The soft bigotry of low expectations

Why are we moving energy poverty goalposts in the wrong direction?

I love ♥️ that the world is committed to connecting everyone to modern energy by 2030. Understandably, a lot of attention will go to the 675 million people who, as of 2021, still lack access to electricity. But the way we’re measuring the goal is a problem – and that problem just got worse.

This post gets into the nerdy weeds of energy data and measurement but has massive implications for how the world tackles poverty and climate change. The goals that are selected and the indicators used to track progress will affect literally billions of dollars of investment. ‘What gets measured gets done’ is a cliché, but I think it’s accurate. And I think we’re at great risk of not quite measuring the right thing, which will do a gross disservice to both those living without electricity and the billions (!) who are counted as electrified but still live in stifling energy poverty. 

I’ve long grumbled about the definition of “modern access” used by the standard setters at the International Energy Agency (IEA) as too low. As this 2020 methodology note explains:

…electricity access includes a household having an electricity supply connection, with a minimum level of consumption of 250 kilowatt-hours (kWh) per year for a rural household and 500 kWh for an urban household….

So the minimum consumption required to be considered electrified is (assuming 5 people per household) just 100 kWh per person per year or 50 kWh in rural areas. This is not very much electricity. Enough for a few lights, a phone charger, maybe occasional use of a radio or a small fan. No fridge, no kettle, no washing machine, no laptop. Nothing anywhere close to a common understanding of the term “modern.” 

Let’s modernize modern energy

That’s why colleagues and I proposed a “Modern Energy Minimum” of 1,000 kWh per person per year as a new global standard. This metric would provide a clearly trackable measure of progress and a goal for governments and donor agencies. This metric would not replace IEA-defined access but provide a 2nd additional target after reaching first basic access. 

The Modern Energy Minimum is obviously more ambitious because it’s more electricity use. But it’s also more closely connected to other development goals like job creation and industrialization because it’s wider. Unlike the IEA metric, it tracks non-residential power too. Most people usually think about electricity in terms of household appliances and the size of a power plant is often explained by how many homes it serves. Yet the majority of the world’s power – over 70% – is used in factories, office buildings, data centers, and other parts of the economy. 

I like to think of it this way: electricity at home makes our lives better, while electricity outside the home allows us to earn a better living. (Okay, even if you work at home, you’re likely using a lot more power at a data center somewhere than in your home laptop.) We need abundant reliable power both at home and in the wider economy, so let’s start measuring both. 

The low bar gets… lower?

The latest IEA Guidebook for Improved Electricity Access Statistics is trying to reflect the more complex reality than a yes/no for electrification. It makes the case that we need to think about energy access as multifaceted, using the 5-step “multi-tier framework” developed at the World Bank. The Guidebook even mentions the Modern Energy Minimum (pp. 14-15) as an alternative metric 🙌, yet dismisses it as unrealistic 😲:

Although this level represents a good long-term target, the proposed MEM does not mirror the reality of the large population gaining access for the first time who will need to start at much lower levels, which is more accurately reflected in the IEA basic bundle or even below. 

True. Of course, true. People living with zero electricity will not suddenly jump to 1,000 kWh per year. The IEA instead sets their ‘basic bundle’ as the more appropriate definition. 

The IEA defines access to electricity as a household receiving enough electricity to power at least a minimum level of services capable of growing over time. The IEA minimum level of services is defined as the “basic bundle”, which includes more than one light point providing task lighting, phone charging and a radio, or broadly equivalent to a range of around 50-75 kWh per household per year, depending on efficiency levels (p. 13). 

So the new IEA standard for energy access is not even the old 100 kWh, but rather 10-15 kWh per person per year. My house will use that much in the time it takes me to write this post. 

What’s going on here? The people at IEA are smart and dedicated, so why would they further lower a ridiculously low bar? 

Strong endorse: chickens on book covers!

We should measure energy poverty more like we measure income poverty

Income poverty is a useful comparison to illustrate the many problems here. In 2015, the world committed to ending extreme poverty via Sustainable Development Goal One. How we’re measuring progress is very clear. SDG1, Target 1: “By 2030, eradicate extreme poverty for all people everywhere, currently measured as people living on less than $1.25 a day.”

I can imagine several objections.

1. $1.25 per day is too low. Yes. People living at $1.26 per day are still unbearably poor. But this is explicitly an extreme poverty line, not an end goal in itself. No person on the planet hits $1.26/d and declares income success. Plus the World Bank adjusts the threshold to reflect changes in purchasing power, so the new measure is now $2.15 per day. 

2. Income growth is a ladder not a binary event. Also true. That’s why the World Bank measures poverty headcount ratios at $2.15 per day, and also at $3.65 and again at $6.85. This is because, of course, a lot of people earning over $2.15 per day are still very poor. We don't want to forget about them after crossing some low arbitrary threshold. 

3. Even these higher income levels are pretty low. Indeed. I would guess nearly all people everywhere earning $6.85/day ($2500/year) are still hoping for much more. The Bank’s research department is proposing a fourth “prosperity gap” measured at $25 per day. I love this idea. (Why not go even higher and measure at $100/day too? Let’s aim to get everyone on the planet to $36,500.)

4. How can we accurately measure such things? A good question, but the combination of surveys and new data collection techniques can give us a decent sense of the income trends pretty much everywhere on earth. Measuring income is very far from perfect, but it’s good enough to help us track big goals like the % of people above or below multiple income lines. 

Energy poverty measurements are worse conceptually and practically

Now, let's look at energy poverty goals. SDG7, Target 1: “By 2030, ensure universal access to affordable, reliable and modern energy services.” That’s fuzzy. So too is the specific indicator 7.1.1: “Proportion of population with access to electricity.” Who decides what counts as “access to electricity”? The IEA. The UN and all the other development agencies working on energy poverty use IEA definitions and data reporting to determine how far the world is on or off track in reaching SDG7. 

The same big concerns from income poverty could apply to energy too:

1. 100 kWh is too low. And 10 kWh is ridiculous. With a few years to go until 2030, the IEA has just cut the bar by another 70-90%. No person on the planet uses 10 kWh or 100 kWh and declares energy success. But, yes, of course, we need an initial goal for people who are starting from 0 kWh. Maybe 10 kWh or 100 kWh is the right first step. We should just apply the same logic as income and relabel this explicitly as the extreme energy poverty line. 

2. Energy use is a ladder not a binary event. The World Bank’s multi-tier framework is useful in thinking about the many steps to a future where energy is no longer a binding constraint on human capabilities. But picking the lowest possible rung and using that as the equivalent of “modern energy access” seems just wrong. In fact, income and electricity are so tightly correlated, we can see them move together in sync. The old 100 kWh/capita correlates with less than $1/day, while the 1,000 kWh Modern Energy Minimum correlates pretty closely with the World Bank’s 3rd poverty line of $6.85/day. [fn: yes, yes, WB uses PPP exchange rates while our analysis uses current $ so these aren’t exactly comparable. But given the state of all this data, we’re in the same reasonable neighborhood.] So we need, like with income, multiple levels to track energy progress.  

Source: Energy for Growth Hub

3. Even these higher levels are not high enough. Some people believe 1,000 kWh is too high a goal, yet it’s actually not very much electricity. In the United States, the average is over 12,000 kWh per person per year. Our lives and our jobs are intimately bound with abundant reliable power (and we’re not going to give that up). That’s why the world should be aiming not just for 100 kWh for everyone, and then 1,000 kWh each, but even higher. Much higher. With cloud computing, AI, electric transportation, and a growing global middle class, we’re going to need at least 5,000 kWh or maybe even 10,000 kWh for each person on the planet. Let’s plan for prosperity, not poverty. 

4. Can we even measure such things? Don't look too closely at any of this energy data. The IEA estimates are mostly based on national reporting which can differ a lot across countries and is really just an educated guess (that’s also sometimes, um, questionable numbers, like India is 100%?). The World Bank’s multi-tier framework, while useful conceptually, requires a detailed (and costly) survey that will only be done sporadically, so scaling that to a 5-level global tracker is simply not feasible. 

Tracking our proposed Modern Energy Minimum also requires some empirical basis for electricity consumption distribution within a country. Utilities tend to keep that data private, so that cannot be done with current public data. Fortunately, technology can help us. My colleague Stephen Lee has created Open Energy Maps using a machine-learning approach he pioneered at MIT to create building-level consumption estimates. This will soon allow us to measure, within a reasonable margin of error, the headcount at different levels of consumption, including the Modern Energy Minimum. Stephen will soon release the data, but here is a taste that just dropped. So, yes, we will soon be able to track the Modern Energy Minimum, which makes it a viable addition to our energy progress tracking toolkit. 

So why would the IEA lower the bar?

Just as the IEA is making the case for multiple tiers and the importance of people moving up the energy ladder, and new data techniques allow better measurement, why would they drop the access definition to a measly 10 kWh?

• Efficiency gains? The technical answer IEA seems to offer is that less electricity is needed to provide the same basket of services. Lighting is the obvious example since LEDs need much less juice than incandescent bulbs. But this alone can’t possibly justify shrinking the kilowatt-hours to such a pathetic level.

• Technology appeasement? The staunchest resistance I’ve received personally from pitching the Modern Energy Minimum has come from the community of entrepreneurs and nonprofits selling small off-grid options like solar lanterns and pico solar home systems. Several openly complained to me that defining the access goal upwards would make their gadgets obsolete. This is obviously self-serving and a backward way to set goals. Yet reducing the minimum to 10 kWh will clearly allow many more of these players to claim their products are helping meet the UN goal, and presumably such claims abet marketing and fundraising. Interestingly, mini-grid developers who can deliver higher levels of electricity consistent with the Modern Energy Minimum have been among the most supportive. Coincidence?

• 2030 crunch time? With the deadline approaching, any changes to lower the bar will by definition allow more countries and their partners to claim more success. I cannot say with any confidence that was a deliberate factor within IEA, but a lesson from USAID’s Power Africa is illuminating. That initiative had an aggressive goal of 60 million new connections and, under intense political pressure to deliver quickly, initially counted solar lantern deliveries as connections. USAID’s Inspector General, in a mostly positive report, dinged the agency for overstating their impact by doing this, a practice that was later halted. Defining the IEA target downward makes it easier for all to hit SDG7. 

IMO, none of these reasons justify changing the definition. We should be tracking progress at multiple levels, not just making it easier to claim success at the very bottom rung. This change appears especially unfortunate at a time when reliable and abundant electricity is more important than ever for people and for competitive economies. In an age when so much of our lives and our economies run on abundant electrons, it seems doubly wrong to set an even lower standard for the world’s poor. Sure, we can set unambitious interim targets but, no, we cannot allow them to be defined as success. That would leave huge swathes of humanity behind. 

Let’s not gaslight the energy poor

One prominent scholar who shall remain unnamed here once told me that, “people living in the Amazon are happy with 100 kWh. They don't need more.” I’ve written before about the awkward topic of racism in climate and development. Calling a few lights “modern energy access” smacks of a similar attitude that parts of the world should live with less. We should all disagree. 

This all especially reminds me of the late great Michael Gerson, a beautiful writer and unshakeable advocate for the world’s poor, who warned against, “the soft bigotry of low expectations.”