Official statistics disagree about what Northern Ireland households pay for electricity. The Office for National Statistics reports the highest costs of any UK region; the Department for Business, Energy and Industrial Strategy and the Utility Regulator report the lowest. In this paper, MCC Economics & Finance finds the ONS figure the likely outlier, a gap worth around £150 per household per year.
Because these are the numbers that decisions rest on. The Office for National Statistics (ONS) figures feed household spending statistics and the debates on affordability and fuel poverty that draw on them. The Utility Regulator (UR) oversees a regulated electricity market in Northern Ireland and needs reliable cost and consumption data to do it. Policymakers deciding whether Northern Ireland households need support cannot act sensibly while official sources say those households face both the highest and the lowest electricity costs in the UK. And consumers hearing both claims have no way to judge whether they are getting a fair deal. Across Northern Ireland's roughly 790,000 households, a discrepancy of £150 per household per year amounts to over £100 million of apparent annual spending that may not exist.
We present a Northern Ireland electricity cost puzzle. Only four explanations are possible: our own interpretation of the data is mistaken; the Department for Business, Energy and Industrial Strategy (BEIS) understate electricity unit costs; the Utility Regulator understates electricity consumption; or ONS overstate weekly electricity costs. This paper tests each in turn. In our view, the household cost reported by ONS is unlikely to be as accurate as the other estimates from BEIS or the Utility Regulator. We are unable, however, to confirm the source of bias affecting the ONS data.
Yes, directly. According to ONS, electricity costs in Northern Ireland are higher than in any other UK region: £12.60 per household per week over the three years to March 2018, against a UK average of £11.30 (Figure 1). According to BEIS, electricity costs in Northern Ireland are in fact lower than in any other UK region, on both the annual bill and the average unit cost (Figure 2).
Figure 1: Weekly spend on electricity per household across UK Regions, 3-year period to March 2018. ONS weekly electricity spend per household for each UK region, with Northern Ireland the highest of all at £12.60 against a UK average of £11.30.
Two features of the chart matter. Northern Ireland does not merely top the table, it tops it by a visible margin over Scotland in second place. And the ranking is not noise from a single odd year, since each bar is a three-year average, which is exactly why the contradiction with Figure 2 demands explanation rather than dismissal.
The table adds one reading the chart compresses: Northern Ireland's £12.60 sits £1.50 above the England average and £2.50 above the cheapest region, the North East, a spread of around 25% between the extremes of a single country's regions on the ONS measure.
Figure 2: Electricity charges across UK regions, 2018 calendar year, average supplier, including VAT. BEIS annual bills (assuming 3,800 kWh) and average unit costs by region for 2018; Northern Ireland shows the lowest of both measures of any region.
Northern Ireland is bottom twice in this chart: the lowest annual bill and the lowest average unit cost of any region. That double reading matters because it removes an easy reconciliation. If only the bill were low, cheap units with high consumption could square it with Figure 1; with both measures lowest, the ONS and BEIS pictures cannot both be right without Northern Ireland households consuming implausibly more than anyone measures them consuming.
The Utility Regulator's Quarterly Transparency Reports, since renamed REMM Transparency Reports, use the same method and process as BEIS and hence reach a similar conclusion: that electricity costs in Northern Ireland are lower than in other regions of Great Britain. In this paper we explore reasons for this apparent contradiction.
No, and that is the first clue. The ONS methodology is based on household surveys: households are asked how much they pay for electricity, using where possible their last bill or payment, and ONS converts the answer into a weekly cost. The BEIS methodology is based on information received directly from supply companies through a quarterly survey, which BEIS describes in its own words:
The outputs differ as well as the methods. ONS captures what households believe their cost is, including the quantity of energy consumed. BEIS reports what suppliers say the charge per unit is, excluding consumption. The Utility Regulator, whose data enters the comparison below, reports consumption itself. The three could still be consistent if consumption per week explains the difference, since weekly cost equals unit cost multiplied by consumption per week. Using the BEIS unit cost and the ONS weekly cost, the consumption that would make both figures true can be derived.
Table 1: Who produces these statistics. The five organisations behind the numbers: what each does, where its data comes from, and what it collects.
The table is the paper's methodological map, and its point is that the five organisations are not measuring the same thing three ways: one asks households what they remember paying, one asks suppliers what they charge, one counts what meters record. The disagreement in the headline statistics is a disagreement between measurement worlds, which is why the paper's approach is to make the three worlds confront each other arithmetically rather than to referee their reputations.
Table 2: Implied consumption per year, using ONS and BEIS data. The consumption that would make the ONS weekly cost and the BEIS unit cost simultaneously true, derived by dividing weekly cost by unit cost (implied weekly consumption) and multiplying by 52 (implied annual consumption), following the report's column derivations.
The derived column is the paper's central device: if the ONS cost and the BEIS price are both right, consumption is the only free variable, and it must land at 4,200 to 5,000 kWh per household per year. That number is testable against what the network operator actually meters, which is what Table 3 does, and the test is what converts a vague sense that the statistics disagree into a quantified, falsifiable puzzle.
Large and persistent. The implied consumption can be compared with the consumption actually reported by the Utility Regulator. In every year the implied figure exceeds the reported figure, by between 15% and 39% (Table 3).
Table 3: Implied consumption minus reported consumption. A puzzling difference.
Clearly, there is a material inconsistency in, or misinterpretation of, the available data on electricity consumption and costs. The explanation must be one or more of the following:
The average difference in Table 3 is 27%. Because we are not sure whether the difference stems from consumption errors or cost errors, or both, we can call this puzzle the missing 1,000 kWh per year, or the missing £150 per year.
Table 4: The puzzle at three scales. The same discrepancy expressed per week, per household per year, and across all Northern Ireland households, derived from the report's figures.
The scaling is the reason the paper exists. A £2 weekly gap sounds ignorable; £100 million a year of household spending that may not exist does not, and it is the £100 million version that fuel poverty budgets, affordability debates and regulatory judgements implicitly rely on.
Table 5: Connection-to-household ratios are small and do not explain the puzzle. Utility Regulator connection counts against ONS household counts for each year; the ratio peaks at 1.08 in 2014, far short of the 15% to 39% differences it would need to explain.
The connections-versus-households distinction is not a material issue. The Utility Regulator's connection numbers are very similar to the households reported by ONS: a ratio of 1.01 in 2017, a 1% difference. Even the largest ratio in the period, 1.08 in 2014, falls well short of the 15% to 39% differences it would need to explain.
It appears not. If BEIS unit costs excluded fixed charges that could contribute to the puzzle, but the BEIS methodology confirms, in its exact words, that "unit costs reflect the prices of all suppliers and include standing charges". Nor is the payment-type mix the answer. The BEIS methodology states:
We cannot rule it out, but the candidate explanations do not hold up. Combining different sources is risky, so we tested the three simplifying assumptions underpinning Tables 2 and 3, stating each before examining it.
BEIS publishes the average unit cost across all payment types, and even using the charge for credit customers, the most expensive group, paying about 7% more than direct debit customers in 2018, would not fully explain the puzzle. Northern Ireland's unit costs are lower than other regions for all three payment types across the period (Figure 3).
Figure 3: Payment methods do not seem to explain the puzzle. BEIS unit prices by region for credit, prepayment and direct debit customers, 2014 to 2018; Northern Ireland sits below the other regions on all three payment types.
The three panels close a loophole. If the BEIS average had been flattered by a cheap payment-type mix, the puzzle might have dissolved into composition effects; instead Northern Ireland sits below every other region on credit, prepayment and direct debit alike, across the whole period. Whatever explains the puzzle, it is not who pays how.
One part of the BEIS methodology merits further investigation, in terms of what is meant by its statement that "No allowances are made for introductory offers or non-cash benefits that may be available from suppliers". But if such offers would reduce BEIS unit costs, allowing for them would increase the puzzle rather than explain it. Lastly, BEIS confirm the data includes VAT, ruling out a potential 5% understatement.
In our view, unlikely. The Utility Regulator's consumption and connection data come from Northern Ireland Electricity Networks, whose consumption figures are partly estimated because not all meters are read on 31 December. Estimation could plausibly understate the December quarter, when consumption runs high, but by the same logic other quarters would be overestimated, so on an annual basis the net effect should be no systematic bias.
To explain the puzzle, the network operator's estimates would need to understate true consumption materially, every year, for four consecutive years. Northern Ireland's demand has held up in a way Great Britain's has not: electricity sales there are slightly above their 2004 level, while sales in England and Wales have fallen by 22% and in Scotland by 34% (Figure 4). Estimating consumption is harder where demand is not falling steadily, which makes understatement somewhat more plausible in Northern Ireland than in Great Britain. Nonetheless, in the absence of anything more tangible, systematic underestimation seems, in our view, unlikely.
Figure 4: An index of electricity sales (volumes using TWh) to consumers, 2004 = 100. Electricity sales volumes indexed to 2004 for Northern Ireland (104% by 2024), England and Wales (78%) and Scotland (66%): Northern Ireland is the only one of the three where sales are above their starting level.
The divergence is striking. Over twenty years Northern Ireland's electricity sales ended slightly above where they began, at 104% of their 2004 level, while England and Wales fell to 78% and Scotland to 66%. The path is not a steady climb: Northern Ireland peaked at 114% in 2009, fell back to 98% by 2018, and only regained its starting level after 2021. The chart matters as the paper's honest counterweight to its own verdict, since meter estimation is likelier to lag reality where demand is not falling, making understatement by the network operator somewhat more plausible in Northern Ireland than in Great Britain. The paper still judges four consecutive years of material understatement unlikely, but this is the evidence that keeps that judgement a judgement rather than a proof.
This is where the least can be ruled out, so we examined the ONS approach in three parts: whether the sample is drawn correctly, whether it is large enough to be precise, and whether the fieldwork collects accurate answers. We put each to the Northern Ireland Statistics and Research Agency, which runs the Northern Ireland fieldwork on ONS's behalf, and their responses are reported below.
It is difficult to see how it could be the problem. ONS explains its sampling methodology in these words, quoted in full in the report:
NISRA's view is that the Northern Ireland approach is, if anything, more robust than Great Britain's: the systematic random sample is not restricted to postcodes, as the Great Britain sample is, and is therefore less constrained in reflecting households across population areas and house sizes.
It could be the explanation, but the data needed to confirm it is not published.
Before 2016/17 only about 150 Northern Ireland households were captured each year. The sample has since been boosted, in NISRA's words, to improve precision, to around 400. The boost was requested and funded by another department, the Economic and Labour Market Statistics Branch, and may in due course revert to 150, which NISRA note would still be proportionate to sample sizes in other regions, at about 0.02% of households. ONS also publishes three-year averages for each region rather than annual values, precisely to improve the accuracy of the data.
Even so, ONS does not publish standard errors for individual expenditure categories by region. At the whole-of-Great-Britain level the standard error for electricity costs is small, 1.1% for the year ending March 2018, but the error for a regional subsample the size of Northern Ireland's would be much larger. We estimate it would need to exceed 7% for the ONS values to be statistically consistent with the consumption the Utility Regulator reports. NISRA could not share the underlying survey data but agreed to request it from ONS.
Not demonstrably. ONS describes its fieldwork approach in these words, quoted in full in the report:
The published methodology dates from 2012, so we sought confirmation from NISRA that it still reflects practice. NISRA confirmed that interviewers attend households, conduct computer-assisted personal interviews and collect two-week expenditure diaries, although electricity costs are not captured in the diaries. Participation incentives matched Great Britain, £20 for adults and £5 for children, with experiments using £40 under way in Great Britain to test response rates; in NISRA's view there is no systematic fieldwork or incentive difference between Great Britain and Northern Ireland. Interviews run throughout the year, so seasonal timing is unlikely to bias results, and interviewers are trained to avoid rounding bias by seeking bills or statements and pounds-and-pence accuracy.
The survey is onerous, up to 90 minutes covering almost 200 cost categories for a £20 incentive, which could limit the time respondents spend finding their electricity bills, and NISRA agreed the surveys are demanding exercises. NISRA could not share the survey instrument, so we could not examine its design directly, but on the published methodology and our discussions with NISRA the fieldwork is not, in our view, biased in any clear way that would systematically overstate weekly costs.
Two objections apply whatever the answer on any one of them. A well designed survey is not necessarily as accurate as sources of Northern Ireland data that are more focused and contemporaneous, such as network and supplier records. And whatever the sample's quality, a genuinely random sample should sit close to the population average, yet the average implied consumption and the average reported consumption are far apart, so sampling size or bias alone cannot account for the gap.
Of the three, only statistical uncertainty survives as a candidate, and it cannot be tested until ONS releases the regional standard errors.
They point the same way as BEIS. Budget Energy, a Northern Ireland supplier with more than 49,000 domestic customers, 99% of its customer base, offers an independent cross-check: dividing total revenue from its financial accounts by customer numbers gives an average spend per customer (Table 6).
Table 6: A cross-check of weekly electricity costs using Budget Energy accounts. Average spend per Budget Energy customer, derived from the supplier's financial accounts (revenue divided by customers, then by 52 for the weekly figure), against the ONS weekly and annual costs, with the gap shown for each year.
Compared with the ONS information, the Budget Energy cross-check implies a much lower electricity cost, with a gap of a similar magnitude, more than 25%, to Table 3, reinforcing our characterisation of the missing £150 per year. The Budget Energy spend per customer of around £600 or less per year is also consistent with the BEIS data in Figure 2.
This cross-check has limitations. The periods do not exactly align, since Budget Energy's accounts cover years ending June while ONS values cover 36-month periods ending March. Budget Energy, as a newer supplier, needed to offer cheaper tariffs than other Northern Ireland suppliers to attract customers, and those energy-savvy switchers may consume less electricity than the Northern Ireland average.
Table 7: The four possible explanations, and where the evidence points. The paper's argument in one table: each candidate explanation for the puzzle, the tests applied to it, and the verdict reached.
The table shows why the paper's conclusion is worded carefully: three explanations are positively tested and largely closed, while the fourth is not proven but is the only one the evidence cannot close, and the independent supplier cross-check points the same way. That is the difference between claiming ONS is wrong and concluding, as the paper does, that the ONS figure is the likely outlier to be treated with caution.
In our view, the household cost reported by the Office for National Statistics (ONS) is unlikely to be as accurate as other estimates from the Department for Business Energy and Industrial Strategy (BEIS) or Utility Regulator (UR). We are unable, however, to confirm the source of bias affecting ONS' data.
Official statistics present contradictory pictures of electricity costs in Northern Ireland. Our analysis suggests that the BEIS and Utility Regulator data are mutually consistent, whereas the ONS estimate appears to be the outlier. Although the precise source of bias cannot be identified, supplier accounts provide an independent cross-check that supports the lower cost estimates.
.jpg)
The 1995 review cut Britain's distribution charges in real terms for the first time, by 11% to 17%, then reopened in 1996 after the Northern Electric bid.
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
Block quote
Ordered list
Unordered list
Bold text
Emphasis
Superscript
Subscript
Great Britain privatised electricity distribution in 1990, not 1989. The twelve Regional Electricity Companies were sold that December at 240 pence per share, with Retail Prices Index plus X price controls already written into their licences.
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
Block quote
Ordered list
Unordered list
Bold text
Emphasis
Superscript
Subscript
I was delighted that MCC's work was completed on time, and within budget, helping us deliver important changes and improvements, to the benefit of our stakeholders. MCC's report is published on the CCC website.
I am delighted to recommend MCC Economics. Specifically, I worked closely with PJ, who helped us with our Nuclear and CCUS projects. PJ helped us develop new policies and answer questions from our stakeholders. His support helped us deliver important changes and improvements, to the benefit of our stakeholders.
MCC Economics has helped us better understand the most important issues for our stakeholders, including: charges, shareholder returns, debt payments and inflation impacts.
I am delighted to recommend PJ and his team at MCC Economics. We've been working together on National Policy Statements to help meet net zero targets for 2030 and 2050. We initially appointed MCC Economics to support us on offshore wind consultation analysis and have recently reappointed MCC Economics to undertake a larger consultation analysis role across all sectors, including hydrogen, CCUS and networks. I can confirm that PJ and his team have shown excellent spreadsheet skills, alongside very good project management, planning and analysis skills, helping us deliver important changes, and continuous improvements, to the benefit of our stakeholders.
I am delighted to recommend PJ and his team from MCC Economics. They helped us with our price controls for Heathrow airport and for NATS (En Route) plc (the air traffic services provider). Specifically, the MCC team helped us deliver important changes and improvements to our financial models and supporting policy documents, to the benefit of our stakeholders.
I am delighted to confirm that I worked with PJ on a retail project in 2015. The project helped stakeholders understand electricity costs and charges. Specifically, the project helped us explain to stakeholders, internally and externally, why electricity charges differed across the regions (GB, NI & Ireland). PJ was a key member on the project team, which helped deliver changes and improvements in the understanding of energy retail.
.png)
Evidence-based analysis for transparent, defensible, and effective decision-making.
© All rights reserved – MCC Economics 2026.
.jpg)
.jpg)