New analysis from the UK government shows that households are heating their houses less than is required to meet the levels thought necessary to deliver comfort and health. Those on lower incomes are “under-consuming” by a larger margin than those on higher incomes, with only the top richest decile consuming more than the estimated requirement. It seems probable that increased prices for electricity are rationing the poor out of the heat market.
Electricity demand in the United Kingdom has been falling for about fifteen years, with consumption in 2017 at levels last seen in the 1980s.
Figure 1: Electricity Supplied in the United Kingdom, 1921 to 2017 (GWh). Source: Charted by the author from data available from the Department of Business, Energy and Industrial Strategy (BEIS).
The fall is so large and so closely correlated with the introduction of policies increasing electricity prices – note that demand falters in the middle 2000s shortly after the UK introduced its Renewables Obligation subsidies in 2002 – that there is a lurking suspicion that price rationing must be at least an element in any plausible explanation, certainly in more recent years. Even if we allow that the early onset of the 2008 crisis is probably responsible for the initial decline in electricity consumption, the lack of a subsequent recovery in demand might well be largely attributable to the rising cost of renewable electricity subsidies (about £9 billion a year at present) and their associated system balancing and grid costs (Balancing Services Use of System costs, are now £1.3 billion a year as compared to about £300 million a year in the early 2000s).
In regard to the non-domestic sector there is little real disagreement that this decline in electricity consumption can be confidently attributed in large part to deindustrialisation caused by the export of many production processes to jurisdictions with lower costs, electricity amongst them, principally China. This is an EU-wide effect, with industrial electricity prices in the EU28 being 50% higher than those in the G20, and the UK is part of this broader problem.
But the domestic correlate of this effect, in relation to household consumption, is more controversial. Apologists for the UK’s policies tend to argue that the widespread adoption of efficient conversion devices, such as LEDs and better white goods in areas where demand is not particularly elastic, have cut electricity demand without reducing consumer benefit. But analysis bearing directly on this question is in short supply. Fortunately, as part of its monitoring of fuel poverty, the UK government has recently undertaken an examination of energy consumption at the household level that throws some welcome light on the question.
The March issue of the statistical bulletin of Department of Business Energy and Industrial Strategy (BEIS), Energy Trends contains an article,“Comparison of theoretical energy consumption with actual usage”, which puts actual household energy consumption alongside that predicted by the fuel poverty models as necessary to achieve comfortable and healthy levels of heat.
The government’s analysis finds that 69% of households had a theoretical level of consumption that exceeded their actual consumption, the average underspend being about £133 per household per year, or 9.9% of the expected spending. This tendency is stronger in relation to those households classed as “Fuel Poor”, which were underspending by £319 per household per year (19.9%), as opposed to the “Not–Fuel Poor”, which were underspending by £110 (8.6%) per household per year.
Suspicions that this might result from a known bias in the model, which may overstate requirements (see p. 70), are to a degree dispelled by the fact that the distribution over income bands is uneven, with those on low incomes much more affected. Indeed, the underspend decreases as household income rises, and those households that spent more on energy than the model predicted also had incomes 21% higher on average than the rest of the sample. The following chart from the study illustrates the distribution.
Figure 2: Estimated UK household underspend on energy by income decile. Source: BEIS, “Comparison of Theoretical Energy Consumption with Actual Usage”, Energy Trends (March 2019), 72.
It is quite clear, as BEIS itself concludes, that the effect of underspending is “strongly linked to income”. Low income households underspend on energy to a greater degree than higher income households.
Furthermore, fine-grained analysis revealed that “households with children had the largest average under-consumption” and that, generally, “lower income households with dependants are potentially more likely to under-consume than other households” (p. 74), with this effect particularly marked for fuel poor households (p. 67).
Some would argue that these effects are consistent with the view, which I emphasise is not expressed in BEIS’s paper, that such under-spending, particularly in the lower deciles, is largely the effect of price-rationing. In other words, energy prices are sufficiently high to force consumers to trade off their wish for heat against competing demands for that income, with evidence of the trade-off being, inevitably, particularly marked in the lower income deciles. However, given the known bias in the model itself, the under-spend recorded for the middle and higher income deciles is perhaps less significant and our attention should be focused on the causes underlying low energy consumption by the poor. Why are they underspending by so much?
One possibility is that low-income households tend to use electric heating. The regulator, Ofgem, reports that of the 26 million households in Great Britain, about 22 million use natural gas for heating, with 2.2 million of the remainder using electricity. Ofgem also indicates that electrically heated households tend to be of lower income, with around one third of electrically heated households in receipt of an annual income of under £14,500 a year (See Ofgem, Insights paper on householdsother non-gas heating (2015)). That is not surprising, since many electrically heated households are flats, and 25 per cent of all flats in Great Britain are electrically heated, as compared to only 4 per cent of houses. The rented sector, of course, is used heavily by those on lower incomes. Indeed, there is some reason for thinking that the proportion of flats using electricity for heating may actually be rising, as non-condensing gas boilers on shared flues reach the end of their lives, and cannot be replaced with the now mandatory higher-efficiency condensing boilers since the flue gases are too cool to exit the flue safely. In such cases relatively expensive electric heating is the only option.
It follows, therefore, that the 2.2 million electrically heated households are very probably concentrated in the lower income deciles, precisely where the Department’s study has found greater levels of underspending on energy.
That should be of concern to government since energy and climate policies have a much greater effect on the price of electricity than on the price of natural gas. In 2014, BEIS’s predecessor, the Department of Energy and Climate Change (DECC), estimated that by 2020, policies would be making electricity prices to households about 36% higher than they would be in the absence of policies, while the effect on gas prices was to increase them by only 6% (See the Central Scenario in DECC, Estimated Impacts(2014).)
We can consequently conclude that it is very likely that the significant levels of under-spending on energy reported by the Department for lower income households are caused in significant part by electricity price increases resulting from energy and climate policy.
With that in mind, I wrote to the Department last week asking whether their data-set could identify the electrically heated households and so evaluate the hypothesis that the reported “under-consumption”, particularly of the lower income deciles, was correlated with and in part caused by their use of electric heating. The statistician responsible informed me (email 18.04.19) that the question had indeed been examined, but, unfortunately, the data-set did not distinguish clearly between non-gas households that used electricity only and those that used other fuel types for heating as well. Consequently, the Department “did not have a reliable enough sample to accurately test the difference in theoretical and actual consumption between gas heated and electric heated dwellings.” I see no reason to doubt this explanation, but it is obviously a limitation in the government data that should be rectified promptly.
Even without that information we can be confident that higher electricity prices, known with certainty to result from energy and climate policies, are very probably making heating unaffordable for those on lower incomes, and that Government appears to be price rationing the poor out of the heat market in order to reduce emissions.But so far from having second thoughts, the administration is, as the Chancellor told us in his Spring Statement, planning to extend price rationing to still more households, with its Future Homes Standard “mandating the end of fossil-fuel heating systems in all new houses from 2025”. This could well, as Mr Hammond said in his speech, and apparently without irony, deliver “lower carbon” and “lower fuel bills too”, but only through price-coerced under-consumption.
Dr John Constable: GWPF Energy Editor