University of Torino1, Foundation for the Environment / Turin School of Local Regulation2
The consumption of goods with periodic payments involves an important basket of goods and services. Excluding rents, loans and mortgages a relevant share of consumers' disposable income is allocated to services with clear social and environmental implications such as natural gas, electricity, district heating, water and waste. The provision of some of these goods is characterized, for technical or social/legal limits, by a weakness of perfect excludability condition that, in most markets, is able to prevent opportunistic behaviour and free riding. Unlike goods for which there is almost perfect simultaneity between purchase occasion and payment, the perception of goods for which a periodic payment is expected suffer some distortions that obstruct market's functioning and prevent the use of traditional incentive mechanisms. We intend to present a simple and intuitive way to think at arrears and late payments either as an operating cost for utilities providing these services, and water in particular, or as stemming out from deprivation.
The use of an original and unique longitudinal dataset of household billing data from three major Italian utilities (Smat S.p.a. 50.000 observed customers for water services for five years, Iren Mercato S.p.a. 175.000 observed customers for electricity and Egea S.p.a. 34.000 observed customers for natural gas) operating in the metropolitan area of Turin (Italy) and in the Province of Cuneo (Italy), allows us to model the complex phenomenon of arrears and late payments.
METHODS, RESULTS AND DISCUSSION
The descriptive analysis of datasets suggests that a large percentage of households defers payments with regularity: 25% of the water supply bills are paid at least one day after the expiry date, 18% for electricity bills and 8% for natural gas. Nevertheless, half of these bills presents delays not exceeding 30 days representing situations of forgetfulness or liquidity problems promptly solved. The empirical investigation (Probit models for panel data) provides evidence that state dependence, i.e. households who have experienced an event in the past has higher probability to experience again the same event, occurs in the case of unpaid bills pointing out that "history matters" in forecasting arrears. In fact, 65% of households experiencing a delay in the payment of water services bill, also postpones the next payment. At the same time, 88% of those who pay regularly maintain the same behaviour for following payments.
The determinants of these payment behaviours are multiple and in most circumstances driven by economic or social vulnerability; however, according to the information provided by utilities, we verify the effects of different tools to manage disputes, the use of installment plans and different methods of payment. In the case of water services the analysis suggests that the resolution of disputes, especially through installment plans, is associated with a reduced cost of arrears for the enterprise. Interesting experiments in communicating a state of arrears and proposing a resolution (for example texting users through ad hoc mobile platforms) will be examined in the future of the research development.
If descriptives give a first overview of the phenomenon, we enrich the analysis building a compact and consistent measure of the cost of arrears for utilities based on the time between the expiry date and actual payment and the size of the credit. The use of information on administrative costs to manage arrears and an appropriate interest rate representing the opportunity cost of unavailable funds are able to synthetically represent the cost for each customer.
We treat this information using econometric models (Tobit models for panel data) confirming strong, but decreasing, state dependence. The model predicts higher expected costs, for utilities, attributable to consumers who experienced delays in previous payments.
The econometric analysis corroborates the presence of an issue for either companies providing local public service, who face the financial cost of arrears, or for policy makers who need appropriate tools to support vulnerable end users. The further research question pertains to the possibility of using billing data to derive a signal of vulnerability (or presence of a fuel poverty state). We suggest to refer to the economic and sociological literature that studies the poverty states, i.e. periods of life in which the disposable income of the family is below a certain threshold, as a reference point to analyse the dynamics of arrears.
In particular, we propose the use of the Longitudinal Poverty Index to create an index, the Turin-Index, capable to define different levels of arrearage intensity. Through the use of this measure we can provide a dynamic view of the phenomenon based on the history of payments for each household, the intensity of delays and an "emergency effect". The Turin Index is measured between 0 and 1 with the extreme values representing respectively the case of absolute arrearage and the case of "perfect payer". The values in between can be interpreted as measures of the degree of arrerarage intensity. Moreover, the index can be easily aggregated to construct city or regional measures.
In a first stage, we computed the average Turin Index for the three services considered. The Turin- Index shows the great potential to become a bridge for the exchange of information and the identification of critical situations to monitor and to revise between public and private sector; a synthetic indicator able to offer decision makers a snapshot of the phenomenon of arrears and a strong basis to build specific environmental and water policies so as to support users who experiences deprivation. 1.Aassve A., Davia M., Iacovou M. e Mazzuco S., 2007, "Does leaving home make you poor?: Evidence from 13 European countries", European Journal of Population, 23, pp. 315Â–338.
2.Boardman B., 1991. Fuel Poverty: from cold homes to Affordable Warmth, Belhaven Press. London.
3.Bossert W., Chakravarty S. R. e DÂ’Ambrosio C., 2010, "Poverty and Time", Working Paper 2010/74. United Nations UniversityÂ–World Institute for Development of Economic Research, Helsinki.
4.Bradshaw J. e Hutton S., 1983. "Social policy options and fuel poverty", Journal of Economic Psychology, vol. 3, pp. 249Â–266.
5.Calvo C. e Dercon S., 2009, "Chronic poverty and all that: the measurement of poverty over time". In Poverty Dynamics: Interdisciplinary Perspectives (eds T. Addison, D. Hulme and R. Kanbur), pp. 29Â–58. Oxford: Oxford University Press.
6.DOE (Department of the Environment - UK), 1991. "English House Condition Survey 1991: Energy Report". DOE, London, 82, pp. 129Â–137.
7.EPEE (European Fuel Poverty and Energy Efficiency), 2006. "Diagnoses of causes and consequences of fuel poverty in Belgium, France, Italy, Spain and the United Kingdom", http://www.fuel-poverty.org/documents.htm.
8.European Commission, 2003, "Laeken indicatorsÂ—detailed calculation methodology", European Commission, Luxembourg.
9.Foster J. E., 2009, "A class of chronic poverty measures". In Poverty Dynamics: Interdisciplinary Perspectives(eds T. Addison, D. Hulme and R. Kanbur), pp. 59Â–76. Oxford: Oxford University Press.
10.Giarda E., 2011, "Persistency of financial distress amongst Italian households: evidence from dynamic probit models". Quaderni di Dipartimento. Serie Ricerche, Dipartimento di Scienze Statistiche "Paolo Fortunati", UniversitÃ di Bologna, Bologna, IT.
11.Estache A., Gomez-Lobo A. and Leipziger D., 2001, "Utilities Privatization and the Poor: Lessons and Evidence from Latin America," World Development, vol. 29(7), pp. 1179-1198.
12.Hills J., 2011. "Fuel Poverty: The problem and its measurement", Interim Report of the Fuel Poverty Review. Centre for the Analysis of Social Exclusion, London, Case Repot 69.
13.ISTAT, 2011. Indagine sui consumi delle famiglie. Roma.
14.Mendola D., Busetta A., Milito A.M., 2011, Combining the intensity and sequencing of the poverty experience: a class of longitudinal poverty indices, Journal of the Royal Statistical Society, 174(4), pp.953-973.
15.Poggi A. e Florio M., 2010, "Energy deprivation dynamics and regulatory reforms in Europe: Evidence from household panel data", Energy Policy, 38, pp. 253-264.
16.Whelan C. T., Layte R. e MaÃ®tre B., 2003, "Persistentincomepoverty and deprivation in the European Union", Journal of Sociology and Policy, 32,pp. 1Â–18.
17.Wooldridge, J. M. 2005, "Simple solutions to the initial conditions problem in dynamic, nonlinear panel data models with unobserved heterogeneity", Journal of Applied Econometrics, vol. 20, pp. 39Â–54.