Ensuring good performance by boards of SOEs
We are in the season of appointments: of Ministers, State Ministers, Chairmen, Directors, etc. There is always hope that the appointments to the Boards of State-Owned Enterprises (SOEs) will be different this time around and will yield good results. Some will. But many will be seen as falling short. Is the problem structural? Does it require a remedy more complex than simply running names through a committee?
In February 2015 in these pages, I proposed performance contracts for appointees to the Boards of commercially-oriented SOEs (Accountable Management of Government Owned Enterprises, Sunday Times, February 1, 2015).
Last year, I was appointed as Chairman of the ICT Agency of Sri Lanka. This was a different kind of SOE. It was 100 per cent owned by the government, but was not a commercial enterprise like SriLankan Airlines or Sathosa. It was established to provide “free” services to other government agencies and to the private sector, first using a World Bank Credit and then Treasury funds. The Export Development Board and the National Science Foundation are examples of similar organisations. Devising performance indicators for leaders of such SOEs is the focus of this article.
Challenges of devising performance contracts
No one asked me to sign a performance contract. All I had was a self-imposed simple condition: I was not going to take any payment until I had stabilised the agency’s finances.
For performance contracts to be meaningful, the specified obligations have to be within the capabilities of the contracting party. When defining performance obligations in contracts for senior managers I ensured that the key deliverables did not have critical dependencies.
It is relatively easy to set measurable performance targets at the lower levels of an organisation. In profit-driven private enterprises, it is possible to devise objective targets even for those at the top. For example, it would be possible to measure performance of leaders using composite indicators that include elements such as return on investment and share price.
Conventional ways
Take a conventional government department with a simple mandate: manage large volumes of records in multiple media in a conveniently retrievable manner. This would require the mobilisation of resources including physical and technological facilities and people with the necessary skills and attitudes. Unlike in the private sector, personnel decisions require multiple prior approvals from a Ministry and the Public Service Commission. Procurements require Ministry, and in some cases Cabinet involvement. These approvals can take months, “following up” (a perennial activity in government), multiple interactions and modifications etc.
Given these constraints, it would be hard to hold the CEO accountable for timely delivery of outputs. This may be why formal performance indicators rarely apply to heads of government departments. Instead, they are held to informal performance measures, mostly relating to keeping the staff happy, avoiding disasters and scandals, organising the right kinds of ceremonies with due prominence to appropriate dignitaries and so on.
With the state entering into direct provision of goods and services in the 1960s, the departmental model was found impractical. State corporations were created to reduce the rigidities, but they too became sclerotic. Instead of the Public Service Commission, they were subject to the Department of Management Services and the Salaries and Cadre Commission. Their decisions are subject to second guessing by various government functionaries.
When designing e Sri Lanka, one of the world’s first integrated ICT development programmes back in 2002, we realized that successful implementation would be impossible under the constraints of government personnel-management rules. I had, by this time, seen the Salaries and Cadre Commission frustrate attempts to build the Telecom Regulatory Commission as a credible and efficient regulatory agency. Therefore, Cabinet dispensation was obtained to establish the ICT Agency as a special-purpose vehicle exempt from rules that precluded recruiting the right kinds of people from a sector with high salaries.
Following the end of the World Bank Credit, ICTA became completely dependent on state funds. By the time my fellow Board Members and I were called in to restructure an agency in crisis, the pressure to subject it to government approvals with regard to salaries and cadre was intense. But it was the considered opinion of the Board, which was also shared by several informed Ministers, that it would be impossible to fulfill ICTA’s mandate by making it a conventional SOE.
Restructure
Our restructuring plan proposed a block grant to support core operational functions of an agency with a 5-year life span, including cross-governmental platforms and serving as an internal think tank on ICTs for all levels of government. This would not be adequate to cover all operational costs and would actually decline annually, creating incentives for ICTA to raise funds from other government entities to design and deliver ICT-related projects consistent with the National Digital Policy.
The contracts with other agencies would include within them the needed performance incentives and indicators. If they were not properly executed, there would be consequences, including withholding of payments. The performance of the specific task managers would be tied to successful execution. Senior managers would be rewarded on the basis of projects they brought in. The Board would be accountable for the operational block grant and for the core projects.
But that is the theory. The experience of the past 20 months showed that most of the energies of the Board had to be expended on remediation of inherited problems, including previous underpayments of operational and project invoices. When amounts that have been budgeted for and properly invoiced are not paid on time or not paid at all, suppliers are badly affected. Small firms can go out of business. Others, knowing the risks, will either desist from bidding for work or will include risk premiums in their bids. The public loses in all cases. The Board and management are reduced to fire-fighting and “following up.”
In these circumstances, it is extremely difficult to set realistically achievable performance targets for the Board and the CEO. For example, Cabinet approval of the restructuring plan within a year (actually Cabinet had given only three months for this complex task) may have appeared a realistic target. But in actual fact this was not achieved even after 20 months due to no fault of the Board. It was envisaged that restructuring would be approved by December 2018, eight months after the Board started work. But this was disrupted by the dissolution of the Cabinet in October. The Plan was submitted to the Ministry in February 2019 but it only made it as far as the order paper of the last Cabinet meeting of the last government, after months of delays outside the control of the Board.
Projects
Completion of projects would also be a difficult performance indicator. When suppliers are not paid on time for completed components, it is difficult to motivate them to work on the next phase. The wonder is that any work gets done in government under such circumstances.
I had set myself a rudimentary performance indicator. Because staff had not been paid on time on many occasions, I declined the sitting allowances for Board Meetings until the flow of operational costs on which salary payments depended has been stabilised. Seemed simple enough. But even that condition could not be met.
Performance contracts are essential for accountability and good performance. But experience shows they aren’t enough by themselves. Complementary reforms are needed to make performance contracts effective. Remove the critical dependency
Can you imagine running a private company in which all the funds needed for operations are in accounts from which one cannot withdraw funds? The management has recruited employees and entered into contracts with suppliers assuming the funds are available, but their release is sporadic and arbitrary. The firm has to employ extra liaison officers who have been to the same schools as the bank managers and on occasion the senior managers themselves have to visit the bank and plead for funds. Under these circumstances, it is inevitable that salaries cannot always be paid on time and that suppliers are perpetually unhappy, with some refusing to do business any- more. This is the reality of managing SOEs which are beholden to Treasury for revenues.
To implement meaningful performance contracts on the boards of such SOEs it is also necessary to provide them with guarantees of assured and predictable revenue flows. Budgetary allocations cannot be indicative. Boards must be able to make plans and honour contracts.
What happens now is workaround after workaround. Shortly after assuming duties as Chair of ICTA, I was told that funds could not be released to pay salaries due at the end of the month. I exploded, asking why I had been invited to rescue an organisation if the first thing I am told is that operational funds will not be released. I was then told that it is customary, though against the rules, to borrow from other funding lines to pay salaries in such circumstances.
Most people would agree that salaries must be paid on time. If that is a priority and operational funds are late or not forthcoming, all that a Board can do is borrow internally from other revenue lines. This is how a US$1 million prize awarded to ICTA which was to be used for citizen outreach and digital literacy came to be used to pay salaries, before we assumed office. This is also the reason the National Intellectual Property Office (NIPO) is complaining in every available forum about ICTA not delivering a server after the required funds had been transferred three years ago (again, before my Board was appointed).
Winning ways
As part of the rescue, I demonstrated to the Foundation that had awarded the prize that ICTA was performing all the promised outreach functions using other funds, though the prize itself had been used for salaries, and succeeded in closing off that matter. Unless the unpaid operational funds are not settled or some other arrangement is not made by Treasury and the Ministry, there is no solution for the NIPO problem. NIPO cannot write off the funds they transferred to ICTA and ICTA cannot magically materialise Rs. 32 million to buy them the computer hardware they wanted back in 2016-17. ICTA has offered them services from the Lanka Government Cloud equivalent to what they would have got from the servers but even this interim solution has been rejected.
If the budgeted funds are provided on time, there would be no need for workarounds and disputes. There would be no need for the Board to spend its energies on resolving ancient disputes.
The Board and management of ICTA performed well during the past 2 years, in less-than-ideal circumstances. Had the critical dependencies been removed or at least reduced, we could have done even more.