From grants to loans: how social credit scoring models can prevent us from playing God
In 1989, Paul Ylvisaker’s essay Small Can Be Effective made a strong plea for a more professional type of philanthropy that goes beyond ad hoc grantmaking and considers classic financial techniques like lending, insuring and investing. Today we are all familiar with the idea of recycling money through the use of social investment, which allows the same amount of precious philanthropic money to be used for numerous people in need. The same pound can consecutively be invested in education, affordable housing and protecting wildlife habitats. Use of social credit scoring models based on traditional credit scoring models can enhance the effectiveness of this sort of social investment.
Doing good by doing very well indeed has been the new mantra of success among a circle of socially minded investors and entrepreneurs, fuelling microcredit funds and responsible investment forums galore. Traditional credit scoring techniques that are used for all of us dreaming of buying a new house or car have been applied to Muhammad Yunus’s first group of protégées, which has since attracted profit-minded investors to the sector and greatly enhanced social impact.
How social credit scoring works
A social credit scoring model calculates the probability that a loan with a charitable purpose will be paid back on the basis of a number of criteria. A mathematical equation, in which x-variables such as interest rate, duration and social programme area represent the characteristics of a social loan, will spit out a chance between 0 and 1 (the y-variable) that philanthropic investors will ever see their money again. All a social investment fund needs to do to calculate this chance is keep a detailed record of its investments in a spreadsheet: the more variables are known for a sufficiently large number of loans, the better social investment managers can predict the outcome of a social loan. The different loan variables – interest rate, duration, amount of capital on the balance sheet of the social entity, business development stage, geographical location, and social programme area – will all help the model determine the reasons behind a successful social loan. If investing in children’s education pays out more than investing in the advancement of women’s rights, a social credit scoring model will be able to indicate the statistically significant variables. It replaces the gut feeling that a lot of social investment managers, at times erroneously, rely upon.
Once a sufficiently large database has been built, the data is loaded into a statistical programme such as SAS or SPSS, known to some of us from our university days. These programmes then automatically calculate the probability that a loan will be reimbursed by the social enterprise. Such software packages can easily be downloaded for free, since even the trial versions mostly allow for up to 1,000 data points (ie loans) to be processed. Statistically, social investment funds don’t need more than 500 loans to predict the outcome of future social loans accurately. All it takes is five minutes’ work by an investment manager to input the variables for each loan into a basic tracking sheet.
Initial findings
‘The social implications of such models are enormous,’ says Karen Hadem, a consultant at McKinsey & Co’s Social Sector Office. ‘It will enable funds to make the right decision according to the risk appetite of its investors. We have witnessed a similar type of professionalization in the microcredit market, and this was certainly one of the decisive factors in the massive influx of capital. This, in turn, had a significant social impact.’ Paul Cheng, Investment Manager at Venturesome, says that developing a credit scoring model for charitable investments holds the possibility of inducing the same effects.
According to a first study,[1] using data from the US-based PRI Makers Network, it appears that smaller funds seem to outperform larger foundations, and affordable housing projects show superior results to other social investments. Hence, investing in your local fund providing housing for the poor of the neighbourhood appears to be a safer bet than, say, protecting the rainforest on the other side of the ocean. Various factors could be underlying these results.
Hadem refers to the old investment idiom not to put all one’s eggs in one basket. Smaller funds tend to invest in a broad range of social purposes, which might explain why they perform better, whereas larger funds typically concentrate their giving strategy on a limited number of social programme areas. ‘Affordable housing investments are generally granted by more experienced people,’ says Peter Berliner, CEO of the PRI Makers Network, one of the largest social investment organizations in the US. ‘They use straightforward business models in a highly developed context. Other investment areas commonly benefit from less advanced infrastructures, and use more complicated business models.’
It is also found that, counter-intuitively, larger investments are more easily reimbursed than small amounts of money, and projects conducted with the help of non-profit intermediaries seem to be a safer bet than using direct investment channels. Finally, foundations in New York seem to outperform their counterparts in California.
Part of the traditional screening process
Cheng and Berliner agree on the usefulness of credit scoring models to social investment. Berliner in particular stresses the need to build an information culture to provide social investment funds with stronger incentives to analyse and evaluate their performance. However, scoring should be used as part of the traditional screening process, as it can never replace conventional credit experience. Cheng recognizes notable similarities to the introduction of credit scoring techniques to microcredit, but alleges that scoring will not be as powerful to social investment as it is to commercial finance applications, since intuition will always have an important role to play during the assessment process. However, while many of us would feel reluctant to decide to advance cause x or y based on experience and preference alone, clean analytics could help solve the matter from a (social) business point of view.
Stated differently, these models will never earn the right to play God in addressing societal needs, but they can help prevent our social investment managers from acting like that.
1 De Bode and Van Liedekerke (2009) ‘Enhancing the Reach of Modern Philanthropy – Introducing Credit Scoring Techniques to Program Related Investment (PRI)’, Katholieke Universiteit Leuven.
Lisa De Bode is a marketer at Procter & Gamble in Brussels where she has worked on nationwide media campaigns partnering with the National Cancer Foundation. She has worked as a consultant for social investment fund CAN Breakthrough in London and developed the first credit scoring model for social purpose loans. Email lisa.debode@gmail.com
