the clauses that construct a particular dimension.

We interpret the first factor as the traditional safeguarding function of contracts for risks and spill-over. This function of the contract is mainly to limit the incentives towards and opportunities for opportunistic behaviour by one of the alliance partners. The fear of the incentives and opportunities for opportunistic behaviour leads to the design and inclusion of clauses that safeguard ownership and property rights. Hence, clauses with a high loading on this factor concern product or technology ownership, methods, licence agreements, patent rights and confidentiality. The standardized Cronbach's alpha of this factor (0.77) is satisfactory.

We interpret the second factor as a combination of the coordination and the commitment function of a contract. This contract may be very detailed but will generally focus more on the positive (what we want to achieve and how) than on the negative (which legally enforceable measure we put in place to safeguard property or knowledge and how we take the case to court). Contractual clauses facilitating cooperation, which are interpreted as, for example, technical aids to relationship management--such as minutes from meetings, memoranda of agreements or outlines of the nature of alliances--do not necessarily increase the level of detail of a contract in terms of opportunism preemption. Clauses loading high on this factor concern the goal and outcome of the relationship, the project plan, and its project management. The results show that the coordination function serves to make arrangements that guide the day-to-day management of the interfirm alliance.

Additionally, the clauses that concern the investments of the alliance partners as well as the duration of the relationship also support this factor. These two terms are related to the commitment function of a contract: partners may also use the contract as a tangible expression of their trust in each other and their intention to be loyal partners with high levels of dedicated investments and long-term commitments. The contribution of each partner in terms of knowledge, financial and human resources as well as the duration of the relationship are discussed and determined.

Apparently, coordination and commitment go hand-in-hand as our respondents do not differentiate the clauses into separate factors. For that reason we interpret this factor as the 'coordination and commitment' function of a contract. The standardized Cronbach's alpha (0.74) is satisfactory.

We interpret the third factor as the 'external contingency' function of a contract. When parties engage in a long-term and complex relationship, parties may put a detailed contract into place to have a framework for how to (re)act if unforeseeable contingencies arise. Time is a crucial aspect in this function, as over time events may transpire beyond the control of the alliance partners but for which they need to be prepared, such as radical new innovations or the hostile takeover of one of the partners. Such issues may be harmful but unlike opportunism do not have this purpose.

External contingency clauses do not take the inclination towards or possibilities for opportunism as their point of departure. In other words, managers are not actively concerned by their counterparts' opportunistic behaviour at the moment these clauses are designed. However, they are not naive and also acknowledge that unforeseen circumstances affecting the relationship's conditions or context may arise which could change the parties' interests and the particularities of the relationship. Clauses loading high on this factor concern the future risk allocation, relationship adjustment and termination arrangements, and conflict mediation. These clauses deal with issues that may arise from within or outside the relationship. The standardized Cronbach's alpha for the items that construct this factor (0.71) is satisfactory.

To summarize, formal contracts between high-tech alliance partners may serve different functions. In our setting of Dutch innovative collaboration we show that contracts may have the function to safeguard positions and spill-over risks; to coordinate alliance activities and to show commitment; or to safeguard external contingencies.

4.3 Antecedents of contract detail and contract functions

Because our items have different scales, we used PRELIS to calculate the matrix of (polychoric, polyserial or Pearson) correlation coefficients and LISREL to estimate the effects of the antecedents and control variables on contract detail and the three different contract functions (Joreskog & Sorbom 1993, 1996). The descriptive statistics are in Table 5 and the regression results in Table 6 which also provides partial evidence for the transaction cost explanation of detailed contracts.

Our first hypothesis predicted that contractual detail will be greater for alliances involving greater transaction-specific assets. Table 6, however, shows that dedicated assets do not determine the level of detail of an alliance contract ([beta] = 0.01, n.s.). Hypothesis 1 therefore needs to be rejected. But we do find that spill-over risks have a positive and significant effect on contractual detail ([beta] = 0.15, p < 0.01). The greater the risks of spill-over, the greater the number of contractual provisions built into the alliance contract. This confirms Hypothesis 2.

Our third hypothesis suggested that interorganizational trust can substitute for detailed alliance contracts. Table 6 shows that interorganizational trust does reduce the need to specify many details but the effect is not significant ([beta] = -0.06, n.s.). Hypothesis 3 is thus rejected. As for the control variables, our results suggest that large firms strive for more detailed contracts ([beta] = 0.14, p < 0.01). The same applies to the complexity of the interfirm projects: the more complex the interfirm project, the more detailed the interfirm contract ([beta] = 0.19, p < 0.01).

We cannot find significant evidence for the strategic importance of the project to the focal firm and alliance contract detail ([beta] = 0.06, n.s.). However, risk aversion of the focal firm is an important antecedent because it makes the alliance contract more detailed ([beta] = 0.18, p <0.01). Hence, these results confirm valuable insights from transaction costs economics for the governance of high-tech alliances. More detailed contracts are needed to safeguard positions not so much due to dedicated assets but more because valuable knowledge is at stake and spill-over risks need to be reduced. The results cannot confirm that in the setting of high-tech alliances, business agreements are influenced by the social characteristics of the relationship.

Print Get the Mag Weekly Updates