No electricity safeguard survived on paper alone: mechanical rules, real operating control and credible enforcement worked; executive discretion broke the rest
Asked (summary):
Which policies and institutional designs have actually prevented politicians from misusing state-owned electricity utilities in developing countries, and which failed? Named countries and dates for independent regulators, automatic tariff formulas, privatization or concessions, manager performance contracts — and cases where governments overrode, captured or abolished these safeguards.
The evidence is 31 dated cases across 20 countries, drawn from 25 independent source hosts, spanning reforms enacted from 1982 (Chile) to 2021 (India, Liberia) and overrides as recent as April 2025 (Uganda). It is a curated comparative set, not a census of every developing-country reform, and raw case counts summarize this purposive evidence rather than estimating causal effects.
Outcomes by safeguard family
Each square is one case; hover or tap a square for the country, safeguard and outcome. Independent regulators on paper failed most often; privatizations and concessions produced no outright failures in this set but half ended mixed, usually through later political reversal.
Worked — gains achieved and heldMixed — real gains, later override or unresolved strainFailed — safeguard overridden, captured or abandoned
Timeline: enactment to override
Each line runs from the safeguard's start year to its override or end year; a dot alone marks a case still standing (or dated only by its override). Four cases with no dated start are plotted at their override year.
What the counts cannot say
Kenya's 2005 fuel-cost pass-through, combined with tariff increases and efficiency gains, helped eliminate the sector's quasi-fiscal costs by 2009 — affordability was handled in the budget, not hidden in utility losses. imf.org
Peru's 1993 partial privatization with incentive regulation raised productivity from 415 to 1,210 customers per employee, lifted coverage from 48% (1992) to 80% (2007) and cut distribution losses from 22% to 8.2% — gains that held because tariff rules and incentives were contractual. esmap.org
Uganda shows both edges: its regulator ERA (1999) attracted investment and cut distribution losses from 35.4% to 17.7%, and the Umeme concession (2005) cut losses from 38% to about 15% and invested over $800 million — yet government took distribution back on 1 April 2025. erranet.org
Overrides cluster around elections and price shocks: Pakistan's cabinet cut tariffs 5% on 11 December 2002 despite NEPRA's legal authority; Ghana's PURC was overridden in the 2012 election year and an 18% cut was announced in 2017; Tanzania revoked EWURA's approved 8.5% rise in January 2017. ppp.worldbank.org
Design lessons the cases support
Automatic formulas beat discretionary reviews — until suspension is costlessKenya (2005), Brazil (2013) and India (2021) show mechanical pass-through working; the Dominican Republic adopted an automatic path in November 2021 and suspended it on 19 July 2022 under fuel-price pressure.
Regulators need protected tenure, own budgets and published decisionsWhere the executive kept appointment and removal power — Ghana's PURC, Nigeria's NERC (overridden 2015), Tanzania's EWURA (2017), the Dominican SIE (2003) — legal independence did not bind.
Privatization works when the contract is specificChile (1982), Peru (1993), Colombia (1994) and Delhi (2002) wrote baseline losses, investment duties and tariff rules into enforceable terms; Delhi's AT&C losses fell from above 55% to below 10% by 2021, though later tariff populism strained finances.
Manager contracts are the most fragile instrumentThey delivered gains in Tanzania (2002–06), Kenya (2006) and Liberia (2021–24) but rarely outlasted their term: Rwanda terminated at two of five years (March 2006), Nigeria's Manitoba contract lapsed in 2016, Liberia's hires were not renewed in late 2024 with no merit-based succession rule.
All 31 cases
Family
Country
Safeguard
Start
Override / end
Outcome
Verdict
Source
Method: 31 dated institutional cases on political control of state electricity utilities in developing countries, compiled from 25 independent source hosts (no single URL backs more than 4 rows). One row per named safeguard: design family, country, start and override/end dates as stated in sources (year-only where sources give no day), and the sourced outcome. Verdicts classify each case as worked, mixed or failed per the source's own assessment. Counts summarize this purposive comparative set and are not causal estimates; outcome text is trimmed in the table for space, with the full source linked on each row.