Prepaid electricity meters are changing how residential customers pay for power across the developing world
Electricity demand is expected to grow dramatically in the coming decades, with much of the growth coming from the developing world (Wolfram et al. 2012). However, on the ground, maintaining a reliable supply of affordable electricity presents challenges for both service providers and households. The standard model for recovering the costs of electricity provision – like the provision of many other public and private services – is a post-paid billing system in which a customer is charged for consumption over the past month. For poor customers, finding resources to pay the bill presents a major challenge. At the same time, enforcing the payment of bills carries costs for the electric utility, and bills may go uncollected. Kojima and Trimble (2016) estimate that the total annual value of uncollected electricity bills is 0.17% of national GDP on average in sub-Saharan Africa.
New technological solutions
Prepaid electricity meters offer a technological solution to the bill-payment problem. Similar to prepaid mobile phones, they operate on a debit basis. Customers purchase electricity and load it onto their meter, and as long as they have a positive balance, current flows through it and into the home. Importantly, prepaid meters transfer the enforcement burden from the utility company to the customer: if customers cannot afford electricity, they do not consume it. As a result, prepaid electricity meters have grown in popularity as utilities in developing countries expand to serve poorer and less credit-worthy customers. Indeed, market forecasts indicate that prepaid meters will make up the majority of the electricity meter market in Africa by 2020 (Northeast Group 2016).1
In spite of growing interest in prepaid metering to improve service delivery in developing countries, little is known about how alternative payment structures influence consumption habits at the household level or the financial viability of serving low-income customers.
Measuring the effects of prepaid metering
South Africa was an early adopter of prepaid electricity metering during the phase of rapid electrification following the end of apartheid, when electricity was rolled out to poor and rural communities. In partnership with the municipal electric utility in Cape Town, South Africa, we have begun to build an evidence base around prepaid electricity metering that can help inform policy decisions in other settings, where the technology is less mature.
We implement a field experiment in collaboration with the municipal electric utility in Cape Town in which over 4,000 residential customers were switched from post-paid billing to prepaid metering (Jack and Smith 2019). The order of the meter replacement programme was randomised across neighbourhoods, and replacements were involuntary, which allows us to recover a clean estimate of the impact of prepaid metering relative to post-paid metering on a variety of outcomes.
We observe monthly electricity use and payment behaviour for up to a year before and after the meter replacement, and combine it with administrative data on the costs of operation from the electric utility. This allows us to ask the following questions:
- How prepaid metering affects electricity use?
- What are the implications for the electric utility’s bottom line, which depends not only on the change in consumer demand, but also by differences in the timing, completeness, and cost of revenue recovery?
Consequences for customers and the utility: Reduced electricity use and increased revenue
Figure 1 shows electricity use for the around 4,000 customers in the study, as they are switched from post-paid billing to prepaid metering in late 2014 and early 2015. On average, customers reduce their electricity use by 1.9 kWh per customer per day, or around 14%.2 The reduction is stable for up to a year following meter replacement. The largest reductions come from high consumers and from poorer customers. Consistent with past work in this setting (Jack and Smith 2015), we find that customers quickly adopt a pattern of high frequency purchases. On average, customers in the study purchase electricity every three days on a prepaid meter, while they previously spent money on their electricity bill roughly once a month. This change in spending patterns results in additional feedback on the price of electricity, which we show can explain at least some of the reduction in electricity use following the switch to prepaid metering.
Figure 1 Electricity use and the share of customers on prepaid electricity metering
The estimated decrease in consumption results in a corresponding decrease in the revenue due to the utility of around $6.90 per customer per month, and saves the utility around $3.10 in average monthly kWh supply costs. However, more complete revenue recovery, lower recovery costs, and payments that arrive around three months earlier, on average, tip the balance in favour of prepaid electricity meters from the utility’s perspective. Even in a setting where the overall likelihood of non-payment is low relative to other parts of Africa, the value of complete revenue recovery is a substantial component of the gains to the utility from prepaid metering. In our setting, the additional net revenue from a prepaid meter makes up for the higher fixed cost of the meter in less than a year. Accounting for changes in electricity use is important: assuming customer behaviour remains fixed increases the projected returns from prepaid metering (relative to post-paid billing) by almost 50%.
Policy implications
Together, our results indicate that prepaid electricity metering can help overcome revenue recovery challenges, particularly for the types of customers that are least profitable to the electric utility on a monthly billing model (i.e. poor customers who consume little power and are often delinquent on their bills) and therefore generate the highest returns from the switch to prepaid metering (see Figure 2). While these customers benefit from lenient enforcement of bill-payment, they also generate an externality on other customers by undermining the revenue base necessary for infrastructure expansion and maintenance. Prepaid metering makes substantial progress in narrowing the gap in net revenue to the utility across different types of customers in our sample, with implications for expanding electricity access in other settings. While our point estimates may not generalise, the heterogeneity that we observe suggests that prepaid metering will be relatively more beneficial to the electric utility in settings with a more delinquent customer base or a smaller margin per kWh than Cape Town.
Figure 2 Electric utility’s returns from prepaid metering relative to post-paid billing
Our data do not allow for a clear accounting of the mechanisms underlying the consumption response to prepaid metering. Fully understanding how prepaid metering affects customer well-being requires knowing why electricity use changes so much. The impacts we document are substantial, and the welfare implications depend on whether they reflect conscious conservation enabled by greater control and improved information or salience, or whether they reflect constraints that leave households sitting in the dark at least some of the time. We leave a complete investigation of these potential channels underlying our results for future research, but note that in many settings, the choice between prepaid and post-paid metering is likely to affect the extensive margin of access. In these settings, the appropriate welfare comparison may be electrification with prepaid metering or no electricity access at all.
Editors' note: This column is part of our series on electrification.
References
Jack, B and G Smith (2015), “Pay as you go: Prepaid metering and electricity expenditures in South Africa,” American Economic Review, 105(5): 237–241.
Jack, B and G Smith (2019), “Charging ahead: Prepaid electricity metering in South Africa”, American Economic Journal: Applied Economics, Accepted.
Kojima, M and C Trimble (2016), “Making power affordable for Africa and viable for its utilities,” Technical Report, World Bank.
Northeast Group (2016), “Global prepaid metering: Market forecast”, Technical Report, North- east Group, LLC.
Wolak, F (2011), “Do residential customers respond to hourly prices? Evidence from a dynamic pricing experiment”, American Economic Review, 101(3): 83–87.
Wolfram, C, O Shelef and P Gertler (2012): “How will energy demand develop in the developing world?”, Journal of Economic Perspectives, 26: 119–138.
Endnotes
[1] The expansion of prepaid metering is not limited to Africa. The Northeast Group (2016) estimates that an additional 126 million prepaid meters will be rolled out between 2016 and 2026, with growth in all parts of the world. In the US, utilities in 34 states offer some type of prepayment plan for electricity.
[2] This is a very large change in consumption, in line with reductions under critical peak pricing, during which prices rise by up to 600% (e.g. Wolak 2011).