A look under the hood of ECC part 2: How our reports differ from other sources
Avid vehicle market watches will have noticed that there’s a bit of a difference between the numbers we publish and those published from other sources. In a previous blog, we set out the methodology that we use to obtain vehicle registration data for the UK. In this blog, we’re going to unpack what differences there are, and discuss a few potential explanations.
What other similar data sources are there?
Aside from our bulletin, there are two other main sources of vehicle registration information of the kind that we publish. First, the Department for Transport (DfT) publishes a Vehicles statistics series, which is based on data transmitted to it by the Driver and Vehicle Licensing Agency (DVLA). This series is not updated particularly frequently, updates are every quarter and only cover the period up to the end of the previous quarter, making the figures 3-6 months out of date. DfT is working on more frequent updates, and recently started publishing faster indicators, which includes newly registered cars. These updates are published every second Wednesday.
The other main source of data is the Society of Motor Manufacturers and Traders, who publish an update on around the 4th or 5th of the month with total registered cars and vans broken down first by fuel type and, separately, by make. The SMMT has been a near monopoly source of rapid data in this field for some time.
Differences in petrol registrations
In our October 2025 ECC bulletin, we announced that battery electric cars had outsold petrol cars. This was met with surprise by those who pointed at the statistics published from SMMT. Whereas we recorded 31,882 petrol car registrations, SMMT reported 64,360. That is a significant difference indeed. In almost every bulletin, we publish radically different petrol registration figures than SMMT.
A clue as to what is behind this difference comes when we compare SMMT’s figures with definitive data published by DfT.
A curious pattern emerges. SMMT’s figure for petrol car registrations was consistently lower than DfT’s January 2024. Ever since January 2024, SMMT reports a figure for petrol car registrations that is not only greater, but increasingly greater. In January 2024, the SMMT’s figure for petrol car registrations was 25% greater than DfT’s. By May 2025, the SMMT were reporting petrol car registrations that were 52% higher than DfT’s.
The difference in these figures represents something of a philosophical difference between SMMT and DfT. In January 2024, SMMT changed the fuel type categories they include in their monthly reports. Prior to 2024, SMMT reported figures for ‘mild hybrids’. They abolished this category and reallocated the vehicles elsewhere, seemingly mostly into the ‘petrol’ category.
Our approach follows DfT’s more closely. With the exception of the ‘hybrid’ category, which we split into plug-in and non-plug-in using the vehicle’s engine size, emissions rating, and model specification information, we follow the fuel type categories assigned by DVLA. After all, they are the public authority that regulates the vehicles.
Differences in total vehicles registered
The second key difference between our figures and other sources available is sometimes the difference between registration totals. For example, in June of this year, our bulletin recorded a total of 187,655 new cars registered. The SMMT reported 191,316 newly registered cars - a 1.9% difference. DfT’s count of vehicles was 195,673 for this month, a further 2.3% higher than the total reported by the SMMT [1].
These differences fluctuate by month. In September, we reported 314,939 registered cars. SMMT’s total was 312,819, and DfT had 320,389 [1]. The reason for the fluctuation is not immediately clear. Cars are mostly registered by an electronic system with DVLA, and DVLA is ultimately the source of each of the three respective datasets. If registration follows an electronic system, the vehicle’s addition to the database should be instant.
However, there is some reason to think that there can be backlogs of vehicles that require processing at the end of each month. We regularly see historical monthly totals inching up. When we compare our registration figures for dates greater than 3 months ago with DfT’s vehicle statistics, we get figures that are very similar, and only ever around 1% more or less than DfT’s statistics.
Whatever the reason, there is a conclusion that starts to emerge, which is that all early data is provisional, and that a sufficient amount of time is required to pass before we get a definitive or final figure for any given month. We will therefore make it clear that the latest month’s data is provisional and subject to revision. In future publications, you will see a 'provisional' label on our most recent month's data. These figures will be automatically revised in the following month's report as the final data is processed, ensuring our historical data is as accurate as possible.
[1] The DfT figure may have been subject to revision, since DfT republishes historic data each month, updating the counts as additional vehicles are added to DVLA’s database.
