The Developing World Is Connecting to the Power Grid, but Reliability Lags
Most discussions about energy in the developing world quickly turn to the 1.3 billion people who don’t have access to electricity. Many initiatives are focused on serving these people, such as Kenya’s Last Mile Connectivity Project or the US’s ambitions of supporting 60 million new electricity connections through the Power Africa program.
I hear a lot less about the people who have an electricity connection but for whom 
reliability is bad. And, as anyone who has spent time in the developing world knows, poor reliability doesn’t just mean that the lights occasionally flicker. Whole neighborhoods can be left in darkness for hours and even days on end.
The importance of reliability – and not new connections – is particularly meaningful in cities, where nearly everyone has a connection, but reliability is poor. For example, the figure below, from a forthcoming working paper by my colleagues Ken Lee, Paul Gertler and Mushfiq Mobarak, summarizes data from 21 Sub-Saharan African cities. The share of city dwellers who have access is represented in the graph on the left – many are close to 100 percent, and almost all are above 75 percent (except the capital of Malawi). The graph on the right represents the share of those city dwellers who report that their electricity connections works “all the time” or “most of the time.” It’s less than 20 percent in Lagos, Nigeria, home to nearly 20 million people.
Cities are important because one of the starkest demographic trends in the developing world is urbanization. And, in my experience, a lot of rural families are tied to the urban economy – relying on remittances from the husband, older son, or uncle who has moved to the city.
The Reliability Measurement Gap
So, why is reliability the poor stepsister to energy connections? Why aren’t there more programs to promote reliability?
I believe a major reason is that reliability is really hard to measure. As my colleague Jay Taneja says in a recent working paper, “Before electricity reliability can be improved, it needs to be accurately measured.” You can’t fix what you can’t see.
Jay’s paper has some super interesting figures highlighting the difficulty measuring reliability in the developing world. For example, the graph below compares a common metric of outages (called SAIDI, for System Average Interruption Duration Index) measured two ways – as reported by utilities on the vertical axis and by their customers on the horizontal axis. Each point represents a country. If the utilities and their customers were reporting the same thing, the points would all lay on the dashed line. In fact, almost all of the points are below the line, indicating that the utilities are reporting one-seventh as many outages as their customers on average!
Luckily, Jay and a team of engineers at Berkeley are working on some really innovative, inexpensive ways to measure reliability using smartphones and low-cost sensors.
The Cost of Poor Reliability
Why might reliability be a big issue?
For one, poor reliability doesn’t just impact households, but also hospitals, factories, telecom systems, government buildings, etc., all of which are important to economic development. Around the world, non-residential customers use well over 50 percent of 
electricity, and over 70 percent in some of the major developing countries, including India, China and Brazil.
I spoke to an entrepreneur from Lagos last year who was trying to make a go of a company selling beauty products designed for the local market. It’s a business that has very little to do with energy. And yet, the conversation quickly turned to Lagos’ electricity problems. Not only did he have a backup generator, but his internet service provider had a backup generator, his accountant had a backup generator – you get the picture. To start a business in Lagos, you have to invest in a generator. This is a tax on doing business, which makes it hard for new businesses to start and for existing ones to grow.
We need more research to document just how much of a drag this is on the local economy, but I suspect it could be a big hindrance to growth.
Within the residential sector, this isn’t necessarily a story about allowing rich city dwellers to watch TV and keep their apartments air-conditioned. In fact, reliability may impact the poor more than the rich. The chart below was made by an enterprising Accra resident, who hired people to stand on street corners in different neighborhoods and record whether the nearby lights were on. (Another indication of just how starved people are for data on reliability.) The neighborhoods at the bottom of his chart, with 3 or fewer outages over the two-week period, are where the rich people live. I believe that the human observers were able to distinguish grid outages from local backup generators, although I’m not positive.
Finally, the pollution created by all of the backup generators is a major contributor to poor air quality. For example, my former graduate student Fiona Burlig pointed me to estimates from India suggesting that diesel generators contribute 10 to 20 percent of cities’ pollution, depending on the pollutant (here’s another source).
Considering the Trade-Offs Between Reliability and New Connections
Don’t get me wrong. I care about the 1.3 billion people who do not have electricity connections. They are no doubt some of the world’s poorest people. But, that’s only part of the energy access problem. And, I believe that we need to be open to the possibility that connecting fewer people and increasing reliability for existing customers is better for economic development than putting all our eggs in the connection basket. Remember, the already-connected customers include hospitals, schools, etc.
Kenya, for example, has made great strides connecting households to the grid in recent years. According to the latest reports, 15 million more people are connected to the grid 
now than 7 years ago. But, these newly connected consumers aren’t using much electricity – only one fifth of what the average household was using in 2009. This means that Kenya Power’s revenues per customer are dropping, just as it has added a bunch of new infrastructure to its system, infrastructure that it will need to maintain for years to come. It’s possible that building out the system will take resources away from improving the reliability.
There’s a lot of work to do. We need to figure out cost-effective ways to improve measurement technologies, identify the many varied causes of poor reliability, work with utilities to improve their systems for both preventative maintenance and triaging when they face reliability incidents. Those seem like jobs for engineers. Economists can provide estimates of the development benefits to investments in reliability as well as energy access, and they can identify ways to provide utilities with better incentives and more capital to invest in reliability. There are big payoffs to getting these answers right.
Categories
Catherine Wolfram View All
Catherine Wolfram is the William F. Pounds Professor of Energy Economics at the MIT Sloan School of Management. She previously served as the Cora Jane Flood Professor of Business Administration at the Haas School of Business at UC Berkeley. From March 2021 to October 2022, she served as the Deputy Assistant Secretary for Climate and Energy Economics at the U.S. Treasury, while on leave from UC Berkeley. Before leaving for government service, she was the Program Director of the National Bureau of Economic Research’s Environment and Energy Economics Program, Faculty Affiliate of the Energy Institute at Haas from 2000 to 2023, as well as Faculty Director of the Energy Institute from 2009 to 2018. Before joining the faculty at UC Berkeley, she was an Assistant Professor of Economics at Harvard. Wolfram has published extensively on the economics of energy markets. Her work has analyzed rural electrification programs in the developing world, energy efficiency programs in the US, the effects of environmental regulation on energy markets and the impact of privatization and restructuring in the US and UK. She is currently working on several projects at the intersection of climate and trade. She received a PhD in Economics from MIT in 1996 and an AB from Harvard in 1989.
Energy Institute Blog 
New World Bank piece. http://blogs.worldbank.org/developmenttalk/rural-electrification-how-much-does-sub-saharan-africa-need-grid . Key quote (one of several), and likely a dis-incentive to electrification.
“The economic history of today’s advanced industrialized world suggests that the findings of only subtle effects of electrification on economic development are not as surprising as some might think. The diffusion and economic impacts of technological innovation might take years or decades to fully materialize, depending on other interacting processes. As David (1990) notes, in 1900, some 20 years after the first generation stations were installed in London and New York, electricity was still hardly seen in homes or factories. Moreover, even when factories in today’s industrialized countries adopted electric machinery, their productivity did not increase for another 10 or 20 years – a phenomenon that is known as the productivity paradox.”
I wonder what metrics that report is using. I wrote here about how DER could follow the innovation pattern that exploded at the beginning of the twentieth century (https://mcubedecon.com/2014/07/29/what-we-might-expect-for-diffusion-of-new-decentralized-energy-technologies/). Also, tech diffusion and adoption appears to be accelerating, maybe in part due to the increasing economic connectiveness.
I wonder how much of
happened because electrification was centralized? Certainly if other technologies are examined, whether cell phones or color TVs, where the choice of adoption is not controlled by a central authority, the rate is remarkably fast:
I do not have references, but I have also heard that cell phone take-up in developing nations is much faster than, say, fixed lines, and this combines cell availability via decentralized towers and acquisition by individuals.
Two points.
First, in some instances, such as Iraq, grid instability is because the grid offers an easy and rich target for insurgencies to disrupt. The IEEE Spectrum magazine offered a recounting of the experiences of one electrical engineer who struggled, with U.S. backing, to restore portions of the grid to operation after the U.S. invasion. It is a hard slog.
Second, any solution which involves centralizing control over energy distribution or management is inherently a corruptible one. While in developed, democratic nations this is limited to regulatory capture, in other nations such controllers succumb to bribes and worse. The only way to stop this is to figure out a means of decentralizing energy. Decentralized energy is, on average, more reliable, because it cannot be disrupted by forces, whether terrorist or central governments which effectively demand tribute for reliable service.
Even in the United States, I’d go as far as to say that if you want to take back control of your democracy, you first need to take back control of your energy supply.
I agree that many people miss how utilities aggregate political power that serves to protect their investments through centralized power networks. The utilities oppose distributed energy resources and community choice for more than just technical or even economic reasons. They see that their political power will also be diluted.
Reliability lags, popularly called “Dumsor” (put off, put on) in Ghana can be traced to consumer inability to pay. Thermal plant operators are not able to pay for reliable gas supply from Nigeria through the West Africa Gas Pipeline. The country’s main power distributor; the Electricity Company of Ghana is on the verge of being divested mainly owing to its inability to collect revenues owed it. Even prepaid meters are made fool proof by illegal connections.
Owing to the high content of foreign exchange in the cost of power production, current tariff levels are lower than the cost-based tariffs they ought to be, to enable reliable service delivery. I suspect that no amount of nurturing of consumer ability to pay, price discrimination, regulatory reforms or other can match genuine consumer ability to pay to the high costs of power production. The suspicion is mine only and needs to be investigated.
In the other part of the world, particularly California, massive investments into renewable energy has led to a situation where investors may have to pay to get their products consumed. Investment into behavior controlling technology such as thermostat’s, motion and light sensors, also made rebound effects minimal or non-existent in for example LED lights consumption.
The situation in the developing countries call for a strategy that improves cost of production, security and reliability of supply and yet lowers tariffs for consumers. Technologies that can hold costs low irrespective of volume, which can also take many controls from humans, are important to solve the erratic power problems of developing countries like Ghana. In such circumstances, the role of developing fund assistance such as from DFID, USAID, JICA, DANIDA etc., is highly important.
If investors who may need to pay in one way or the other for their products to be consumed would invest in utility scale solar or wind farms in suitable locations in developing Africa particularly Ghana and receive reduced returns on investments and hearty thank you, it will improve the countries energy security, reliability, efficiency and economic growth as well as help reduce reliance of exhaustible, environmentally unfriendly fuels use for power generation. Of course there are alternative investments in the developed world but it takes a big heart to venture into rescue operations. Prof. Wolfram, with a big heart for developing countries’ energy problems, I perceive is calling for a move in this direction.
Gilbert, you raise many points that are directly relevant to my discussion above in relationship to weak institutional frameworks in the developing countries. Resource rich based economies such as Ghana, Angola, Mozambique, Nigeria, Tanzania, and Zambia are at great risk given their weak institutional frameworks and the volatility of commodity markets. These countries must embark on institutional reform if they are to see improvement in the performance of the energy and power sectors, the services provided and cost of service. It can be done. Look at Chile and best of luck.
Without a question, reliability of electricity services in much of the developing world represents a key challenge to the health of the economy due to frequency and duration of power outages. Reliability and connectivity concerns in the developing world are in the main due to the weakness of their institutional frameworks. Weak institutional frameworks in regard to energy and power sectors contribute significantly to: market failure and allocative inefficiency, regulatory capture and potential for collusion, high cost of service, potential for stranded assets, high energy intensity in power generation, increased power losses, reduced effective private sector investment and participation, interference with sector reform, inequitable distribution of electricity services, and economically unviable electric utilities. It is these features of weak institutional frameworks that reduces the chance for adequate reliability services standards, sends the signal to entrepreneurs to invest in and market highly polluting high heat rate diesel generated power, and retains electricity supply and distribution as a public policy instrument with no effective oversight by consumers, utilities, or regulatory agencies. Such weak frameworks are unable to optimize grid and system reinforcement versus connectivity. Such weak frameworks are unable to provide the basis for utilities whether private or quasi-government owned to break even. While the key elements required for the improvement of weak frameworks are basically known such as improved regulatory capacity, fiscal efficiency, accountability, and commitment, framework reform in the developing countries is formidable given the need to also reform policy, market, as well as institutional frameworks. This returns us to the need to address frequency and duration of outages in the absence of adequate institutional frameworks. The valid suggestion made in this blog is to make available mechanisms to monitor frequency and duration of outages as a mean to encourage debate on the quality and reliability of service in any one locality, to empower interest groups, and pressure authorities to proceed with institutional reform.
I generally agree with Chris Marnay’s observation. This is even a problem in the U.S.
An important distinction is that the SAIDI measures transmission level outages and the CAIDI ( Customer Average Interruption Duration Index) measures distribution level outages. In California, I’ve calculated a CAIDI that is 16 times higher for PG&E than the SAIDI. The distribution network is often the real weak link. Moving toward the solutions that Chris outlines are likely more cost effective.
And until recently, moving to diesel in India actually LOWERED GHG emissions because its coal-fired fleet is so dirty. It’s probably cheaper to set up distributed solar to displace that diesel.
Here’s a story on India shifting from coal to solar: https://qz.com/992611/from-coal-to-solar-indias-energy-landscape-is-almost-too-hard-to-keep-up-with/
Reliability is a normal “good,” i.e. more of it is always better than less. Further, measuring reliability is clearly useful. But, these features don’t add up to a compelling case that the only way to raise social welfare is raise reliability. The optimum level of reliability is one where the cost of providing it plus the cost of not having it is minimized. Unfortunately, the cost of providing it can be high, especially where investment capital is limited, and cost rises asymptotically as the grid nears perfection. The minimum social cost reliability level varies across regions and customers; further, it varies enormously across end-uses, from critical life support systems to schedulable pumping. Critical service interruption provides the bulk of outage costs, and raising the overall level of grid reliability above the point of minimum total social cost is a blunt instrument for ensuring critical services are provided. It’s more promising to serve critical loads in more locally targeted ways, such as microgrids, back-up generation, and UPS. In countries with poor reliability, this is what tends to happen. Electricity delivers wildly diverse levels of value added across energy services, so a more creative approach would be to identify the interruption of which services causes how much loss of social welfare, then devise systems the minimize total social cost. It must be recognized that not all unserved grid load is equal, and there are both alternatives to centralized power generation and delivery and methods of prioritizing and scheduling loads.
HI Catherine, interesting post as usual: In particular the idea that reliability might be more important for economic development than new connections (though the new connections argument has a moral element to it – all good development should have both moral and economic drivers).
One point however: while the 1.3 billion unconnected households gets a lot of play, it seems slightly misleading to suggest that there is no focus on reliability. For example if you look at the World Bank Group evaluations, they constantly show how the majority of financing goes toward investments in countries with relatively high access rates, and are directed towards improving the quality, reliability and efficiency of supply rather than providing access. This is something that the World Bank actually laments – e.g: http://ieg.worldbankgroup.org/evaluations/world-bank-group-support-electricity-access
I think it is important to keep our focus on reliability, and i think we should know about development gains (economic vs moral) that might result. That said, i think all the chatter about the 1.3 billion is an effort to address the fact that historic (and current) finance has been skewed towards established customers precisely because they represent a market (which is necessary for the utilities). Portraying the current chatter (which is an effort to address this) as if its a problem which marginalizes what should be a focus on reliability, seems out of touch with what has actually been happening.
The underlying problem in most of the countries, including South Asia [India Pakistan Sri Lanka Bangladesh] is the culture of politician and bureaucrat corruption. Piwer ‘must’ be maintained in area where the powerful live and work, even if industry gets cutoff. With the surging summer heat the influentials demand more to keep their A/Cs going. How is that to be managed in a system with limited production and grid capacity? Rolling [scheduled or unscheduled] outages.