1# RECORD 1ce55be799df6cb17762fb0541fe95340687f3de2# TITLE: The willingness to pay for access to piped water: a hedonic analysis of house prices in Southwest Sri Lanka3# YEAR: 2012.0 VENUE: nan DOI: 10.1007/s12076-012-0079-y4# AUTHORS: Caroline van den Berg; Céline Nauges5# PAGES: 16 CANDIDATE TABLE PAGES (1-based): [7, 10, 11]6# ABSTRACT: We apply a hedonic house price analysis to value connection to the piped water network in Southwest Sri Lanka based on a household survey that was conducted in 2003-04. We find that the willingness-to-pay for a piped water connection is on average about 5 percent of monthly household expenditure, which is at the lower end of the range from estimates obtained in case studies in other developing countries. In addition, the willingness-to-pay for piped water decreases as a proportion of income when income goes up. An important policy conclusion is that connection to the piped water system should be considered in relation to the availability and quality of alternative water sources. If alternative sources with good water quality are available, the need for a piped water connection is less urgent.789===== PAGE 1 (score 5, numeric-paren 0, table-words 0, var-words 5) =====1011Lett Spat Resour Sci (2012) 5:151–16612DOI 10.1007/s12076-012-0079-y13ORIGINAL PAPER14The willingness to pay for access to piped water:15a hedonic analysis of house prices in Southwest16Sri Lanka17Caroline van den Berg · Céline Nauges18Received: 3 January 2011 / Accepted: 2 February 2012 / Published online: 23 February 201219© The Author(s) 2012. This article is published with open access at Springerlink.com20Abstract We apply a hedonic house price analysis to value connection to the piped21water network in Southwest Sri Lanka based on a household survey that was con-22ducted in 2003-04. We find that the willingness-to-pay for a piped water connection23is on average about 5 percent of monthly household expenditure, which is at the lower24end of the range from estimates obtained in case studies in other developing coun-25tries. In addition, the willingness-to-pay for piped water decreases as a proportion of26income when income goes up. An important policy conclusion is that connection to27the piped water system should be considered in relation to the availability and qual-28ity of alternative water sources. If alternative sources with good water quality are29available, the need for a piped water connection is less urgent.30Keywords Household behavior · Piped water services · Hedonic pricing method ·31Development planning and policy · Sri Lanka32JEL Classification D12 · H4 · O2 · Q25 · Q52331 Introduction34The World Health Organization and UNICEF’s Joint Monitoring Programme (2010)35estimated that in 2008 about 884 million people in urban areas in developing coun-36tries lacked access to safe water. Yet, these numbers do not capture the often low37The opinions reflected in this paper are the opinions of the authors and not opinions of the World38Bank, the University of Queensland.39C. van den Berg ()40Middle East and North Africa’s Sustainable Development Department, Word Bank,411818 H Street NW, Washington DC, USA42e-mail: cvandenberg@worldbank.org43C. Nauges44School of Economics, University of Queensland, Brisbane Qld, 4072, Australia454647===== PAGE 7 (score 21, numeric-paren 0, table-words 2, var-words 11) =====4849The willingness to pay for access to piped water: a hedonic analysis5015751Table 1 Descriptive statistics for connected and unconnected households52Non-53connected54households55(mean)56Connected57households58(mean)59Mean com-60parison test61(p-value)62Average house value (in LKR) (ln)6313.016413.31650.00066Structural characteristics67Average number of rooms681.23691.28700.03371Average number of sealed toilets721.17731.25740.00175Red bricks/cement used for walls76(proportion)770.94780.98790.00080Tiles used for roof (proportion)810.82820.74830.00084Storage tank (proportion)850.46860.41870.02788Neighborhood characteristics89Average household income in90municipality (in LKR) (ln)919.7929.78930.00094Samurdhi recipients in municipality95(proportion)960.18970.20980.00099Household located in Negombo100(proportion)1010.741020.261030.000104Household located in Kalutara105(proportion)1060.721070.281080.003109Household located in Galle (proportion)1100.551110.451120.000113Average distance to pre-school (km) (ln)1140.571150.081160.019117Average distance to hospital (km) (ln)118−0.04119−0.601200.000121Average distance to road (km) (ln)1221.341231.021240.000125Average distance to beach (km) (ln)1261.391270.771280.000129Socio-economic characteristics130Average household monthly income (in131LKR) (ln)1329.421339.581340.000135Average household per capita monthly136income (in LKR) (ln)1377.931388.041390.004140Proportion of Samurdhi recipients1410.211420.151430.004144Average number of years of education145by head of household (ln)1461.931472.121480.002149Water sources150Piped water (proportion)1510.001521.001530.000154Other water sources as main source155(proportion)1561.001570.001580.0001594 Empirical results1604.1 Descriptive statistics161Descriptive statistics by connected and unconnected households are presented in Ta-162ble 1. Simple tests of means between both types of households are included (last163164165===== PAGE 8 (score 11, numeric-paren 0, table-words 0, var-words 11) =====166167158168C. van den Berg, C. Nauges169column). We only discuss significant differences. Note that 39 percent of the house-170holds are connected.171Average monthly income (including total monthly income of all wage earners in172the household, plus any other income, plus any money remitted to the household173by a family member working outside the country) for connected households is about17430 percent higher than the average of unconnected households. The mean comparison175tests also suggest that connected households tend to live in houses with more rooms176and sealed toilets than unconnected households. Unconnected households are more177likely to have tiles as roof material than connected households, while the opposite178holds for red bricks/cement for walls. Furthermore, a larger proportion of uncon-179nected households have a storage tank.180As far as neighborhood characteristics are concerned, there are significant differ-181ences in terms of distance to pre-school, hospital, beach, and distance to a paved road182between connected and unconnected households. Particularly, unconnected house-183holds tend to be located further away from these amenities. Average municipality184income is higher for connected than for unconnected households. Non-connected185households are concentrated in the Negombo district and connected households in186the Galle district. The Kalutara district is mixed.187As far as socio-economic characteristics are concerned connected households have188higher total and per capita income, and their household heads have more years of189education than unconnected households. In addition, 21 percent of the non-connected190households are Samurdhi recipients, but only 15 percent of the connected households.191Unconnected households mainly collect water from four sources: public taps, pub-192lic wells, neighbors (private tap or private well), and private wells. Nauges and van193den Berg (2009) show that the lowest time cost (to collect water) is borne by house-194holds having private wells, followed by households relying on water provided by195neighbors. It takes on average about 5 minutes to go to community sources (pub-196lic tap or well), and waiting time can be up to 24 minutes. Water from non-piped197sources, except vendors, is free. However households who collect water from wells198need equipment, which can be a bucket and a rope (for the vast majority of house-199holds), a hand pump, or more costly and technically more advanced equipment, such200as an electric pump. Overall, unconnected households are satisfied with the quality201(i.e., taste, color, smell, and safety) and service of non-piped water. More precisely,20290 percent of households collecting water from public taps, neighbors, and private203wells, believe that water from these sources is safe. This percentage is slightly lower204for households relying on public wells: 60 percent of these households think that205there is “no risk” or “little risk” in drinking water from the public well.206Connected households were also questioned about their judgment of the piped207water service. The majority believes that water supplied through the network is safe.208However, there were frequent complaints about piped water being available for less209than 24 hours a day. Indeed, in the rainy (dry) season, only 31 percent (22 percent) of210connected households have 24 hour service; and 36 percent (42 percent) have access211for 12 hours or less a day.212To sum up, satisfaction with access to and quality of water is very high in this part213of Sri Lanka, whether a household has access to a piped water connection, a private214well, or depends on a neighbor for water supply. This means that the willingness215216217===== PAGE 10 (score 28, numeric-paren 0, table-words 7, var-words 11) =====218219160220C. van den Berg, C. Nauges221Table 2 OLS estimation results of the conventional hedonic house price model222Dependent variable (ln of house appraisal value)223Coefficient224Std. err.225p value226Constant2277.575∗∗∗2281.0132290.000230Structural characteristics231Number of rooms (ln)2320.640∗∗∗2330.0522340.000235Number of toiletsa (0/1/2)2360.467∗∗∗2370.0522380.000239Wall is made of red bricks or cement (otherwise reference )2400.929∗∗∗2410.1132420.000243Roof is made of tiles (otherwise reference case)2440.143∗∗2450.0542460.008247Overhead storage tank (not available reference case)2480.323∗∗∗2490.0482500.000251Neighborhood characteristics252Mean household income by municipality (ln)2530.355∗∗∗2540.1122550.002256Percentage of Samurdhi recipients by municipality257−1.856∗∗∗2580.4142590.000260Distance to road in km (ln)261−0.071∗∗∗2620.0222630.001264Negombo district (reference case)265Kalutara district (0/1)266−0.208∗∗∗2670.0562680.000269Galle district (0/1)270−0.110∗2710.0512720.032273Water source274Household has access to non-piped sources only (reference)275Household has access to private piped water connection (0/1)2760.195∗∗∗2770.0482780.000279Number of observations2801,523b281Adjusted R-squared2820.41283∗∗∗, ∗∗, ∗Indicate significance at the 1, 5, and 10% level, respectively284aPresence of toilet: 0 = no toilet, 1 = one toilet only, 2 = two or more toilets285bThe sample on which the model is estimated is smaller than the sample of 1,552 households discussed in286Sect. 4, because of non-response in some of the explaining variables, which reduces the total number of287included observations to 1,523288Most of the remaining variables are significant at conventional levels and have289the expected signs. Structural characteristics are highly significant, and so are the290neighborhoods characteristics. Introducing socio-economic characteristics as a proxy291for impact of perceptions and knowledge shows that the connection to the piped water292network has a significant, positive coefficient, but that it is lower than in the traditional293hedonic pricing model (Table 2).8 This tendency of hedonic pricing models to provide294an upper bound of willingness to pay has been noted by different authors (Vaughan295and Ardila 1993; Carson et al. 1996).296In the next subsection we use this piece of information to derive the marginal297willingness to pay for a private piped water connection.2988Note however that the estimated coefficients of a piped water connection in the conventional and adjusted299models are not statistically different, in the sense that the corresponding 95% confidence intervals overlap.300301302===== PAGE 11 (score 26, numeric-paren 0, table-words 6, var-words 10) =====303304The willingness to pay for access to piped water: a hedonic analysis305161306Table 3 OLS estimation results of the adjusted hedonic house price model307Dependent variable (ln of house appraisal value)308Coefficient309Std. err.310p value311Constant3125.818∗∗∗3131.0963140.000315Structural characteristics316Number of rooms (ln)3170.556∗∗∗3180.0523190.000320Number of toiletsa (0/1/2)3210.334∗∗∗3220.0523230.000324Wall is made of red bricks or cement (otherwise reference )3250.847∗∗∗3260.1113270.000328Roof is made of tiles (otherwise reference case)3290.138∗∗3300.0533310.009332Overhead storage tank (not available reference case)3330.219∗∗∗3340.0483350.000336Household socio-economic characteristics337Monthly household income (ln)3380.267∗∗∗3390.0333400.000341Household is Samurdhi recipient342−0.271∗∗∗3430.0603440.000345Education of head of household in number of years (ln)3460.031∗3470.0183480.086349Neighborhood characteristics350Mean household income by municipality (ln)3510.310∗∗∗3520.1093530.005354Percentage of Samurdhi recipient by municipality355−1.553∗∗∗3560.4063570.000358Distance to road in km (ln)359−0.071∗∗∗3600.0213610.000362Negombo district (reference case)363Kalutara district (0/1)364−0.175∗∗∗3650.0553660.001367Galle district (0/1)368−0.094∗3690.0503700.062371Water source372Household has access to non-piped sources only (reference)373Household has access to private piped water connection (0/1)3740.152∗∗∗3750.0473760.001377Number of observations3781,491b379Adjusted R-squared3800.45381∗∗∗, ∗∗, ∗Indicate significance at the 1, 5, and 10% level, respectively382aPresence of toilet: 0 = no toilet, 1 = one toilet only, 2 = two or more toilets383bThe sample on which the model is estimated is smaller than the sample of 1,552 households discussed in384Sect. 4, because of non-response in some of the explaining variables, which reduces the total number of385included observations to 1,4913864.4 The marginal willingness-to-pay for a private connection387We now turn to the derivation of the marginal willingness-to-pay for connection to388the public water system. Coefficient ˆβ in Tables 2 and 3 denote the difference in ln389value between connected and unconnected houses and the absolute increase in house390value from the connection to the piped network is V NC(e ˆβ −1) or in percentages:391V NC(e ˆβ −1) × 100392V NC393= (e ˆβ −1) × 100.394395396===== PAGE 12 (score 13, numeric-paren 0, table-words 5, var-words 3) =====397398162399C. van den Berg, C. Nauges400Table 4 Predicted increase in house value after connection according to the basic hedonic pricing model401(overall, monthly equivalent, % of monthly household income)402Predicted increase in house value (LKR)403Monthly equivalent of404increase in house value405(LKR)406As % of monthly407household income408(non-connected)409Mean410[95% confidence411interval]412Median413[95% confidence414interval]415Mean416[95%417confidence418interval]419Median420[95%421confidence422interval]423Mean424[95%425confidence426interval]427Median428[95%429confidence430interval]43197,149432[47,839;151,437]433100,430434[49,454;156,550]435810436[399;1,262]437837438[412;1,305]4396.6%440[3.2%;10.2%]4417.0%442[3.4%;10.9%]443In the conventional hedonic pricing model, ˆβ = 0.195 (0.152 in the adjusted model),444which translates into a relative increase in house value (after connection) of about 22445(in the adjusted model reduced to 16) percent. However, ˆβ is estimated with error.446The 95% confidence interval for ˆβ is [0.061 and 0.247], which corresponds to a 95%447interval of [6% and 25%] in percentages when the basic hedonic pricing model is448used.4494.5 The average monthly income equivalent450We define the average monthly equivalent of the increase in house value due to con-451nection as 1/12 of the discounted annualized increase. We use a discount rate of45210 percent, which is the rate charged by a self-help group or microfinance institution.453In Table 4 we report in the first column the predicted increase in house value after454connection to the piped network (based on the estimated parameter ( ˆβ = 0.195) of455the connection dummy variable), and the monthly equivalent, in absolute terms and456as a percentage of monthly household income. We find that non-piped households457on average would be willing to pay each month about LKR 810 (median LKR 837),458which is about 6.6 percent of their monthly household income, to get a connection to459the piped network.9 When using the adjusted model that takes a better account of the460perceptions and knowledge, the results are significantly lower and non-piped house-461holds are on average willing to pay only LKR 614 (median LKR 701) which is about4625 percent of their monthly household income as shown in Table 5. These numbers463are in the range of North and Griffin (1993) for the Philippines, De Oca et al. (2003)464for Mexico, Nauges et al. (2009) for Central America, and Suparman et al. (2009) for465Indonesia.466In Table 6 we show the monthly equivalents for different income groups. The first467quartile is the 25 percent of households with the lowest per capita income, and so on.468Table 6 shows that the expected willingness-to-pay for access to a private connec-469tion decreases as a proportion of income when income goes up, although there is a470kink whereby the expected willingness-to-pay decreases from the first to the second471quartile, but it increases in the third quartile, after which it decreases in the fourth4729The total increase in house value due to the water connection is on average LKR 97,149, which translates473into a monthly value of LKR97, 194 × 1/12 × 10 percent = LKR 810; and an annual value of LKR 9,715.474