【正文】 50設(shè)計和建造合同已經(jīng)導(dǎo)致 16億美元的成本超支 ,其中大部分即可追溯到設(shè)計和現(xiàn)場條件不與期望。在這一章我們認(rèn)為合同的形式和獎勵機(jī)制可以剪裁的幫助緩解這個采購問題。我們認(rèn)為 ,簡單的職業(yè)活動 ,我們定義為易于設(shè)計和小的不確定性對生產(chǎn)所需要的 ,應(yīng)當(dāng)獲得使用固定價格合同 ,應(yīng)伴有高水平的設(shè)計完整性 (防止需要嗎適應(yīng) ),所以最好是通過競爭獲得招標(biāo)。 為我們的直覺處方源于緊張關(guān)系提供獎勵降低成本 ,避免昂貴和浪費 ,暨南請求之后變化。這種效率損失 ,常常會由于磋商價格有真的鎖定現(xiàn)有供應(yīng)商的希望使用需要改變自己的優(yōu)勢。 我們繼續(xù)認(rèn)為付款程序的選擇 ,如固定價格成本加成合同 ,是與后續(xù)的決定綁在一個 procurer 面 :不管授予競爭性招標(biāo)采購合同或通過談判的潛在的供應(yīng)商。在美國公共部門的律例管理采購 ,通常是基于美國聯(lián)邦航空局法規(guī) ,大力支持使用具有競爭力招標(biāo)。在招標(biāo)的優(yōu)點是無偏見的獎勵的計劃 ,它沒有反映優(yōu)化 ,事后的適應(yīng)。然后我們繼續(xù)描述如何選擇合同決定使用獎勵機(jī)制。s notorious Big Dig which required drastic changes in construction plans. Construction projects begin with the preparation of plans, specications, and blueprints. For reasons largely unpredictable and out of the control of both buyers and contractors, modications and revisions will be changes require contractors and their subcontractors to adapt the construction process without direction from a prior written contract. This is the sense in which construction contracts are inplete. According to inplete contracting theories of therm(Coase, 1937。 decisions to hire subcontractors, or|using the term coined in the literature, their \make or buy the public sector, construction projects are typically procured using petitive bidding. The most mon contracting format is called designbidbuild. First, the buyer39。 bids submitted exante, not costs incurred expost, establish the terms of pensation. Primary contractors, who submit bids, and the subcontractors they hire are liable for all costs. Consequently, forward looking, rational contractors will assess the likelihood that design changes will occur and incorporate anticipated adaptation costs into their bids. I develop a model representation of the designbidbuild process and propose a 11 measure of contractual inpleteness to address four quantitative questions. What is the eect of inpleteness on cost if a subcontractor performs work? What is the e ect if a primary contractor performs work?Do these eects dier? Finally, what are the dollarvalued magnitudes? The baseline prediction is that cost increases in the degree of inpleteness under both arrangements, but with a larger marginal eect for work performed by a subcontractor. I apply the framework to bridge projects procured by the California Department of Transportation (Caltrans). A vast body of empirical work qualitatively analyzes how rms are organized. These traditional studies, lacking cost measures, only address the third question listed above: does inpleteness(or some other parameter) a ect the probability of subcontracting? There is very little quantitativeevidence about the impact of rm boundaries on economic outes such as cost. This has been a major critique in this literature (Hubbard, 2020。 Klein, 2020). This paper o ers one of the rst attempts to bring the missing data, examine the cost primitives of the theory, and, stated bluntly, to show that rm boundaries matter. Such a research design is made possible because bids reect cost. The main nding is that inpleteness has a negligible eect on cost for work performed by a prime contractor. The eect is large for subcontracting。s speci cations list construction work items and corresponding quantities. On bridge contracts, tasks range from heavy engineering jobs such as installing structural concrete, steel, asphalt, and drilling to ancillary tasks such as trac striping, fencing, and landscaping. For each task, bidders decide whether to perform work themselves or hire a subcontractor. They also submit a unit price bid expressed as dollars per unit of quantity. Unit price bids are aggregated according to a scoring rule to determine the low bidder. Inpleteness is inherently a di cult concept to measure. A measurement should capture contractors39。 nor do I model any other exante actions taken in anticipation of bargaining. Instead, I treat the predictions of theory in a reduced form manner and use the model to show why the inpleteness proxy a ects forecasted unit costs and thesubsequent subcontracting a subgame, I model subcontract model the strategic aspects of bidding which also depend on quantity changes. These are scoring auctions. The total bid is calculated by multiplying unit price bids with original quantities,then summing those values across tasks. For each task, the winner is paid its unit price bid times the quantity actually installed. Di erences in original and nal quantities induce strategic bidding behavior. The basic intuition described by Athey and Levin (2020) shows that bidders skew unit price bids above cost on tasks expected to overrun on quantity and below cost on tasks expected to underrun. By skewing, a bidder earns a higher prot without a ecting its total bidand hence probability of winning the contract. Bid skewing is risky. If the overrun (or, for thatmatter underrun) does not occur, the winning bidder suers a lose. The principles for allocating unit price bids are analogous to concepts from modern portfolio theory (Markowitz, 1952). The specic modeling choices match industry practitioners39。 objectives. They choose organizational arrangements tha t minimi challenge in conducting an performance based make or buy study is to account for endogeneity in subcontracting decisions. The panel data structures provides controls. This study of contractual inpleteness falls into the class of empirical work that considers uncertainty and plexity of a transaction. Seminal contributions include Monteverde and Teece (1982), Masten (1984) and Masten et al. (1991). They nd a higher degree of plexity is associated with a lower probability of subcontracting. Recent work by Gil (2020) (movies), Acemoglu,Aghion, Grith, and Zilibotti (2020) (Ramp。 Bajari et al., 2020) restricted attention to just one dimen