Exploring the affordability and value of solar home solutions.
If you’ve ever handled kerosene or set it alight, you’ll know it is disgusting. If you use it the way most of Kenya’s low income families use it–in a tiny tin with a small wick—for lighting your home, you’ll inhale a lot of smoke, the entire place will reek of gas and smoke, your children will strain to read under its dismal light, and if light is needed in another room, you’ll need to precariously tote around the tiny lamp, hoping not to spill kerosene all over the floor.
We repeat, kerosene is awful. But at around KSh 60 – 75 (USD 60 – 75 cents) per liter , illuminating kerosene (or paraffin) is also really, really cheap, which explains why three quarters of Kenya’s poorest households use it as a primary energy source for lighting their homes.
It’s the disgusting part—not the savings—making many Kenyans increase their energy expenditures to acquire solar home solutions. We have seen some claims that solar lighting solutions like M-KOPA and BBOXX make sense to users because they save money. For example, BBOX estimates that its clients save $2 per month on average. Stephan Faris writing for Bloomberg referenced a widely-quoted study (also referenced on M-KOPA’s website):
“According to a 2014 survey, an average off-grid household in Kenya spends about 75¢ a day on energy, or $272 a year—$164 on kerosene, $36 on charging their mobile phone, and $72 on batteries. M-Kopa estimates a customer saves about $750 over the first four years by switching to its basic solar kit.”
But our work on the Kenya Financial Diaries made us sense that these estimates of budget savings were off for low-income households. Very few of our off grid respondents ever paid to charge their phones or buy batteries. In fact, when we looked at Diaries households’ total expenditures on all energy—including lighting, phone charging, and cooking (energy that is not replaced by a solar unit)—we found the median household was spending only KSh 262 per month (or about 4.2 percent of median total consumption expenditure of KSh 6,223 per month), equating to about just $32 per year.
Since the Diaries sample is small and focused on low income people, we also looked at available national datasets to get a more representative estimate of Kenyans’ energy spending and to model just how long it would take for different income groups to achieve net savings by only shifting their existing expenditures. For expenditure data, we referenced the nationally representative 2005 Kenya Integrated Household Budget Survey (KIHBS) and the 2008 and 2015 panels of a large survey led by Tavneet Suri and Billy Jack to assess the impacts of M-PESA. Though some of this data is quite old, kerosene prices remain in a similar price range, though they fluctuate regularly with the price of oil and actually declined in 2011 after a reduction in kerosene taxes. To assess the financial attractiveness of solar lighting solutions, we compared estimates of kerosene spending to the pricing models of two leading off-grid solar solutions:
- M-KOPA’s basic home kit, which requires a deposit of KSh 3500 and payments of KSh 50 per day for one year to own the system;
- BBOXX’s “LIGHT” product, which requires a deposit of KSh 590 upfront and monthly payments of KSh 590 for the first three years (after which accessories, like lights are paid for) and Ksh 440 for the next seven years which covers maintenance and payments towards full ownership after year 10.
The first plot in Figure 2 below shows survey-based estimates of monthly kerosene spending for rural off-grid households at each national consumption decile, from those living at consumption per capita levels among the poorest 10 percent to those living at consumption levels enjoyed by the richest 10 percent. We draw primarily on the 2008 representative M-PESA panel survey for this data, since it is our most recent source. Though still a bit outdated, as noted above, kerosene prices today are quite similar. The second plot shows the number of months to net savings for all levels of kerosene spending in the range of KSh 50 and 1000 per month for the M-KOPA and BBOXX systems, with vertical lines referencing the mean kerosene spending levels for rural, off grid households in deciles 1, 5 and 10.
According to most recent estimate from 2008, the mean rural off-grid household spends around KSh 370 per month on kerosene. Presuming that household spends around KSh 100 on phone charging per month and purchased a solar home system to replace both kerosene and phone charging costs, they would only achieve net savings on M-KOPA in 48 months and on BBOX they would not generate any savings as its monthly financing cost of KSh 590 exceeds their kerosene spending. Rural off-grid households that classify in the poorest 30% of all Kenyans, spend around KSh 295 per month on kerosene and at this spending level, investing in one of these solutions is an even greater stretch. With M-KOPA, they would only be able to achieve net savings in between 45 to 76 months (depending on how much they currently spend on phone charging).
Another way to compare the financial benefits of the M-KOPA and BBOXX systems is by computing net present value (NPV). Figure 3 below, shows the NPV of a solar home system investment after 5 years of ownership at 5 different levels of monthly kerosene spend (and assuming the household spends KSh 100 per month on mobile phone charging). The NPV can be interpreted as the cumulative savings that are generated over the 5 year period time discounted by the central bank rate of 10.5% (cash flows farther in the future are discounted at a higher rate than cash flows nearer to the time of purchase).
Based on these models, it’s unsurprising that these kinds of solutions are not yet ubiquitous among the poorest. Recent estimates suggest that only 5.7 percent of the poorest 40% of all Kenyans are using a solar solution as their primary source for lighting. But, even in the middle of the income distribution, many are willing to pay more for solar than they spend today on dirty and inconvenient sources of energy. Solar units provide cleaner, better quality light. They are more convenient, reducing trips to shopping centres to buy kerosene. They allow users to potentially listen to the radio with more regularity and charge their phones with more convenience, even if there is no cost saving for many years. More convenient charging can even entice some to upgrade their phones to smartphones knowing they have a reliable way to charge every night. All of this value means that even some very low income Kenyans have found a way to pay for solar solutions (See our blog, Solar Stories to understand their experiences).
This analysis highlights some key takeaways that go beyond solar alone:
- When talking about delivering exciting technology in low income markets, we need to be realistic about affordability and continue to seek lower cost solutions to make them accessible for the poor.
- For solutions that deliver new and better value, what people spend today tells us a bit about their budget capacity, but it is not a reliable indicator of willingness to pay.
- For solutions that offer a new and better service, cost savings is not the most important metric of value.
This post contains a correction from the version posted on 20July 2016. Here we update the current price of kerosene to reflect the range of charges around the country.
 Kerosene prices from Energy regulatory committee http://www.erc.go.ke/index.php?option=com_content&view=article&id=162&Itemid=666. Kerosene prices reached a high of about KSh 95 in 2011. Population kerosene use for lighting obtained from 2016 FinAccess household survey. The estimate presented here corresponds to the FinAccess 2016 estimated proportion of Kenya’s adult population (ranking in the bottom 40% of an asset-based wealth index) that uses kerosene as the primary energy source for lighting their home.
 This is based on data from the 2008 M-PESA panel rather than the 2005 KIHBS, allowing us to get a more recent data point. But, since the M-PESA panel combines kerosene, charcoal and firewood, we use the KIHBS ratio of expenditure on kerosene to these other estimates to arrive at this figure. This is a higher end estimate, given that only Kenya locations with at least 1 M-PESA agent were included in the M-PESA panel sample and so the resulting sample was more urban than the national distribution.
 None of the publicly available national household surveys that we reviewed separately tracked spending on phone battery charging. Very few Diaries households paid for charging, usually charging with friends or at a workplace. Those who did pay occasionally paid KSh 5 per charge. Few without electricity have smartphones requiring frequent charging.