1. Introduction Infrastructure projects in India are famous for delays and cost overruns.

Recently commissioned, Bandra-Worli Sea Link amply demonstrates the state of project delivery system in the country. What was planned as a Rs 300 Crore project to be completed by 2004 has actually cost Rs 1,600 crore along with a delay of five years. Indeed, very few projects get delivered in time and on cost. The delays and cost overruns have become hallmark of infrastructure projects in India. The quarterly reports of the Ministry of Statistics and Programme Implementation (MOSPI) stand testimony to a saga of unfettered delays and cost overruns. Yet, the extents and the causes behind the time and the cost overruns have remained understudied. As a result, the types of the policy interventions required to rectify the malady have also remained unidentified. Delays and cost overruns have significant implications from economic as well as political point of view. Due to delays in project implementation, the people and the economy have to wait for the provisions of public goods and services longer than is necessary. Thus, delays limit the growth potential of the economy. Similarly, cost overruns reduce competitiveness of the economy. Services provided by infrastructure projects serve as input for other sectors of the economy. Cost overruns in these projects lead to an increase in the capital-output-ratio for the entire economy. Simply put, delays and cost overruns reduce the efficiency of available economic resources and limit the growth potential of the entire economy. Definitions: Every infrastructure project has to undergo several stages: from planning of the project, to its approval, to awarding of contract(s), to actual construction/procurement, and so on. Broadly put, a projects lifecycle has three phases; development, construction, and operation-and-maintenance phase.

In the beginning of the development phase, the project sponsoring department prepares estimates of time and cost (funds) needed to complete the project. An expected date of completion is also announced. The actual date of completion is invariably different from the expected date. We define time overrun as the time difference between the actual and the initially planned (i.e, expected) dates of completion.5 The time difference is measured in months. A related term used in the paper is the implementation phase or implementation period. It is defined as the duration in which a project is planned to be completed, i.e., the duration between the date of approval of the project and its expected date of completion. Therefore, for each project we can define percentage time overrun as the ratio of the time overrun and the implementation phase for the project (multiplied by one hundred). Clearly, the time overrun and therefore the percentage time overrun can be positive, zero or even negative. Similarly, we define cost overrun as the difference between the actual cost and the initially projected (i.e., expected) cost of the projects. The initially expected cost is called the initial project cost. This is the estimated cost of project works. It is estimated when a project is planned and generally is arrived at using current input prices. The actual cost becomes known only at the time of completion at the end of phase two. Percentage cost overrun for a project is defined as the ratio of the cost overrun and the initially projected cost of the project (multiplied by one hundred). Again, percentage cost overrun can be positive, zero or negative. Causes Technical and Natural Factors: The estimation of project time and cost for infrastructure projects is a characteristically complex exercise. Though the estimation techniques have become better and sophisticated in recent times, they are still imperfect. As work on a project starts, its future unfolds and the authorities along with the contractor become better informed about the specific technological and material requirements of the project works. For example, during construction phase of a road project, an unexpectedly poor quality of soil may necessitate changes in the engineering, the design and the quality of bitumen required, from what were initially planned. Such changes may require extra time as well as funds. In some cases the actual circumstances, in contrast, may turn out to be favourable and the parties may find that they had made excessive provisions of funds and time. Similarly, flood or any other event of force majeure may cause delay as well as destroy the project assets. Alternatively, the natural conditions may turn out to be rather conducive, saving construction time and costs. Therefore, due to imperfect estimation and natural factors the actual project time and cost will generally be different from their expected values. The Contractual Failures: As discussed before, the construction phase of an infrastructure project starts with signing of a construction/procurement contract. In terms of Figure 1, the contract is signed between the authority (employer) and the contractor at date t=1. It specifies the works that are to be performed, or the good that is to be delivered by the contractor. The degree of precision in this initial contract has direct implications for cost overruns. Suppose the initial contract is contingency-complete in that it fully specifies all the works that are to be carried out in each possible contingency that may arise during the construction phase. Under such a contract, cost overruns can be avoided altogether. Since every contingency has been completely planned for, no additional (un-specified) work is ever required. Now, if the initial cost is determined for each contingency, there will be no cost overruns. Moreover, the contract price can be fixed or can be contingency specific. In 12 either case, the parties obligations have already been spelt out fully. No additional payment is required to be made by the either party. In fact, delays on the part of contractor can also be

avoided with the help of suitable penalty clauses. Therefore, in principle, complete-contingentcontracts can ensure that project is completed in time and within (contingency-specific) budget. In reality, the initial contract cannot be complete. Parties cannot predict every possible scenario that may unfold during the construction phase. As a result, the initial contract cannot completely specify every relevant aspect of the project works; different states of nature during construction require different modifications in the project works. For example, on a railways project depending on the local conditions, it may become necessary to have more of mannedcrossings or railway-over-bridges than were initially intended. The bounded rationality of the parties along with imperfect forecasting techniques makes it impossible to specify every contingency and the relevant tasks to the last details. As a result, the initial contract leaves out several project works. This is especially true of infrastructure projects, which are inherently complex and have long building phase. Formally speaking, the initial contracts for infrastructure projects are intrinsically incomplete. Once the contractor starts the work at the project site and the future unfolds, the need for additional works arises invariably. Additional works require more funds and hence cause cost overruns. In some cases, extra time is also needed. So, some of the cost overruns are caused by what we have called the contractual incompleteness. Construction projects are typically more complex and therefore more difficult to plan and execute than is the case with non-construction projects, say, those involving purchase of equipments. So, the degree of incompleteness of the initial contracts is expected to be higher for construction projects. As clarified in Section 2, most projects in Road, Railways and Urbandevelopment sectors are construction projects. Majority of projects in Civil Aviation, Shipping and Ports, and Power sectors too involve construction and are complex even otherwise. Each project in these latter sectors is generally unique in terms of its requirements. So, learning from previous projects is limited. If our claim about the causal relation between the contractual incompleteness and the cost overruns is correct, the cost overruns experienced by projects in the above sectors should be significantly larger than by other sector projects. Organizational or Institutional Failures: As argued above, execution of infrastructure projects requires active cooperation of several departments within as well as among various ministries. Government departments are hierarchical organizations. A large body of literature shows that there is a conflict between the individual and the organizational objectives at every stage of hierarchy. As a result, hierarchical organizations are inherently weak in inducing the desired efforts from the people involved.15 This is especially true of government organizations. Therefore, infrastructure projects have to face the consequences of organizational failures within the sponsoring ministry itself. On top of it, these projects need joint efforts of several other organizations. In India, different departments are responsible for different project activities. For example, project implementation, shifting of power lines, water lines, sewer lines, cutting of trees, environmental clearances and other such activities are performed by different departments. Executions of these activities are highly dependent on joint and timely efforts of the departments involved. However, interdependence of efforts means that it is easy for departments to shirk and pass the blame

on others. So, in addition to intra-organizational failures, infrastructure projects in India are vulnerable to inter-organization failures. Several reports, including the official ones, corroborate our claims.16But, how can we measure the implications of these failures? As mentioned earlier, most projects in road, railways and urban-development are construction projects. Projects in these sectors generally require environmental clearance from the central as well as the state agencies. Moreover, compared to those in the other sectors, these projects require much more active cooperation of several departments for land/property acquisition, shifting of power lines, water lines, sewer lines, approval of under or over-passes, etc. Laxity on the part of just one department or dereliction of duty by a few officials can hold-up the entire project. So, these projects are highly vulnerable to delays caused by all kinds of organizational failures. The same is the case with projects in civil aviation, shipping and ports, and power sectors, though to a lesser extent. Majority of projects these sectors too involve construction or setting of network points. In several cases, ecologically sensitive land has to be acquired. This means more regulations and increased vulnerability to inter-organizational failures.

Time Overruns: any delay in implementation in itself should cause cost overrun for the
project. This should happen simply on account of inflation itself. In most cases, initial cost estimates are arrived at using the current input prices. If there are delays, inputs will become more expensive and, in turn, will cause an increase in the project cost. Moreover, certain overhead costs have to be met as long as the project remains incomplete. Delays should increase these costs also. Also, a long delay may cause depreciation of project assets, necessitating expenses on repairs or replacements. This means that in addition to the above factors, time overrun on account of any other factor is also an underlying cause for cost overruns. Economic Factors: Each project is located in some state(s). Several departments of the state government concerned play rather crucial role in project implementation. After all, activities like land-acquisition, shifting of utilities, etc., are performed by the state government concerned. Moreover, economic and geographical features of the state may affect the project time and costs. For example, if a state has better transport, power and telecommunication infrastructure in place, it will be easy to execute projects in the state. Generally, richer states are said to be in possession of superior infrastructure. In contrast, due to law and order as well as difficult terrain, project implementation is likely to be difficult in the North-Eastern states and Jammu and Kashmir. To check statistical validity of these conjectures, states have been clubbed in four categories. Five richest states, in terms of per-capita income, are grouped together.18These are Haryana, Punjab, Delhi, Gujarat and Maharashtra.