When in July last year the Treasury announced that the UK’s biggest lenders were revealing how much they lent at a local level, like most people who work closely in trying to understand and address financial and social exclusion, we in Birmingham felt that this was a significant breakthrough.
As Danny Alexander stated at the time, this had the potential to be a major step forward in terms of transparency with the potential to ‘enable smaller lenders identify gaps in the market and allow businesses to hold their local bank to account where they aren’t lending’.
After major campaigns by the likes of Community Investment Coalition, the data was finally released in December through a framework agreed by the British Bankers Association, Council of Mortgage Lenders and the treasury. The data provides details at postcode sector level of more than £1tn of lending across Great Britain for residential mortgages, loans to small and medium businesses and unsecured personal loans.
In Birmingham we were keen to look at this data to see what it told us about how banks lend in local areas of the city, in particular in relation to our deprived localities and neighbourhoods known to be vulnerable to financial exclusion. The information has the potential to support a lot of our existing work, such as interventions supporting people who have no access to mainstream banking services and to stopping people going to high-cost lenders. Towards the end of last year the Birmingham Fair Money campaign was launched, uniting not for profit lenders and credit unions with the aim of disrupting the growth of high cost lending in Birmingham. This was followed by the launch of the fair money manifesto.
We therefore decided to analyse the data released by the banks, with a particular focus on the personal lending data. We mapped the lending patterns across Birmingham and tried to combine it with population data and deprivation data to see if there is correlation between patterns of deprivation and more or less lending. You can download the full report here.
As with any analysis the report contains lots of lovely looking maps, tables and charts. However what does the information tell us? Can we draw any significant conclusions?
In summary it tells us a little but has the potential to tell us so much more.
This data does help establish a picture of lending practices within Birmingham, and on the surface there does seem to be some geographical disparity in lending across the city. Credit unions have noticed that some of these areas are locations where they expected there to be less lending due to bank branch closures.
However on the whole it hasn’t really given us a clear picture of the personal lending in our city.
The first big drawback of the data is that it only shows the total amount of the outstanding balances of all loans in each postcode sector. There is no other accompanying information. So it is not possible to establish how many loans are outstanding in each area, and therefore calculate the average loan value. We have used census information to establish loans per capita, however this only gives a partial picture.
Another big downside to the data is that as the data release is only voluntary, it only contains a proportion of all lending. For example participating lenders were: Barclays, Lloyds Banking Group, HSBC, RBS, Santander UK, Clydesdale & Yorkshire Bank and Nationwide Building Society. These lenders only count for account for about 60% of bank lending to SMEs, 73% of mortgage lending and 60% of unsecured personal loan markets in Britain. On personal loans it is estimated that the figures for participating lenders represent under 30% of the total national unsecured credit market.
As with many aspects of open data, the UK is significantly behind other countries in relation to financial lending data. For example we can look quite enviously across the Atlantic to colleagues in the US who have access to many loan data sets. Through the home mortgage disclosure act, the Federal Reserve release a great deal of data including information about applicant’s ethnicity, loan amount, loan denials, and a marker of whether the interest rate is high interest or not. It is released every year and provides communities with ways of looking racial and geographical disparities in lending.
There is no doubt that if we had this level of data in this country we could do a lot more. Credit unions in Birmingham, such as Citysave, have suggested that there is potential power in this information. They build their lending strategies by understanding their members and the communities they can potentially serve. To be able to compete against high interest lenders, and to build effective interventions to stop people turning to local loans sharks, this data could if it was more detailed, paint a clear picture of where to operate in the city and with which communities.
In summary, while it’s certainly a ‘thumbs-up’ that banks are starting to reveal this information, there is still a lot more that can be done – with little cost to the banks. More transparent organisations can only build more trust in the financial system as a whole. More information to local organisations can only improve how we serve our communities and protect the most financially vulnerable in society.
To add how grateful we are for the review of data, and I now have our own lending data overlaid against the bank lending data in Birmingham. It shows clearly that we are lending more pro rata than the banks in several key parts of Birmingham where there are lower income levels and greater exclusion. Areas such as Ladywood, Hodge Hill, Aston and Perry Barr are all areas where we have focused resource under the Growth Fund contract working with good partners. Similarly it isn’t a surprise that the reverse is true in areas of higher income such as Edgbaston and Sutton Coldfield where we are lending less than the banks pro rata, and where we have invested less outreach resource. What is most useful in a city that has 187 identified communities, for an organization where resource is light such as our credit union, is to look for areas where we should be focusing resource. Richards analysis has shown clearly how much more we need to do. & we are already talking with our authority about how we could do so. I would only add that we would like to see lending data from all lenders. We are happy to commit to full transparency, and believe it to be the basis of ensuring our communities have access to affordable and fit for purpose financial solutions. Thanks again, Richard.
The Community Investment Coalition (CIC) campaigns for access to fair and affordable finance for all communities. We campaigned for disclosure of lending data as we believed this transparency was critical to giving those tackling financial exclusion information about the geographic location of communities struggling to access affordable credit. Birmingham CC has made a great start in using this data to look at local lending patterns. It is important that other councils do the same, so that we can get a broader picture across Great Britain about patterns of lending by the main high street lenders. (The voluntary framework doesn’t currently cover Northern Ireland).
Richard is absolutely right to say that this data only gives us a partial picture. CIC welcomed government treasury minister Sajid Javid’s comments at CDFA’s annual conference yesterday, about the government’s commitment to transparency, the need for more data and the government’s willingness to legislate to make this happen, if need be.
The voluntary framework of lending data is a great start. But it is exactly that – a start.
Hi Richard
It looks to me that the data that’s been released does give us some more questions to ask about lending. The Treasury press release says that this excludes credit card debt and I’m not sure what the breakdown of that might be across postcodes, although I’d imagine that a lot of people in the most deprived areas don’t have credit cards.
The data will increase in value if it is released consistently over time. I can’t see anything in the press release or the first half of the mortgage lenders’ websites that gives any indication whether this is an ongoing release, nor how complete or, indeed accurate, the lenders believe their data to be..
It would be really useful if the organisations publishing these open data sets could use the Open Data Institute’s Open Data Certificates pages to give us some more detail about what they are releasing here.
Thank you for providing the analysis you’ve done on this and the blog post about it. What do you think the next steps could be so that a range of people with interests in this data can get some accurate insights to help them in their work? How does this data help CitySave identify areas where high-interest lenders are most prevalent, for instance? Because that could be really interesting and useful.
All the best
Simon
Thanks for the comments.
This is an ongoing release – if you have a look at the bottom of this press release http://www.cml.org.uk/cml/media/press/3781 it indicates that future data will be made available quarterly, relating to a date point six months in arrears. The next datasets, relating to balances at end-September 2013, will be published in April 2014.
We are already starting to do a lot of interesting analysis with the data – comparing with Experian endebtedness data, CCJ data, and other credit union datasets. The analysis with Citysave data set throws some interesting questions.
However i have to return to the point that it still only gives us a partial picture. Just the number of outstanding loans in each postcode would get us a number of steps forward. Hopefully future releases will enable us to do that.
Getting high interest lenders to release data would be a good step forward….remember the current data sets only relate to the mainstream banks.
It would also be good if the banks also adhere to the the open data principles of data release. For example HSBC releasing the data in PDF format is very frustrating
The debate continues