公有建築計畫經費及標價審查模式之研究

Abstract

政府機關辦理公共工程，從可行性研究、規劃、設計、招標、施工以至竣工驗收，於多個時間點，必須進行決策。其中，計畫經費上限究應為何，屬於政府資源分配層次的重要課題；另招標階段，當廠商標價偏低時，機關決標與否的決策，常困擾著機關，本研究針對此二個決策事項進行探討。

首先，由於公共工程預算占政府部門預算之百分比甚高，如何客觀的決定工程計畫經費，關係到政府預算是否有效分配；然而，我國對於公共工程計畫經費審查之規定，僅規範程序而無量化的標準，審查人員往往以自身的經驗決定總計畫經費，審查的結果，其可信度有待提升。本研究提出創新的公共工程計畫經費審查程序，整合兩個模式：其一，以層級分析法為基礎的多目標準則評估模式，以確定審查準則及其權重；其二，以模擬為基礎，建立計畫經費累積成本分布曲線及其上下區間。經由層級分析法所得計畫成本估算得分，對映至成本累積曲線，可獲得計畫經費審查結果。

再則，依政府採購法規定，機關以最低標之總價決標方式辦理採購，當廠商標價顯不合理時，機關得拒絕決標予該廠商；然而，實務上，機關很少採用該規定，主因係該規定缺乏客觀的審查程序。過往研究多以總標價來審查標價之合理性，而本研究提出以工項單價層級的標價審查模式，以改善總價審查的缺點。模式中，最低標價以各符合資格之投標廠商及機關電子標單工項單價進行分析；此外，模式定義差價比，以衡量機關底價與最低價標間之差異。

Public construction project budgets often occupy a high percentage of the yearly total budget in many countries. Thus, accurately reviewing these project budgets is a crucial task to government officers for effectively allocating these budgets. However, for example, in Taiwan, regulations for reviewing the construction project budget only describe the governmental administration process in a qualitative manner. Without a systematic and quantitative method, government officers review the project budgets simply based on their own experience, eventually making the budget reviewing results unreliable. This work presents a quantitative procedure to support reviewing the budgets of construction projects. The procedure is developed by integrating an analytic hierarchy process (AHP) and a simulation-based cost model. The AHP is applied to accurately reflect the reviewers’ evaluations with respect to the budget determination criteria. While each cost item is assumed to be a variable, the cost model generates a cumulative cost distribution for setting the boundaries of the project budget. The merits of the procedure are demonstrated by to a practical Taiwanese project.

A lump-sum lowest bid method allows project administrators of public construction projects to reject any bid if they determine the bid price is unreasonable. In practice, however, few bids are rejected because of the lack of an objective process for evaluating the lowest bids. While past research assessed the bids based on total bid price, this study proposes a bid evaluation model developed based on the level of unit prices. In the proposed model, the lowest bid is assessed according to the unit prices electronically obtained from the owner and all qualified bidders. Moreover, a differential ratio is defined to measure the deviation between the unit prices submitted by both the owner and the lowest bidder. The merits of this innovative model are demonstrated by applying it to two subprojects of a building construction project. Practitioners and governmental officers have confirmed the feasibility of the model provided the languages or clauses describing the modeling algorithms are explicated in the general or special conditions of tendering documents.