The Rate of Diesel Car Ownership across Great Britain

Over the past six months, there has been growing levels of attention placed on diesel cars, due in part to the social narrative surrounding air quality levels and the increasing awareness concerning the real world emissions of local pollutants from diesel car operation. Research conducted by a variety of different organisations has demonstrated that a gap is present between the level of emissions measured by the official tests (such as the New European Driving Cycle) and those observed on the road for diesel cars (the International Council on Clean Transportation is a good source of information). This situation has generated calls for action from the government, which is viewed as being partly responsible for the widespread use of diesel cars due to a circulation tax policy which seemingly favoured the registration of diesel cars.

The figure above displays the proportion of new car registrations in Great Britain by fuel type between 2001 and 2015 (source: Department for Transport – Vehicle Statistics). There has been a convergence during this time period, going from an 80-20 split between petrol and diesel to a parity in registrations, which has persisted for the past five years. However, as air quality is a local problem, looking at national statistics only provides a partial understanding of the situation. We’re now capable of understanding the structure of local car fleets that are registered across the local authorities of Great Britain, which provides intelligence regarding where diesel cars have integrated to a higher or lower degree.

The map above reports the proportion of local authority car fleets that are diesel fuelled as of 2016.  A large range is apparent in this statistics, with Eden in Cumbria representing the local authority with the highest level of diesel registrations at 51.22% whereas the London borough of Sutton is the local authority with the lowest level at 24.27%. The proportion of diesel fuelled vehicles also appears to be strongly linked to the level of rurality, with areas of the England & Wales that are more sparsely populated tending to have high rates of diesel ownership. Interestingly, urban areas that have prominent air quality problems (such as Greater London) seem also to have amongst the lowest rates of diesel car ownership. This could have important implications for how to target a diesel car scrappage scheme, which is currently been discussed as a means through which to encourage the replacement of older diesel cars. For instance, restricting access to the scheme to those owners that reside in Air Quality Management Areas (i.e. where there’s an air quality problem) could prove ineffective if the problem is being generated primarily by individuals driving in highly polluting cars from further afield. 

Linking Emission Surcharge Cars to London’s Demographic Structure

So far, we’ve seen that there is significant geographical variation in the registrations of vehicles which are not compliant to the proposed Emission Surcharge and that this variation clusters in certain hot-spots and cold-spots across the neighbourhoods of London. It is likely that the geographical variation and clustering that we have observed is associated with the underlining demographic structure of London’s populace. In order to consider this issue, we’ve linked the car fleet compliance data with the 2011 population census to see how the rate of non-compliant vehicles is related to a number of key demographic characteristics.

The scatterplots presented above display the rate of non-compliant vehicles (on the Y-axis/vertical axis) against a number of London’s demographic attributes, some of which are protected characteristics included in the Equalities Act of 2010. Starting at the top left and then reading across and then down, we see that greater rates of non-compliance are associated with lower levels of household income, higher levels of unemployment, higher levels of people with bad or very bad health, higher levels of black, asian or minority ethnic (BAME) populaces, higher levels of muslim populaces and higher levels of households classified as lone parents. This indicates that certain demographic cohorts are likely to be more exposed to the introduction of the Emission Surcharge, meaning that the impact of this proposed policy may be greater on certain types of people as compared to others. 

Hot-spots and Cold-spots of Emission Surcharge Cars across London

From the previous two posts discussing the structure of London’s car fleet in relation to the proposed Emission Surcharge (also referred to as the T-charge) amendment to the London Congestion Charge (which will see a £10 supplementary charge imposed on cars that are not compliant to the Euro 4 emissions standard), we have seen that a wide geographical variation in the registration levels of cars coming under this regulation is present. The observation of geographical variation does not necessarily mean that there is any inherent organisation in the patterns which we see and could simply reflect a random allocation of cars across the neighbourhoods of London. In order to test whether any spatial organisation is present, we can conduct an analysis called a Local Indicator of Spatial Association (LISA) which determines if the registration levels of pre-Euro 4 vehicles in one neighbourhood is related to the levels observed in the surrounding neighbourhoods. Thus, the LISA allows for spatial clusters which share similar registration levels of cars not compliant to the Emission Surcharge regulation to be identified.

The map inserted above displays the results of the LISA and indicates that a series of spatial clusters are present. Hotspots, which are shaded in dark red, represent clusters which display relatively high levels of pre-Euro 4 car registrations. The LISA indicates that hot-spots are present primarily in the mid-west of London (both north and south of the Thames) as well as to the south-west of London. Cold-spots, which are shaded in deep blue, represents clusters which exhibit relatively low levels of pre-Euro 4 car registrations. The LISA implies that cold-spots tend to be more common than hot-spots and have diverse locations across London. Indeed, there doesn’t seem to be any clear splits in the clusters (e.g. North-South, East-West, Interior-Exterior), which is likely due to the sporadic patterning of London’s population.

The T-Charge: How does the Registration Rate of Pre-Euro 4 cars in London Compare to rUK?

With the proposed T-Charge amendment to the London Congestion Charge likely to lead to the introduction of a supplementary fee for cars which do not meet the Euro 4 emission standards, one issue to consider is if London has particularly high registration levels of non-compliant cars compared to the rest of the United Kingdom (rUK). A simple histogram of the registration rates of pre-Euro 4 cars across all of the neighbourhoods (lower super output areas) of the UK is presented in the figure below and displays a clear bell-shaped distribution, with the average registration rate of pre-Euro 4 cars being 35.5% with a standard deviation of 9.6%.

The boxplots in the figure below takes this cumulative distribution and separate it out by the Government Office Regions of the UK. Here, we see that the neighbourhoods of London tend to have higher levels of pre-Euro 4 car registrations compared to the other Government Office Regions. The difference is quite substantial when we compare highest to lowest, with London displaying an average of 41.9% registration rate of pre-Euro 4 cars whereas Scotland displays an average of 24.8%. One interpretation of this difference is that Scotland tends to have a much younger car fleet compared to London.

The Proposed T-charge Amendment to the London Congestion Zone: Where are the affected cars located?

The recently elected Mayor of London, Sadiq Khan, has tabled a proposed modification to the London Congestion Charge which will involve a supplementary charge levied against cars with older European emission bands (details can be found @ https://www.london.gov.uk/press-releases/mayoral/mayor-unveils-action-plan-to-battle-toxic-air). At the moment, the information provided by the Mayor of London's office indicates that cars which are not compliant to the Euro 4 emission standards will be required to pay an additional £10 per day to enter the charging zone. Some obvious questions to ask are how many cars will be affected by this modification and where the affected cars are located?

The figure inserted gives some preliminary statistics about London's private car fleet (data are for the end of 2015). Around 2.4 million private cars are registered in the Greater London Area, with the top ring chart showing that 41% of these vehicles are not compliant with the Euro 4 emissions standard. The bottom ring chart shows that 84% of these non-compliant cars have a petrol engine whilst 16% are fuelled by diesel. The histogram illustrates that the proportion of pre-Euro 4 car registrations varies substantially across the neighbourhoods of London (Lower Super Output Areas or LSOA).

The map inserted above shows the proportion of pre-Euro 4 cars registered per neighbourhood (LSOAs). We can see substantial differences across London. Some boroughs, such as Hackney (48.3% of the fleet) and Newham (46.7% of the fleet), have relatively high levels of per-Euro 4 car registrations. Others, like the boroughs of Westminster (35.8% of the fleet) as well as Kensington and Chelsea (33.9% of the fleet), have a relatively low level being pre-Euro 4. Illustrating the spatial variation that exists in the proportion of vehicles affected by this proposed modification represents one of the first steps in understanding the impact the amendment will have on the vehicle stock as well as households.

The Rated Nitrogen Oxide Emissions Factors of the United Kingdom’s Local Car Fleets

Recently, the discussion concerning the emissions from car operation has shifted away from being primarily focused on carbon dioxide, which is a prominent global pollutant contributing to climate change, towards the emission of local pollutants such as nitrogen oxides, particulate matter and black carbon. These local pollutants reduce air quality and can have significant impacts on public health, contributing towards respiratory and cardiovascular disease, and are become a prominent issue in local and national political debate.

 Interestingly, when you do a simple correlation analysis between the mean rated grams of carbon dioxide emitted per kilometre (gCO2/km) and the mean rated grams of nitrogen oxide emitted per kilometre (gNOx/km) of the car fleets registered across the local authorities of the UK, a insignificant relationship is estimated. This indicates that the global emissions factor (measured in mean gCO2/km) and the local emissions factor (measured in mean gNOx/km) of the local car fleets is not connected.

 The map below illustrates the mean gNOx/km of the car fleets registered across the local authorities of the UK in 2014. In this instance, it is evident that some of the areas which are experiencing air quality problems (such as the boroughs of London) have car fleets with relatively low NOx emission factors whereas some of the more rural locations of the UK (such as the South West of England and Wales) have fleets with comparatively high NOx emission factors. This finding has a number of important implications, notably that the gradual technical improvement of the car fleet in terms of NOx emissions is not necessarily transferring into improvements in the air quality in certain areas.

 

The Impact of the London Congestion Charge on Hybrid Electric Vehicle Registrations

In the last post, I illustrated the spatial adoption of Hybrid Electric Vehicles (HEVs) across the local authorities of the UK. From a casual inspection of the Choropleth map, it seems as if registrations appear to be highest in the south east of England. A technique exists which allows us to determine if there is a statistically significant spatial patterning in the registrations of HEVs. This technique is called spatial autocorrelation analysis and considers if the registrations of HEVs in a particular local authority are related to those observed in neighbouring local authorities. A visual representation of spatial autocorrelation is the Local Indicator of Spatial Association (LISA) map displayed below which strips out all of the local authorities which do not appear to be connected to their neighbours and concentrates on areas that are exhibiting significant local patterning. The interpretation of the LISA is quite straightforward, with regions highlighted in deep red (i.e. London) representing spatial hotspots for HEV adoption whilst the regions highlighted in deep blue (i.e. the South West England, Wales and the North of England) are characterised as spatial coldspots of adoption.

A question which naturally drops out of this assessment is why is London such an obvious hotspot for the adoption of HEVs? One explanation is the presence of the London Congestion Charge (LCC), which until recently had an exemption in place for HEVs meaning that owners of these vehicles did not have to pay the daily fee to enter the centre of London. To further evaluate whether the LCC has had an effect over the registrations of HEVs, two approaches seem appropriate. Firstly, we can see if the distance from the LCC seems to be connected with HEV registrations. Secondly, we can consider if interaction with the LCC (measured by the proportion of the populace who drive a car to work in the LCC) appears to be connected with HEV registrations. These evaluations are presented in the scatterplots below, and indicate that as distance from the LCC increases, registrations of HEVs tend to decrease while as interaction with the LCC increases, registrations of HEVs tend to increase. To summarise, it looks as if the exemption of HEVs from the LCC has had a significant impact on the spatial demand for these vehicles in the UK.