Eric T. G. Wang, Terry Barron, Abraham Seidmann Management Science / Vol. 43, No. 12, December 1997

1. Contracting Structures for Custom Software Development: The Impacts of Informational Rents and Uncertainty on Internal Development and Outsourcing Eric T. G. Wang, Terry Barron, Abraham Seidmann Management Science / Vol. 43, No. 12, December 1997

2. outline • abstract • the model • internal bargaining • investment

3. abstract • custom software development projects • information asymmetries • user valuation and developer costs • relationship-specific investments • positive externalities for the user or the developer from the other party’s investment

4. market prices for software are not helpful in solving either the valuation or the cost problems • analyze the unique nature of the software development agreements that can be reached between the user and the developer

5. compare the value of using internal and external developers • understanding the factors relevant to the outsourcing decision

6. when internal and external developers have identical cost functions • internal development definitely yields the larger net value

7. the model is similar to the two-stage contracting • emphasis on software development bargaining outcomes under • informational asymmetries • relation-specific investments • investment externalities

8. the model

9. events and timing system’s value is affected by α user investment β developer investment ω user private information q feasible system specification switch cost relationship-specific investment requirements analysis focus on α and β are not contractible

10. at bargaining stage • user => value • developer => cost • assume ω and θ are not correlated • θ(ω) will not change after learning ω(θ) • assume that there is only one potential outsourcer

11. central management • internal develop • budget balancer • external develop • has all bargaining power • welfare weight

12. user’s investment will increase the system’s value at a diminishing rate • creates positive investment externalities and also increases the system’s value • developer’s investment will increase the system’s value at a diminishing rate • creates positive investment externalities and also decreases the system’s cost

14. internal bargaining • α and β has been made • two-sided incomplete information • user (developer) knows realized ω(θ) but not realized θ(ω) • optimal truth-revelation machanism • central management • maximize V(q, ω) - C(q, θ) • expect not to incur a budget deficit

15. let

17. the first-best software specification is

18. user’s expected payoff • developer’s expected payoff

19. central management’s problem is to set the software specifications, the user’s charges, and the developer’s payment • maximize the system’s net value

20. characterizes the set of incentive-compatible and individually rational mechanisms

21. compensate for the costs of investments so that the user and the developer are willing to participate at the beginning of the project

22. in order to induce the user and developer to reveal their private information • some premium over expected cost is paid to the developer • if the expected system net value is large enough to cover both the user’s and the developer’s informational rents • central management can expect not to incur a budget deficit

24. the central management can expect to at least balance the budget while at the same time inducing truth revelation and achieving bargaining efficiency

25. investment • first consider the case where the specific investments create no externalities • Vβ=0 and Cα=0 • the efficient investments are

26. first-order conditions

27. theorem 2 • Given the efficient mechanism in period 2 (development), the ex ante investments are also efficient regardless of whether the investments are contractible, provided that the investments create no externalities.

28. central management’s beliefs are set at • user can predict the resulting specification of the new system • user knows her own optimal reporting strategy perfectly, she can predict the exact specification as a function of θ

29. at the outset, central management displays a mechanism (or a contract) • instructs the user and the developer to make investments equal to α* and β*

30. result shows that under-investment will not arise when the investments generate no positive externalities • if the investments do generate positive externalities, under-investment occurs

31. external bargaining • user’s problem in period 2 for a realized ω can be formulated without loss of generality as

32. incentive compatible and individually rational payment extra cost that the user has to pay the developer

33. optimal system specification for outsourcing informational rent required for inducing truth-revelation

34. investment • user offer a contract based on • realized ω • investment α • beliefs developer’s investment • letting denote the user’s optimal system specification given

35. with the developer’s beliefs about user’s investment fixed at • developer’s optimal investment • first condition

36. given beliefs about the developer’s investment • user’s optimal investment • first condition

37. equilibrium

39. if the value and cost of developing the system are the same for internal and outsourcing • how the cost is evaluated • internal • no real resources have been transferred out • outsourcing • outsourcer’s profit margin • real dollars that the organization has to pay

40. numerical example • k = 1, s = 2, t = 1, t0 = 1

42. Figure 3 Figure 5

43. Figure 4 Figure 6

44. managerial and research implication • an internal developer can be expected to have a substantial advantage over an outsourcer • rational firm will behave in just this way • 186 projects from 5 firms • 64% custom / insource • 17% custom / outsource • 11% package / insource • 8% package / outsourced