In countries where electricity outages and “load shedding” are a regular feature of life and a major obstacle to jobs and economic growth, failure to achieve reforms has often been attributed to governance and incentive constraints. In new research, we find systematic empirical evidence consistent with case studies and experiential anecdotes showing how governance failures and public-sector incentives impede reforms.
 

Main Insights

Using the Utility Performance and Behavior Today (UPBEAT) data covering 190 utilities from 93 countries over the period 2012-2023, we find that reliability is worse when there is greater corruption in the political system as measured by the International Country Risk Guide (ICRG).

These results are not driven by variation in income per capita. Yes, reliability is worse in countries that are poorer (Figure 1). But the empirical estimates show that corruption indicators are significant in explaining variation in reliability even after controlling for income.
 

Figure 1: Outage Duration (left) and Frequency (right) against GDP per capita

These results are also not driven by privatization reforms. Yes, utilities under private ownership tend to perform better than public utilities, but the gap in performance is smaller in countries with greater control of corruption (Figure 2). The empirical estimates show that the same utility, regardless of ownership, would likely perform worse over time when it operates in a context with higher corruption in the political system.
 

Figure 2: Outage Duration (left) and Frequency (right) by Ownership and Corruption

As researchers, we were surprised to see the cross-country data revealing significant correlation between corruption and reliability because there is limited variation in corruption measures, especially changes over time within countries. When we plot the reliability data against corruption indicators, we see many countries clustered at intermediate values, with few at the higher end of control of corruption (Figure 3). And yet, across multiple empirical specifications (reported in the paper), the statistical estimates are robust and significant.
 

Figure 3: Outage Duration (left) and Frequency (right) against Control of Corruption

For many practitioners with deep experience in the energy sector, this evidence could evoke a “we knew this already” response, as suggested by the extensive case literature linking corruption and sector performance. The systematic significance of corruption indicators in large sample data underscores the value of these case studies, reflecting the experiential knowledge of technical experts who have grappled with political constraints to reforms.

Drawing on these case studies, we offer the following hypotheses about the mechanisms through which challenges in political institutions matter for reliable electricity:

• Hypothesis 1: “Grand corruption” among elites at the helm of government agencies undermines the overall system-wide planning, and  the implementation of reforms such as privatization, which may be selectively applied and insufficient to deliver reliable electricity across the system. 
• Hypothesis 2: Norms of non-payment of bills by powerful agencies and widespread theft of electricity in society, often in collusion with frontline utility staff, erode revenue, compromise technical efficiency and operational discipline.
• Hypothesis 3: Weak incentives of utility managers and staff, especially if jobs in the public sector are driven by clientelism (political patronage), reduce accountability for performance.
• Hypothesis 4: Salience of privately targeted, short-term benefits (e.g., free or subsidized electricity at election times) in political contestation militate against reforms.
• These channels can be sequential or overlapping, and are not mutually exclusive, with the nature of political contestation for power being at the root. Testing these hypotheses is outside the scope of our current paper, which focuses on analyzing cross-country variation using available data. Generating credible evidence on the channels through which governance and incentive structures matter, and how they can be overcome, requires bespoke data in specific country-contexts.
   

Overcoming the Underlying Governance Problem

We propose a forward-looking agenda where these hypotheses are tested by moving from generic reform prescriptions to context-specific policy experimentation with the following governance interventions:

➡️	Measure internal culture, frontline staff incentives, and management practices, leveraging the data for internal deliberations to transform motivation and performance within utilities.
➡️	Invest in customer outreach to shift norms around theft and bill payment.
➡️	Use accountability mechanisms at the “first mile” of executive government—in political institutions of local government—to align incentives with reform and curb harmful interference in the energy sector.

Strategic Shift in the Policy Approach is Needed

Rather than expecting standard reform packages to work in politically constrained settings, we propose an approach of “learning by doing” with policy experimentation and impact evaluation of governance interventions in partnership with reform-minded  leaders. There are few proven answers on how to improve governance in the energy sector, but practical ideas exist that can be tested to make reforms feasible even in difficult political contexts, so that electricity can be relied upon to grow economies and jobs.

Source : World Bank