L10-accountability Flashcards
how do we solve the 4 failures of implementation?
- the 4 failures of implementation
- the rules are ignored
- the rules are imported
- the rules are broken
- the rules are resisted
conclusion
Accountability
- Monitoring, rewards and punishments can induce people to enforce/respect the rules - But all are blunt and produce imperfect incentives
Electoral accountability
- Voters can punish politicians - But even with true information they often lack the skills, power and collective action to do so
Social accountability
- Direct citizen pressure can be transformative But the poor often lack info, allies and collective action
the problem with developing countries and development
In developing countries the state and institutions are weak:
- Formal rules are not enforced/respected - Rules are ignored by bureaucrats - Rules are imported by donors - Rules are broken by politicians - Rules are resisted by citizens
-> investments are risky:
- Would you build a car factory in Paraguay? Spend your nights studying hard at a university in Liberia? Hire scientists to develop new software in Laos
-> development initiatives often fail
- Nigeria's Vision 2020 had little impact - 1MDB dev bank in Malaysia *harmed* growth - WB projects have no impact on long-run gov
-> how do we change people’s behavior?
= by holding them accountable
= by offering rewards for effective implementation of the rules and punishments for breaking the rules
(bigger stick to punish + bigger carrot/reward to reward good behavior)
defintiion: principal-agent model
= principle hires an agent to do something for them (bc you are busy, you can’t do it themselves) BUT information assymetry (they have more info on what they’re doing than you do)
!agent not doing it for loyalty or smth to the principle, it is bc they have incentives to do a good job, they want e.g. The income
*principle agent is universal, used in many organizations/countries
- Principals employ agents to help achieve development
- But cannot see everything agents are doing (‘information asymmetry’)
- Principals set a contract with incentives (e.g. Threat of being fired, reward of bonus)
- Agents respond to those incentive
!agents have their own objectives
Contracts and hierarchy are central to bureaucracies and development
- E.g. Zambia: ministry of health, underneath the provincial health office, etc. Etc.
Versions of accountability to promote development
E.g. To prevent political corruption we might establish an agency to monitor politicians and pass laws with large punishments - ‘horizontal accountability’
- International Commission against Impunity in Guatamala = int'l experts that would support and conduct prosecution of corrupt officials - When Morales was president, he found out he was being investigated for corrupt financing -> kicks the UN out - So it only works as long as you have support for it
E.g. To make sure teachers work hard we might give a bonus to those that improve students’ scores more - ‘performance-related pay = more money if your students get better grades
- But in practice not much effect
E.g. To make aid work we might pay countries for progress in achieving development targets, not just with grants or loans - ‘cash on delivery’
- Pay only if you achieve development, e.g. For every kid financed you get $1000 - Pay people for outputs, not inputs - Lots of advantages: encourages people to do it in the way that works best in their country (countries find local/homegrown solution = more organic than just importing rules) - Good at: aid is a lever for good governance, the money is irrelevant: we are doing it for development, to boost good governance Problem = it is terrible at financing: money comes after the spending
Why accountability mechanisms often fail:
- Multiple tasks: e.g. Teaching basic skills and encouraging creative thinking: incentives encourage a bias to whichever is easier to measure (teachers will only teach A B C, whilst maybe D is also important)
- Multiple principles: reporting to multiple politicians creates confusion and discourages supervision (e.g. Supervisor, rector, dean = all above teacher in principal-agent relation, but sometimes conflicting messages)
- Nr of people able to access employment under India’s employment guarantee scheme (NREGS) is 9% higher where bureaucrats report to a single politician - Moral hazard: agents take excessive risks knowing they will be bailed out by the principal if things go wrong
- E.g. Local govs in China building ghost cities: they build more than was necessary, national gov bailed them out -
The principal’s objectives: if the principal doesn’t want development, the agent won’t make it happen
- E.g. Zuma: as principle tried to extract money from bureaucracy for own benefit
- If principle doesn’t bother to boost dev, agent won’t either - Compliance is hard in developing contexts
- Poverty and weak state limit how much people can control their response
- Teachers don’t turn up on time bc transport is unreliable -> firing them when they don’t show up won’t fix the problem
- Valid excuses undermine enforcement and compliance
- Nurses absent from 44% of clinics in India in 2003 -> 2006 reform: pay reductions for absent nurses (monitoring with time-stamp machines) + presence increased for 6 months, but then nurses started getting exemptions (e.g. Bc lived far away or were sick) or reported broken time-stamp machines = absenteeism returned to normal = people find ways around accountability - Incentives compete with informal institutions/norms: financial motivation can crowd out intrinsic motivation
- E.g. Israel: fine for picking up your kid late -> no longer felt moral obligation to pick kid up on time - Discouraging risky but valuable investments: where corruption is compensating for a weak state/institutions
- E.g. USAID cut support to Afghanistan’s health sector which was saving 100.000 children per year due to a lack of receipts
- Cracking down on corruption makes bureaucrats risk-averse, e.g. Brazil’s random audits reduced gov spending and worsened healthcare indicators
- Multiple principles: reporting to multiple politicians creates confusion and discourages supervision (e.g. Supervisor, rector, dean = all above teacher in principal-agent relation, but sometimes conflicting messages)
!long list, don’t need to remember any single detail, just overview that there are issues
4 maybe most important (he said most important, then: well maybe most important)
Electoral accountability
Voters are the ultimate principal in democracy
- Politicians are the agent - Elections allow us to REWARD and SANCTION politicians - Demand-side accountability, not supply-side - 'one person, one vote' = each person has say into who is eleced
E.g. The Botswana Democratic Party was voted out of office in oct 2024 after 58 years (where being re-elected before that bc they did well)
- Voters angry at unemployment, low growth, corruption -> voted him out = electoral accountability in action - Replaced by the Umbrella or Democratic Change
Brazil: Where audit reports of corruption are released before an election, Brazilian mayors are less likely to be re-elected
BUT voters are poorly informed about politicians’ performance
- There's only a very weak correlation between prior beliefs and true information Accurate information is necessary for accountability
Can we boost electoral accountability simply by providing true information to voters?
Dunning paper (reading) - shows people don’t have good info on how politicians perform (people over or undercore the performance of “their” candidates)
Wrong info -> punish wrong person
Same article: how do they respond when you give them info?
Field experiments in 6 countries informed voters about politicians’ performance
And recorded who they voted for
‘bad news’ should lead to voting against the incumbent
Information had 0 impact on voter behaviour in any country
No accountability
=> electoral accountability model is not working, just educating people does not make them do a better job -> why didn’t they respond to politicians’ performance?
- Clientelism inverts accountability: politican holds you accountable to voting correctly: otherwise you don't get money - Maybe social norm of corruption -> doesn't matter who you vote for -> maybe partisan alignment or ethnic voting \+stuff on the next card
Electoral accountability as blunt tool:
8
- Political skills are needed to interpret information
- Who is responsible for healthcare?
- ethnic voting: only rewarding co-ethnics, only punishing ‘others’
- Clientelism short-circuits accountability:
- One vote won’t make a difference
- But determines whether you get $10, or a job
- Lobbying and corruption are more important : if you have more money you can campaign more, can transport people to the voting boot
- ‘one dollar one vote’ = election results reflect wealth
- Lack of social contract
- Without taxes, voters feel less stake in punishing bad politicians
- With natural resources and we’re not taxed, we don’t care bc its not our money
- Psychological disconnect: no relationship with the gov to keep them accountable
- Lack of voter coordination/collective action
- Accountability only works if many voters react
- Voters are cynical: ‘all politicians are thieves’
- Your vote doesn’t really make a difference: it doesn’t matter if one person votes for someone, loads need to –> collectively hiring an agent and being collective principle -> how to collectively decide to fire the agent: we all have to agree = diff bc different opinions (difficult to set standard for good performance)
- With Botswana example it worked: achieved collective action, but this is hard
- Backfire
- Why should i be honest if no one else is?
- If they’re corrupt, the system is corrupt, i should get my share to -> give up, accept corruption, normalize it
- In Lagos, anti-corruption messages make citizens more willing to pay a bribe
- Backlash
- Politicians react with competing or misinformation
= political accountability is complex, depends on other institutions
Social accountability vs electoral accountability
Electoral accountability is a ‘long route’ (elections only every ~4 years)
- Relies on politicians responding to voters - AND politicians being able to ENFORCE changes through the state
If there’s no elections in a while -> direct citizen engagement to apply pressure on politicians outside of elections
Definition: social accountability
= accountability through direct citizen engagement
‘voice’ - or client power
The short route of accountability
- Complaints - Writing letters - Querying politicians - Reporting in the media - Citizen/NGO monitoring of services - Protests - Town hall meetings - Participatory budgeting - Parent-teacher associations
= led by civil society
OR by one part of the state against another (e.g. Monitoring agency asks citizens to help gather information on the Ministry of Infrastructure)
Social accountability: problems and responses (examples)
(examples in reading)
Problem: only 20% of gov spending reached schools in Uganda (corruption by local gov officials)
-> response: providing info to monitor school grants: school clsoer to newspaper seller -> headteacher know about grants -> schools receive more of the grant (+80%) -> more students (+20%) and better grades (+6%)
Problem: high infant mortality rates due to poor attendance and care at primary health clinics
-> solution: encouraging ‘community monitoring’ by NGOs led to 33% reduction in Under-5 mortality in Uganda
Problem: infrastructure projects are often abandoned by contractors or building quality is poor
-> solution: an online portal where citizens and NGOs can track and upload progress and photos of gov projects in Kaduna State, Nigeria
-> worked for one year: website and app no longer work, relied on donors and was abandoned
Social accountability often fails
- Community monitoring of road projects in Indonesia did not prevent corruption or poor road quality
- Audits - vertical accountability - were much more effective
- Citizens lack information, skills and confidence
- E.g. For attributing responsibility - Dependent on allies in the state for success
- E.g. Judiciary, legislators, statistical agencies publishing data
- Need info from someone in gov to make it possible: usually one ministry stealing, one to stop it, they ask your help to stop it -
backlash: elites punish citizens who speak out
- Citizens are made ‘passive’ by a powerful state - The rich exit to the private sector
- Rich powerful people are better at getting gov to work, they have more power
- If rich people have already given up, exited the system = you can’t rely on them (e.g. They have private healtcare, private schooling etc. -> don’t care about if the gov is working, don’t care about public sector)
- Reducing pressure to improve public services
- In the Philippines, private schools reduce the quality of public schools - Pressure depends on collective action
- Free-riding: citizens are busy
- One person protesting, submitting a complaint letter or whatever = doesn’t achieve anything
- More numbers necessary to have more power, to get response
Collective action is intrinsically hard
Conclusion
Accountability
- Monitoring, rewards and punishments can induce people to enforce/respect the rules - But all are blunt and produce imperfect incentives
Electoral accountability
- Voters can punish politicians - But even with true information they often lack the skills, power and collective action to do so
Social accountability
- Direct citizen pressure can be transformative But the poor often lack info, allies and collective action
