Photo courtesy of AP

What role should citizens expect a government to play in public life? The theoretical case for state intervention in the economy in the public finance discourse is based on an unstated assumption: that the state is highly functional and well governed. Public Choice theory does introduce the concept of government failure, but what of states where failure is endemic?

In states that are afflicted with such a problem the question arises as to whether state intervention will be in the public interest. There are many recent instances of policy failure in Sri Lanka, the ban on agricultural chemicals and fertiliser is a classic example. How could such a grievous error have occurred? This is not an isolated incident either, ill-conceived policies are frequent in the economic and the social sphere; the ban on burial for Covid deaths, the family background report that restricted female labour migration and the ban on turmeric imports are other examples.

Government failure does occur even in advanced countries but these are occasional lapses or errors. The frequency of errors in Sri Lanka however raises a question: does the Sri Lankan state suffer from systemic failure? Sri Lanka suffers from many weaknesses in its institutional architecture that have been extensively discussed from a legal perspective. Do these weaknesses contribute to policy failure? This essay examines these weaknesses to provide a possible explanation for the frequent failures in policy. It is offered here only as a hypothesis, a possible explanation for  frequent policy failure and is based on the observations of the author.

Policy failure and accountability

Failure can originate in poor policy selection, design, implementation or a combination of these. Misaligned incentives can result in poor policy selection or design; inadequate consultation or technical skills can result in bad design. Even if a policy is well designed, implementation cannot be taken for granted in a complex policymaking environment, substandard execution can result in inferior or even harmful outcomes.

Before the policy making process can be examined, it is necessary to start with a more fundamental political question: what is the role that the state should play? The author suggests that the question be approached from a purely pragmatic perspective. In the author’s view, the level of state intervention should be determined by two critical factors:

  1. the degree to which citizens are able to hold the state to account, and
  2. the extent of state capacity.

A state that is accountable to citizens is more likely to adopt policies that improve public welfare, one that has efficient administration has the ability to design and implement policy properly, ensuring good outcomes.

Accountability mechanisms

Tarunabh Khaitan [I] uses a three dimensional framework to gauge the accountability mechanisms liberal democratic constitutions put in place to check the political executive. These operate vertically, by demanding electoral accountability to the people; horizontally, by subjecting it to accountability demands of other state institutions like the judiciary and fourth branch institutions; and diagonally, by requiring discursive accountability by the media, academics and civil society.

Vertical accountability refers primarily to elections; horizontal accountability refers to mutual scrutiny by other bodies of state; the legislature, parliamentary committees, courts, public audit, independent commissions, ombudsmen etc. More broadly other independent bodies such as an independent Central Bank provide important checks on executive excess. Diagonal accountability refers to scrutiny by the media and civic engagement.

For sound public policy horizontal accountability is critical. The bodies that perform these functions are specialised, permanent and engage with government on a day-to-day basis. They are in a unique position to ensure citizen centric public policy but to function effectively they must be independent and empowered.

Public policy must try to maximise the enhancement of public welfare but there are a number of reasons why this may not happen which are summarised below and discussed thereafter.

  1. Policy is not made by citizens themselves but by their representatives who must act on their behalf. This gives rise to a problem of agency; the interests of the representatives may differ from those of citizens and unless constrained, the agents may pursue their private interests.
  2. Horizontal accountability mechanisms are the principal means for constraining public agents on an ongoing basis. This is achieved through the balance of power between different arms of government; chiefly the executive and the legislature. Within the executive itself, it is necessary to strike a balance between elected and appointed agents – to overcome an excessive focus on short-term policy that is driven by the electoral cycle. Although an independent, permanent bureaucracy is is not an accountability mechanism it can provide important checks on the executive and is essential for proper policy formulation and implementation.
  3. Vertical accountability: elections are integral to the system of checks on power but can only work in conjunction with other mechanisms. They provide periodic accountability but cannot provide routine scrutiny of policy which must be carried out through horizontal accountability mechanisms. Elections are also subject to short term swings of popular opinion and in a heterogeneous society can give rise to identity based politics. They may also be undermined by clientelist politics where politicians focus on distributing highly visible benefits to patrons who are able to use this to influence a broader section of voters to ensure reelection.

Where clientelist and identity based politics prevail, mechanisms of diagonal accountability are less effective since voters become less concerned with promises of public services or abstract discussions of policy.

Accountability is necessary to ensure that politicians work broadly in the public interest but there is a need to balance responsiveness with responsibility. Responsiveness is necessary but some popular polices may provide benefits in the short term and may be harmful in the long term. Some unpopular polices may bring long term benefit but carry short term costs. Policy needs to be responsive to public concerns but they must also balance short term and long term costs and benefits, the horizontal accountability mechanisms must ensure this.

Ultimately the quality of public policy depends on the incentives and pressures faced by politicians and the process by which policy is made.

Agency problem in government lead to clientelist and identity based politics

The central problem is one of agency, once appointed, the agents (politicians) may not act in ways the benefit the principal (the public). Adserà, Boix, and Payne (2000) [ii] identify the problems that arise when a principal (the public) delegates decisions to an agent (policy-makers); differing interests and informational asymmetries between the principal and the agent.

First, policymakers and voters may have interests at odds with each other: the former may be simply interested in enriching themselves while in office; or, even if they are honest, their ideas about what enhances the welfare of the public may differ from what the public itself wants.

Second, the principal and agents may differ in their corresponding levels of information about the state of the world, the policies to be pursued and their welfare consequences. If the public is less well informed than the policymaker, the latter can more easily impose her preferences or even exploit the public. In short, the delegation of decision-making and policy implementation responsibilities, a “must” in modern representative democracies, automatically opens up the possibility for significant inefficiencies and corruption among political practitioners. (Adserà, Boix, and Payne, 2000)

Informational problems and the political incentives for the provision of public services

Information asymmetries lie at the heart of the agency problem. Keefer and Khemani (2005) dissect the information asymmetries within the agency problem to explain the reasons why contested elections can fail to create accountable government. Their analysis provides valuable insights into the incentives to provide public services faced by elected representatives.

Their argument in summary is: due to the inherent complexity and long time horizon over which they make take effect, citizens find it difficult to evaluate the quality of public services or the role that a politician may have played in their provision. As a result, political promises based on improved services carry little credibility. Instead citizens rely on tangible benefits to evaluate politicians. As politicians are unlikely to be rewarded for  improved services they focus on providing more tangible short-term benefits. This leads to clientelist politics, the distribution of handouts for votes.

Keefer and Khemani identify three imperfections in political markets:

  • lack of information among voters about the performance of politicians,
  • social fragmentation among voters manifested as identity based voting,
  • the lack of credibility of political promises

Information limitations reduce citizens ability to hold politicians accountable and encourage them to cater to special interests instead of providing broad social services.


With respect to information, citizens cannot easily evaluate the quality and efficiency of public services. Even when they can do so, they cannot easily determine who is responsible for this:  service providers, higher level ministry officials, or politicians. Since the connection between politician and policy is obscure  means that politicians have difficulty in making credible promises with regard to improved service.

The key imperfection that is that citizens cannot observe the contributions a specific politician makes to policy or the connection between policy and their welfare. Quality of services depends on behaviour of day to day providers than on the easily observable actions of remote politicians. Although policy decisions by politicians – teacher management, preventive health care initiatives can influence provider behaviour and incentives citizens still need to know which policies have contributed to better quality services.

Second measurable benefits or costs may not emerge until several years after the policy has been taken making it difficult to difficult for voters to reward or punish politicians within the electoral cycle.

For example if teacher quality or attendance has improved voters cannot distinguish whether it happened due to their own pressure on the teacher, because the teacher decided to do a better job, because of generalised reform in teacher quality or because of a targeted intervention of the politicians. Imperfect information means citizens have difficulty in assigning credit or blame to the politician.

Politicians get some credit for easily observable buildings and jobs but little for the quality of services available.

Credibility of political promises and clientelism

Due to this information problem, Keefer and Khemani argue that politicians can make credible promises only to voters with whom they have built a personal reputation.  Credible politicians are, for example, locally influential people who have helped families with loans or jobs or assistance with legal or bureaucratic difficulties. The promises of such influential people are what people can rely on so national politicians use such locally influential people (patrons) to mobilise votes.

It makes little sense for politicians to promise reforms that benefit large numbers of voters when only a few voters believe these promises, they can only make credible promises to narrow groups of voters.

Patrons strengthen their position in a community when they are able to deliver tangible benefits to their clients but have weak incentive to provide public goods because they benefit non clients as well as clients. Promises of buildings and jobs become the currency of political competition at the expense of universal access to high quality education and health care.

If the poor are uninformed on the impacts of policy, broad social services can also suffer as politicians make concessions to organised and informed special interests. Special interests, single businesses, groups of manufacturers, farmers, public employees can influence policy in their favour but if the poor are unable to see the connection between the diversion of resources to special interests leading the lack of broader services will not act.

Khemani (2004) finds that as elections near, state governments in India increase expenditure on easily observable public investment projects and decrease them on more broad-based public services. State governments In India also increase targeted tax breaks to narrow groups of producers before elections.

Social polarisation and service provision

Even in the best of times a heterogeneous citizenry that differs in income, ideology, religion, language will attach different values to different public goods that makes it hard to determine optimal policy.

In divided societies identity based politics of ethnicity and religion can come to dominate political discussion at the expense of questions of social services. Studies of electoral politics in India show that identity characteristics along ethnic, linguistic,and religious lines dominate political behaviour (Weiner and Field 1974). This is also true of Sri Lanka.

Like uninformed voters polarised voters are less able to hold politicians for their performance in office.

Responsibility and responsiveness

Representative democracy entails governments that are both responsive and responsible. Peters and Linde 2018 [iii] define responsiveness as the short-term match between what people want and what they receive from government in terms of policies, responsibility involves policy that serves more long-term interests that have not as yet been articulated as demands from voters. They argue that a certain level of responsiveness allows governments to build a “buffer” of support that will allow them to take decisions that are not necessarily responsive but responsible.

The authors view is that in flawed democracies, where relationship with voters is patron-client, similar forces work in a distorted fashion. If the horizontal and diagonal accountability mechanisms are weak, government policy can focus on distributing rents to clients instead of social services. This builds a reservoir of support which allows politicians to pursue personal objectives that do not improve public welfare. Therefore in flawed democracies a minimally responsivegovernment can behave irresponsibly and pursue policies that have little to do with public welfare.

Horizontal accountability and policy formulation

The preceding discussion would appear to suggest that elections can fail to create accountable governments. Yet, many advanced countries seem to have governments that are accountable and able to formulate good policy. This is due to the horizontal and diagonal accountability mechanisms that are present in a functioning democracy that come into play during the process of policy formulation.

Public policy is complex, goals are not easily determined because different constituents will demand different things. It is also necessary to balance short and long term objects, resource limitations and practical problems of implementation. People may also react in unpredictable ways that may negate the objectives. It requires adequate consultation, careful study and be free of influences from vested interests.

The quality of policy depends on the process by which it is made. Good policy requires not only government but also the involvement of the legislature. Debates in the house allow MPs, especially from the opposition to discuss policy and voice the concerns and interests of their constituents. They can refer matters for more detailed scrutiny through parliamentary committees.  Although unseen, the bureaucracy is heavily involved. The identification of problems, the analysis of data, the development of possible solutions is done mostly by the bureaucracy. A permanent, meritocratic bureaucracy, independent of the political executive is a vital element in the process. Collectively these bodies create horizontal accountability.

Senior bureaucrats have direct experience with the operation of public policies. The experience and knowledge enable them to argue from positions of great strength about the financial and administrative difficulties of policy proposals, the repercussions likely to be encountered from the affected groups and alternatives for dealing with policy problems. The role of the bureaucracy in policy making is informative, suggestive and analytical.

The bureaucrats’ position is further strengthened by their permanence in the administrative organisation as compared to the frequent rotation of a minister. The average time spent by a minister with a department is much less than the average time spent by a bureaucrat. The minister’s tenure in the office very often falls short of the time required for a policy to be formulated, implemented and evaluated. Bureaucrats however are normally appointed for a career in the public service. This puts them in the position to acquire vast knowledge of a specific public sphere. Due to their expert knowledge of the work done in their departments; of the results and impacts of existing legislation, and also because they can devote all their time to the administration of their departments, they hold a unique position in the system.

The bureaucracy is the permanent and professional part of the executive wing of the government. An independent and meritocratic bureaucracy is neutral and is not affected by the change of political parties. Since they are permanent they are able to take a long view on the consequences of policy and form an important check on the short term impulses that drive elected representatives-both from the government and the opposition.

Centralisation of power and weak accountability could be the reason for policy failures 

The core of the problem in Sri Lanka is the acute weaknesses in horizontal and diagonal accountability mechanisms. The country suffers from an over powerful executive that dominates all branches of government resulting in only minimal level of horizontal accountability.

In the Westminster tradition, the Cabinet operates in an atmosphere of collegiality; the Prime Minister is only first among equals and while some may be dominant figures a degree of discussion does take place. Decisions are taken as a group and they are collectively accountable to Parliament. Under Sri Lanka’s Executive Presidentialism in effect, all authority rests with the President who has no requirement to consult anyone and is only nominally accountable to Parliament. In general, when dealing with complex problems several heads are better than one, this is especially so in a crisis. Sri Lanka’s parliament is weak and the practice of crossovers means the government is able to ensure a permanent majority by offering positions.

This leads to a highly personalist system of government with decisions taken with little consultation or discussion with the cabinet, legislature, bureaucracy or other stakeholders.

“The presidential system encourages a personalised style of politics that favours charismatic politicians or populists and such politics is often at odds with the basic norms of constitutionalism.” (Linz, as quoted by Edirisinha) [iv]

Edirisinha also argues that the shift of power away from Parliament has devalued both Parliament and Cabinet which is no longer the focal point for policy debate and formulation.

This may explain the many ad-hoc decisions; strikingly evident in some of the U turns in policy, for example tile imports being banned, then the ban being lifted and then banned again within the space of a couple of days. This also reflects the relative influence of special interests; two business groups, the manufacturers and the importers, lobbying for different policies.

The legislature’s powers of scrutiny are weak, for example the committees where much of this happens can only make observations, they cannot compel the executive to act. They suffer from a lack of research capacity and are under the control of government representatives.

The bureaucracy is neither permanent, independent or meritocratic. Although a tiny minority enter through competitive exams the majority are political appointees whose loyalty is to the politicians who appoint them, not the public or the constitution. The selection is based on political favours not merit. As the career prospects of even those independently selected are in the hands of politicians many fear to act independently, those who do are often penalised. The lack of meritocracy means its technical skills are weak, politicised appointments and political control over it means that it is subservient to the politicians.

In short, all the elements of horizontal accountability that work together with vertical and diagonal mechanisms to check and shape policy are weak or non functional. With such weak institutional arrangements it is hardly surprising that policy is rarely in the public interest.

The evolution of Sri Lanka’s political institutions

The preceding discussion has attempted to lay out the weaknesses in Sri Lanka’s institutions which may explain the frequent failures of public policy. Yet in the early decades after independence the state succeeded in implementing good policies in health and education. How was this possible?

Post-independence Ceylon had both accountable government and an excellent independent administration, the country’s high social development indicators are a legacy of that institutional arrangement.

The dismembering of the administrative system in 1962, the loss of its independence in 1972 and its subsequent politicisation severely diminished a critical element in formulating and implementing policy. The gutting of constitutional systems of accountability in 1978 reduced the role of the legislature from an effective check and counterbalance to little more than a rubber stamp. The centralisation of power and absence of counterweights opens the door to the influence of vested interests in policy. In the absence of accountability, corruption and rent seeking have flourished, distorting the incentives of policy makers. The net result is systemic decay of the institutional arrangements of sound government.

The public presumably want democracy to mean greater popular participation in decision making and economic social advancement but are repeatedly disappointed fundamentally due to the lack of accountability.  As elected politicians dominate, the entire system of government is driven by short-term impulses. The politician must impress the voters to ensure election but once elected, face no pressure to advance their interests.

Since other checks on power are minimal, accountability to citizens is limited to elections but the system has been perverted by patronage and clientelism. Unfortunately the practice of clientelism fits in with the essentially feudal nature of Sri Lanka’s society. Loyalties to family and caste are at the very heart of Sri Lankan politics. People have traditionally looked to influential ‘notables’ in their area to provide solutions for their problems. The patrons used by politicians fit perfectly into this structure.

Over the past two decades clientelist politics have been financed through debt or money printing which has allowed largesse to be distributed at little visible cost. Employment within the public sector has doubled, large scale construction projects have raised the profile of their sponsors but as none of this has been funded by taxes, citizens have experienced mainly benefits not costs and have been happy to vote in governments that have made promises of increased spending.

Similarly, in trade policy catering to narrow special interest groups (from business to agriculture) to the detriment of the wider group of consumers has proved beneficial.  The short term benefits from protectionism are visible to the producers but the long term costs in terms of higher costs to consumers, lowered productivity and reduced economic growth are not.

The separation of powers could have averted the current debt crisis. An independent Central Bank guided by clear rules, a legislature aided by fiscal rules on deficit and debt empowered to scrutinise government spending supported by a powerful Auditor General could have prevented the expansion of the deficit and its funding through money printing.

The costs are now becoming apparent in the macroeconomic imbalances that have lead to the ongoing balance of payments and debt crisis but connecting these to their causes is a difficult task for citizens.


This essay argues that the question of intervention by government in a market must be informed not only by the instances of market failure or questions of equity but also by the quality of government itself. While government failure is discussed in the literature, the problem of systemic failure deserves special consideration in countries where governance is poor.

The essence of the problem is the difficulty in constraining the agents appointed by people to act in ways that benefit the citizenry. This depends on the accountability mechanisms that are found within the constitution and the structure of government which may be classified as vertical, horizontal and diagonal. The most important from a policy perspective are the horizontal mechanisms, which are able to exercise regular scrutiny over government. Vertical mechanisms may be weakened by information asymmetries and undermined by clientelist or identity based politics.

Sound policy may be described as carefully thought through interventions designed to improve public welfare. This essay provides a possible explanation for the rarity of such policies in Sri Lanka. It is however an untested hypothesis.

[i] 2021. Killing a Constitution with a Thousand Cuts: Executive Aggrandizement and Party-state Fusion in India. [ONLINE] Available at: [Accessed 14 June 2021]

[ii] Adserà, A., Boix, C. and Payne, M. (2000). Are You Being Served? Political Accountability and Quality of Government. [online] Available at:

[iii] Linde, J. and Peters, Y. (2018). Responsiveness, support, and responsibility. Party Politics, 26(3), p.135406881876398.

[iv] Constitutionalism and Sri Lanka’s Gaullist Presidential System. (n.d.). [online] Available at: