Monday, July 2, 2012

Cost and time over-runs in road projects

A recent oped in Businessline had some stunning figures on cost and time over-runs in road projects in India. Thillai Rajan and Govind Gopinath use data from the central road projects to find that
Nearly 69 per cent of them are riddled with cost overruns and 89 per cent of them are faced with time delays... The average cost and time overruns have been 16.25 per cent and 41.2 per cent, respectively... Had the projects been developed within the budgeted cost, we would have been able to build 16.25 per cent km of roads more at the same cost... if there were no delays, we would have been able to add 41.2 per cent km of roads more in the same period. Since overruns can usually be controlled by better management without any substantial infusion of capital, these numbers indicate the potential for achieving additional road capacity without additional investment.
Their findings with respect to public private partnerships (PPPs) in road sector, which has had the largest number of PPPs in India, are strikingly contrarian,
It was seen that 88.1 per cent of PPPs had cost overruns as compared with 54.37 per cent in non-PPP projects. However, a higher proportion of non-PPP projects had time overruns (92.23 per cent) as compared with PPPs (80.95 per cent). While PPPs have a superior track record in managing time overruns because of private-sector efficiency (and probably also because of the inbuilt incentive to complete projects faster), it was not the case as far as cost overruns are concerned. The average percentage cost overruns in a PPP is close to three times that of what was observed in a non-PPP project. These numbers achieve even more significance if we consider that the average project size of a PPP (Rs 382.46 crore) is about a half more than that of a non-PPP project (Rs 253.89 crore).













These are very illuminating figures and should inform policy making within the National Highways Authority of India (NHAI) in particular and in the road sector in general. Two very fundamental takeaways,

1. The blind belief in PPPs as the panacea to our road sector needs is clearly misplaced. On both fronts, PPP projects are as bad as non-PPP ones in addressing the twin challenges of cost and time overruns. In fact, as the authors write, inflating costs to boost bottomlines from such projects, so widespread across the world and in other sectors, appear to be happening in roads in India.
 
2. Both contracting and project management principles need makeovers. The former, in particular, is important because the current incentives, as the authors indicate, are clearly not aligned towards mitigating cost overruns. As the case of PPPs indicate, given the pervasiveness, to the point of inevitability, of contract renegotiations, the private sector appears comfortable with cost overruns on their balancesheet. In simple terms, the underlying financial assumptions of roads contracts need revisiting and project management, especially land acquisition, need tightening.

In road projects, it would be interesting to examine the specific reasons for cost over-runs. An NBER working paper by Eduardo Engel, Ronald Fischer, and Alexander Galetovic examined data on Chilean experience with re-negotiations of 50 PPP projects in the 1993-2006 period and finds four observable predictions - "(i) in a competitive market, firms lowball their offers, expecting to break even through renegotiation, (ii) renegotiations compensate lowballing and pay for additional expenditure, (iii) governments use renegotiation to increase spending and shift the burden of payments to future administrations, and (iv) there are significant renegotiations in the early stages of the contract, e.g. during construction."

Further, they also found that of the $2.3 bn re-negotiated and awarded bilaterally, 84% corresponds to payments for additional works, while the remaining 16% correspond to additional payments for works included in the original contract. The potential for additional works (or quantities of works) are especially high in road projects where alignment changes to accommodate problems in land acquisition (like litigation and popular opposition) are commonplace. If the cost over-runs due to such problems are rationalized (if payments are made at the same unit rate as in the original contract), the concern should therefore be with limiting additional payments on the original contract rates. But the latter forms a small share of the total renegotiated contracts.

In view of this, the low hanging fruit is in reducing the time over-runs. The massive variations in the extent of time over-runs between road projects in various states, and in particular the high percentage of time overruns in states like Bihar and UP, hold important lessons. Since other states manage to complete projects with many times less time overruns, it is important that measures be taken to shore up project management practices in the laggard states, both at the government level and private contractor levels.

3 comments:

Anonymous said...

Funny that the study did not take into account quality of roads. From my experience in India and the U.S., I can tell that over 50% of roads in India are substandard and do not comply with the norms. The result is that these roads are to be redone more frequently than required. A good way to analyze the road construction efficiency would be to do a cross-country comparison of cost, quality (compliance), time taken, and longevity.

Vikash Sharda, Chennai said...

Study does not look to be factually correct. If time overrun is 41 months which is equivalent to around 3.5 years. Inflation cost increase itself would be around 21% assuming inflation of 6% p.a. Average cost increase of around 16% above in less than even inflation cost increase which is not correct. Then you have cost overrun due to scope change and design change which should also be added to the above cost increase of 21%. It appears that study has not looked into the cause of cost overrun and impact of estimated cost.

Govind said...

This is because revised cost estimates take into account inflation figures and thus, the 16% over and above inflation and primarily caused due to operational inefficiencies