Willingness to Pay for Sanitation Services in Dagupan City, Philippines 1 D. S. Harder*, A. U. Sajise** and E.M. Galing*** * Resources Environment and Economic Center for Studies (REECS), Quezon City, Philippines (E-mail:
[email protected]) ** University of the Philippines at Los Banos (UPLB), Los Banos, Laguna (E-mail:
[email protected]) *** WB-WSP, Ortigas, Metro Manila, Philippines, UK (E-mail:
[email protected]) Abstract Using the Contingent Valuation Method (CVM), the Study looks at the septage management and sewerage 2 services demand of households with latrines in Dagupan City, Philippines vis a vis the cost of providing these services. The survey covers a total of 1200 respondents for the septage and sewerage demand study. Under the Willingness to Pay (WTP) Study, certainty corrections and protest vote screening were applied to WTP responses to control for hypothetical bias. For the septage and sewerage studies, income was shown to significantly influence WTP. WTP was also found to increase across income groups, indicating the plausibility of implementing a socialized pricing scheme for the septage and sewerage fees. Another important implication of the results is that increased demand for sanitation facilities would only take place as general income levels of Dagupan City improves. Under certain assumptions, the individually rational and financially viable Septage Fee is around PhP 46.00/ mo with optimal desludging frequency of three years. This means that a self-financed Septage Program is possible for the city. However, the case is different for the Sewerage program since none of the proposed fees (including the average WTP of Php102/mo for the whole sample) is sufficient to cover the huge investment costs associated with a self-financed Sewerage infrastructure. Thus, the LGU has to source funds elsewhere. Keywords Septage; Sewerage; Willingness to Pay (WTP); Sanitation 1.0 BACKGROUND This study on household sanitation behavior and demand for sanitation services in Dagupan City is part of the continuing activity under the Program for Sustainable Sanitation for South Asia (SuSEA). This is implemented in six pilot provinces nationwide that include the municipalities of Bauco, Alabel, Guiuan and Polomolok as well as the cities of Dagupan and General Santos. The overall objective of SuSEA is to increase the access of poor Filipinos to sustainable sanitation services. It is composed of two national components: 1) Establishment of a national sustainable sanitation program (NSSP); and 2) Formulation of a national sustainable sanitation communications and hygiene promotion program (NSSCHPP). Among the objectives of the NSSP is for all provinces and cities and half of the municipalities to come up with a sustainable sanitation program, as well as a 100% increase in sustainable sanitation investments by 2016. This links closely with the objectives of the study that aims to understand the sanitation decisions households make and their willingness to pay for sanitation services. In particular, using the contingent valuation method, the study looks at the septage management and sewerage 3 services demand of households with toilets in Dagupan City vis a vis the cost of 1 This study is an initiative of the SuSEA Program of the World Bank Water and Sanitation Program (WB-WSP) with Coffey International as the resource agency/contractor of the Study 2 Sewerage demand covers only the Central Business District 3 Sewerage demand covers only the Central Business District providing these services. It reveals what people say they would do given the hypothetical scenario of providing sanitation services in their area. 2.0 THE CONTINGENT VALUATION METHOD (CVM) 2.1 The CV Design Contingent valuation, a method used to study household preference and behavior, directly asks people in a survey how much are they willing to pay (or willing to accept) for a good or service. It is referred to as ‘contingent’ because people’s willingness to pay (WTP) responses is anchored on a specific hypothetical scenario and description of the good/service to be provided. The hypothetical market includes a statement of the proposed change (i.e. sanitation services) and an institutional mechanism in which the proposed change is to be provided and financed (i.e. waterbill). Since not all goods and services are traded in the market (i.e. reduction in water pollution), information on market demand is usually unavailable, thus making Contingent Valuation a useful and practical decision-making tool for valuing well-known goods. Initial applications of contingent valuation studies in the Philippines have been on water and sanitation (Choe et. al. 1996; Lauria et. al. 1999). 2.1.1 The Contingent Valuation scenario The two hypothetical CV scenarios presented in the survey includes the septage and sewerage management plans. This is compared to the existing practice or status quo (Table 1). Table 1. CV Scenario for Septage and Sewerage Study Status Quo Management Plan Septage Management Plan Only <5% of hh in Dagupan a have desludged their septic tanks 100% households w/ septic tanks will be desludged Coverage of the Plan Sewerage Plan 0% sewerage connections for households & establishments 100% sewerage connections of hh and establishments in the Central Business District Septage Management Plan DENR composting facility (Sual Pangasinan) Septage Treatment Facility @ Dagupan City Where sludge/ wastewater is brought Sewerage Plan Canals that eventually discharges to the river Wastewater Treatment Plant (WwTP) @ Dagupan City Septage Management Plan Private desludgers LGU & Dagupan Water District or authorized desludger Management Sewerage Plan None LGU & Dagupan Water District or private operator Septage Management Plan One time payment fixed fee per month added to water bill Payment Scheme Sewerage Plan None Volumetric (based on water consumption) Frequency of desludging Septage Management Plan As needed Once every 3-5 yrs (scheduled desludging by area/zone) 2.1.2 Elicitation method In framing the WTP question, the referendum dichotomous choice (DC) elicitation format was used. This format uses bids that are randomly assigned to respondents and asking them whether they would vote to pay the assigned amount that will be added to their monthly water bill. They could either accept or reject the bid offer. This type of elicitation format is realistic since individuals typically make purchase decisions based on fixed market prices (‘take it or leave it’). This also minimizes if not totally eliminates the incentive for respondents to lie or to engage in strategic behavior in order to influence the provision of the good. The hypothetical scenario further includes a script that reminds respondents of their budget constraint to get more valid WTP responses. Certainty corrections were further applied to WTP responses. In particular, respondents who said that they are willing to pay for the sanitation service but are unsure of their response are re- classified as ‘no’ respondents. This is usually done in Contingent Valuation analysis to correct for hypothetical bias. Overall, ‘no’ responses increased from 781 to 831 after the WTP responses to the offered bid were adjusted for certainty. The data was also subjected to protest vote screening. The study only considered valid ‘no’ responses if the reason for voting no is due to lack of preference for the good or income constraint. 2.1.3 Setting of bid prices One of the crucial decisions to make in any CVM study is the selection of bid prices. The research team did an initial focus group discussion with households. The initial bids were then pre-tested to 151 randomly selected households from 4 barangays in Dagupan, namely: Barangays 2 & 4 (within the Central Business District), Barangay Lomboy (island barangay) and Barangay Mangin (interior/river barangay). The choice of pre-test sites took into account the income disparities between zones or barangay classification. The final bids used for the survey are Php50, Php100, Php150, Php200 and Php250. 2.2 General Empirical Strategy: CVM and the Random Utility Model (RUM) The random utility model is the overall framework used to analyze the survey data. The model posits that given two choices (i=1,2), a household chooses the first alternative over the other if the utility gained from the first alternative is higher than the utility derived from the other. If we let V i be the indirect utility gained from choosing alternative i, then rationally choosing alternative 1 means that: Note that the choice problem falls under the class of discrete or dichotomous choices. Since the analyst only observes the actual choice and some but not all variables that affect the choice, a random error term ε i is included and V i is considered as the deterministic component. The probability of observing choosing alternative 1 can be written as: An interesting application of this model is a hypothetical scenario used to analyze and explain the stated preferences of households for the proposed Sanitation Programs in Dagupan City (i.e. Septage and Sewerage Management Plans). These services will be available for a household in exchange for a reduction in income. Suppose that the household is offered a price or bid (b) in exchange for the hypothetical service. Agreeing to pay the bid or price implicitly implies that: Pr(y 1 = 1) = Pr V 1 x 1 , m 1 ( )> V 2 x 2 , m 2 ( ) ( ) where: v 0 is the status quo utility v 1 is the utility after the proposed change t i is the bid offered z i is a vector of demographic variables (e.g. education, age, gender, etc.) y i is respondent i’s income level θ i is a vector of other control variables (e.g. classification) Assuming a linear utility function we can restate the earlier probability statements. Specifically, we can rewrite the status quo indirect utility and decompose it to its’ deterministic and random components (ε ij for j=0,1): (2) Similarly the indirect utility function for the proposed change is expressed as: (3) Notice that in the previous equation we are subtracting the bid from the income level. This is how we capture the trade-off from reducing income but paying some amount t i . Using equations (2) and (3), we can alternatively write equation (1) as: (4) or Simplifying difference terms for the coefficients yields: (5) The next step would be to estimate the simplified β coefficients. This involves assuming a specific functional form for the probability distribution Pr(•). A common assumption is the logistic function. The log likelihood function is derived through taking the summation the log of the logistic function for each observation or respondent. The usual estimation procedure for obtaining the estimated coefficients is the Maximum Likelihood Estimation (Greene, 2002). Finally the Mean WTP is calculated using the following formula: (6) 2.2.1 Some Econometric Issues There are two econometric issues associated with the nature of the data for the study. The first one involves the nature of the CVM data for the Septage Management Plan. In particular, different hypothetical scenarios were offered to households with and without water connections. Connected households were offered the hypothetical scenario of whether they would be willing to pay some bid amount for the SMP. Non-connected households, on the other hand, were offered a hypothetical scenario of whether they would want to be connected or not. This was followed by the same hypothetical scenario offered to connected households. We conjecture that there is a potentially a larger hypothetical bias for non-connected households because they essentially face two hypothetical scenarios. The likelihood of actually experiencing the septage infrastructure is less real for this subgroup of respondents. A consequence of this is that the probability of being offered a two-staged scenario is conditional on the household being connected at the time of the interview. Since, the choice of being connected could be linked to certain household characteristics, we expect that some will be more predisposed of having water connections. This leads us to the potential problem of sample selection. Simply pooling connected and non-connected households and following the specification outlined above for the septage sample can lead to inconsistent estimates and hence biased WTP values. To address this issue we first estimate an endogenous switching regression model where both the selection and outcome variables are discrete or dichotomous. Consider first the selection mechanism. A household chooses to connect to DCWD water system if: The outcome variable on the other hand, pertains to the household’s answer to the CV scenario. The household agrees to pay the offered bid if: Since, the error terms of the selection and outcome are hypothesized to be correlated we can observe the conditional choice of agreeing to a bid as: Instead of estimating the conditional probability of observing an agreement to a bid, we can opt to estimate the joint probability. To do this we can assume that both error terms are normally distributed and estimate a bivariate probit expressed in equation terms as follows: Looking at whether rho is significant or not at an assumed confidence level easily tests the presence of sample selection. If rho is statistically different from zero then the decision to connect is Pr(connection = 1) = Pr(V connected (g) + ε connected > V not connected (g) + ε not connected ) = Pr(ε ' > V not connected (g) −V connected (g)) where ε ' = ε connected − ε not connected Pr(yes) = Pr(V yes (g) + η yes > V status quo (g) + η status quo ) = Pr(η' > V status quo (g) − V yes (g)) where η' = η yes − η status quo Pr(yes | connection = 1) = Pr (η' > V status quo (g) −V yes (g)), (ε ' > V not connected (g) −V connected (g)) Pr(ε ' > V not connected (g) −V connected (g)) Pr (η' > V status quo (g) −V yes (g)), (ε ' > V not connected (g) −V connected (g)) = φ(η, ε, ρ)dηdε ε ' ∫ η' ∫ where φ is the bivariate normal density function necessarily related to the decision of agreeing to a bid offer. A simple application of the econometric model in the previous section to a pooled sample of connected and non-connected households will lead to biased estimates of WTP. 2.2.2 Variables Used in the Various Analysis The covariates used in the implementation of the empirical strategy are shown in Table 2: Table 2. Variables Used in the Regression Model Variables Description bid Bid offered to respondent and assumes the following values: 50, 100, 150,200, 250 know Respondent’s score on sanitation issue test age Age of respondent civil Civil Status of respondent: 1 - Divorced/separated 3 - Married 5 - Widowed 2 - Live in 4 - Single educ 1 has value of 1 if respondent finished some elementary; 0 otherwise educ 2 has value of 1 if respondent has graduated from elementary; 0 otherwise educ 3 has value of 1 if respondent finished some high school; 0 otherwise educ 4 has value of 1 if respondent graduated from high school; 0 otherwise educ 5 has value of 1 if respondent finished some years in College; 0 otherwise educ 6 has value of 1 if respondent graduated from College; 0 otherwise educ 7 has value of 1 if respondent have post-graduate schooling; 0 otherwise educ 8 has value of 1 if respondent finished Vocational; 0 otherwise hseown has value of 1 if respondent/ family of respondent fully owns the house; 0 otherwise agehouse Age in years of respondent’s house ageown Interaction term between age of house and house ownership: agehouse*hseown yrslive Number of years household has lived in the house yrslive2 Square of the yrslive diarr has a value of 1 if any household member has had diarrhea in the past three months; 0 otherwise buss has a value of 1 if a business is being run in the house; 0 otherwise Income 1 has value of 1 if household’s monthly income is <PhP 4,999; 0 otherwise Income 2 has value of 1 if household’s monthly income is between PhP 5000 and PhP 9,999; 0 otherwise Income 3 has value of 1 if household’s monthly income is between PhP 10,000 and PhP 20,000; 0 otherwise Income 4 has value of 1 if household’s monthly income is between PhP 20,001 and PhP 35,000; 0 otherwise Income 5 has value of 1 if household’s monthly income is >PhP 35,001; 0 otherwise hh_all Total number of household members age_septic Age of septic tank Class 1 has a value of 1 if baranggay is classified as a Central Business District baranggay; 0 otherwise Class 2 has a value of 1 if baranggay is classified as a Coastal/ Inland baranggay; 0 otherwise Class 3 has a value of 1 if baranggay is classified as a Island baranggay; 0 otherwise Class 4 has a value of 1 if baranggay is classified as a River baranggay; 0 otherwise 2.3 Sampling Frame For the septage and sewerage study, a total of 1200 4 respondents were sampled, distributed between 850 and 350 respondents, respectively. A 12% non-response rate was noted for the whole sample. The allocation of the sample is based on the following formula: The above equation provides the least cost allocation for a fixed sample size given the variability in income. We assume that cost differences are insignificant since Dagupan’s terrain is homogeneous and its’ land area (4,446 has) is not that large. Thus, this variable can be ignored in the computation. Variability in income 5 is based on barangay clusters, which is the only available disaggregated information from the baseline study 6 . However, for the sewerage study, a simple proportional allocation was used because barangays under this scenario belong to the same type/cluster. Thus, the ‘income variability’ cancels out in the equation. For the septage study, sample allocation for each barangay was further categorized into households connected and not connected to the Dagupan City Water District (DCWD). Considering all this, the sample size allocation for the septage management and sewerage options are shown in Tables 3 and 4, respectively. Table 3. Sample Size Allocation, Sewerage Respondents Name No. of Households Type of Barangay Total survey Households Herrero Perez 509 CBD 23 Lucao 1395 CBD 64 Pantal 3087 CBD 142 Poblacion Oeste 683 CBD 31 Barangay I (T. Bugallon) 124 CBD 6 Pogo Chico 903 CBD 42 Tapuac 913 CBD 42 TOTAL 7614 350 Table 4. Sample Size Allocation, Septage Respondents 4 The total sample takes into account cost and time considerations in doing the survey 5 Incorporating income variability is important since demand is basically determined by income and prices. 6 2007 Baseline Study Report: Dagupan City. Sustainable Sanitation in East Asia (SuSEA). It would have been ideal to use the income variability for each barangay, but since the raw data from the baseline study was not available, further disaggregation is not possible. n i = N Ni σ i ci Ni σ i ci k ∑ | \ | ¹ | | where N is the total sample size N i is the total population in the ith income strata σ i is the income variance for the ith income strata c i is the cost of obtaining a sample from the ith income strata Name No. of Households Type of Barangay Non- connected Households a Connected Households Total Households (survey) Bacayao Norte 289 River 3 8 11 Bacayao Sur 307 River 3 9 12 Bolosan 660 River 6 19 25 Lasip Chico 166 River 2 5 7 Lasip Grande 267 River 3 8 11 Mamalingling 230 River 2 7 9 Mayombo 1222 River 12 36 48 Pogo Grande 380 River 4 11 15 Salisay 336 River 3 10 13 Tambac 424 River 4 12 16 Bonuan Binloc 1185 Coastal & inland 8 25 33 Bonuan Boquig 1978 Coastal & inland 14 42 56 Bonuan Gueset 3085 Coastal & inland 22 65 87 Caranglaan 1281 Coastal & inland 9 27 36 Malued b 1567 Coastal & inland 11 33 44 Tebeng 463 Coastal & inland 3 10 13 Calmay 943 Island 7 21 28 Salapingao 418 Island 3 10 13 Herrero Perez 509 CBD 6 19 25 Lucao 1395 CBD 17 51 68 Pantal 3087 CBD 38 114 152 Poblacion Oeste 683 CBD 8 25 33 Barangay I 124 CBD 2 5 7 Pogo Chico 903 CBD 11 33 44 Tapuac b 913 CBD 11 33 44 TOTAL 212 638 850 a - 25% of total household covered b - In the actual survey, the sample size in Barangay Tapuac was adjusted from 86 to 69 since is was found out that 78% of the houses in the area are apartments and excluded from the survey. The remaining respondents were transferred to Malued with sample size increasing from 44 to 61 (all hh). 3.0 SURVEY RESULTS The discussion of survey results is divided into two parts. First is a qualitative description of the data in terms of environmental and health hazards that relate to the design, construction and maintenance of existing household sanitation facilities. The second part outlines the results of the econometric strategy discussed in the methodology section. 3.1 Overview of Current Sanitation in Dagupan City 7 3.1.1 Environmental problem/priorities Base on household ranking, the more visible environmental threats that the local government should give priority to include solid waste (36%) and flooding (30%). It is common knowledge that majority of the barangays in Dagupan are vulnerable and susceptible to the disturbing effects of inundation. Poor sanitation (26%) only ranks third among the top environmental problems of Dagupan City from the perspective of households respondents. This is not only in terns of lack of access to latrines but also by the threat posed by badly designed sanitation facilities including the lack of proper sludge and wastewater disposal systems. Receiving water bodies often absorb the pollution load coming from households. Thus, it is not surprising that DENR-EMB included Dagupan River as one of the rivers nationwide that pose contact risk to public health and the environment. It has shown a very high total and fecal coliform level 8 , about 30 times higher than the DENR standard. 3.1.2 Current sanitation facilities & the environment Survey results show that 43% of households have single vault septic tanks that do not conform to the standard design prescribed under the Sanitation Code of Dagupan City. These vaults provide only minimal treatment and limited sludge storage compared to two or three chambered septic tanks. Although most of households reported that their septic tanks are water sealed (at least for one chamber), 15% of the remaining septic tanks are bottomless or do not have concrete flooring. This could lead to wastewater seepage. In terms of location and access, the Sanitation Code of Dagupan City further stipulates that septic tanks ‘should not be constructed under any building and shall be located such that desludging equipment can have convenient access to the manhole’. Nonetheless, almost half of the respondents interviewed had septic tanks located inside the house, with some directly under their toilets. A third further reported that their septic tanks are more than 20 meters away from the nearest access road. This concern however, can be addressed by using longer vacuum pumps (i.e. vacuum pumps of Manila waters has a maximum length of 80 meters). The only constraint is the width of the access road that should be able to accommodate the desludging truck. This could be a problem particularly for the island barangays. Another disturbing observation is the 13 meters average distance of the nearest water source to the latrine, which is lower than the specified distance of 25 meters and above under the Clean Water Act. However, the survey shows that only 6% of the respondents are familiar with this provision under the Act. In general, the close proximity of toilets and water sources coupled with poorly constructed septic tanks could lead to groundwater contamination that is hazardous to public health. A faulty septic tank can also lead to hydraulic overloading 9 particularly during flooding. 7 This section uses the original database (n=1200) as basis for discussion 9 Hydraulic overloading can force the waste out through the septic tank before it receives adequate treatment. It can lead to anaerobic conditions in the drainfield and might not give solids time to settle out before being pushed through the system (http://www.stormwatercenter.net/Assorted%20Fact%20Sheets/ Tool7-Non_Stormwater/SepticSystems.htm; accessed 4/28/2010). Most respondents perceive that water from the septic tank goes underground (48%) or stays inside the tank (38%) 10 . Only 9% believed that wastewater from septic tanks eventually finds its way into the water bodies. More than three fourths of the respondents still had the misconception that septic tanks can provide 100% treatment for sludge and wastewater, thus controlling for water pollution. For the septic tanks to function as designed, it has to undergo regular maintenance through the removal of built up solids. However, the survey shows that only 13% of households have desludged their septic tanks for the past 10 years. This is despite majority of respondents agreeing to the statement that septic tanks should be desludged once every three to five years as required under the Sanitation Code of Dagupan City 11 . Desludging is done mainly as a response to emergency situations such as septic tank overflow, clogging or damage caused by flooding, earthquakes, etc. Poor maintenance reduces the effectiveness of septic tanks to treat waste before it is discharged to the environment. Also, there is no existing sludge treatment and disposal facility within Dagupan. The nearest DENR approved composting facility is in Sual, Pangasinan. This resulted to indiscriminate disposal of untreated or poorly treated sewage into the environment. The survey shows that of those that had their septic tanks emptied, 40% do not know where the septage is brought while 30% buried their waste in a pit (i.e. manual desludging). The rest disposed of the waste in nearby water bodies (13%), in a farm (5%), or anywhere (7%). This practice is prohibited under the Sanitation Code of Dagupan City, which stipulates that, ‘no discharge of septage or sludge shall be allowed in manholes, drainage areas, canals, creeks, rivers or other receiving bodies of water or land’. Unceremonious discharge of untreated sewage into vacant spaces and receiving water bodies provides a vector for pathogens that could transmit disease to humans. In terms of health impacts, 10% of the respondents reported incidences of diarrhea in the past three months 12 . Meanwhile, occurrences of other waterborne diseases are negligible and no reported outbreaks were observed during the duration of the study. 4.1 Econometric Results This section covers the Willingness to Pay (WTP) of Dagupan City households for the Septage Management Plan, and WTP of the Central Business District (CBD) households for the Sewerage Plan. The final models used are corrected for both uncertainty and protest votes (CC/PC model). In particular, the discussion includes characteristics of the respondents, results of the CV analysis including simulation exercises that helped identify some important provisions of a potential Septage Management Program and Sewerage Plan. For instance, we determine the most efficient desludging frequency (only for the SMP) and pricing policy. Efficiency is referred to in the context of an intertemporal decision wherein discounting or time preference plays an important role. Likewise it also refers to that which is rational for the individual household as well as financially feasible to the Local Government. 4.2.1 Implementing A Septage Management Program in Dagupan City After screening for protest votes, 790 useable responses were used for the WTP analysis under the Septage Management Plan (7% lower than the original sample). This includes both households connected and not connected to the Dagupan City Water District (DCWD). 10 These respondents view septic tanks as simply a storage of sludge and urine rather than a primary treatment system 11 Apparently, this item, despite approval from respondents, gained the most comments. Those who got it wrong were respondents who had larger septic tanks with only few household users. 12 Three percent of respondents also had diarrhea in the past week of the survey. On average, they spent Php604 for treatment, including in patients. Respondent Characteristics (pooled sample) Based from the socio-economic profile of households, those that are willing to pay for the Septage Management Plan are relatively younger. Also, nearly a third are married with an average of six members per family. In addition, majority of these respondents are relatively well off in terms of asset & income. Nearly a fourth earns monthly income greater than Php20,000 compared to 16% of those who are not willing to pay. Households who support the Plan are likewise better educated. In terms of location, a higher percentage of septage respondents from the Central Business District and island barangays are willing to support the SMP while a greater number of refusals to pay were observed from coastal/inland and river barangays. Likewise, respondents who are willing to pay have older houses and septic tanks and have lived longer in their current residence. In general, septic tanks have a mean age of 16 years for both subgroups. Respondents from the island barangays used to have hanging latrines, some access public or shared toilets within their neighborhood while others had to reconstruct their toilets and septic tanks after the damages brought by typhoons Ondoy and Pepeng, including the 1990 earthquake. Bid Distribution Cursory evidence shows that in general, percent of respondents saying ‘yes’ on willingness to pay for the Septage Management Plan decreases with the offered bid (Fig 1), thus conforming to theoretical expectations. 59.28 36.42 26.92 18.00 10.32 0 10 20 30 40 50 60 % y e s r e s p o n s e 50 100 150 200 250 Bids (Php) Fig 1. Bid Distribution of ‘Yes’ Responses, the Septage Plan Similarly, percentage of those saying yes to the offered bid increases with income (Table 5). Table 5. Distribution of WTP Responses by Income Bracket, Septage Respondents Willingness to Pay Income bracket No % no Yes % yes All Income 1 74 69.81 32 30.19 106 Income 2 191 73.75 68 26.25 259 Income 3 194 69.53 85 30.47 279 Income 4 62 63.27 36 36.73 98 Income 5 25 53.19 22 46.81 47 Total 546 100 243 100 789 a a – excludes one don’t know response for income Willingness to Pay for a Septage Program As elaborated earlier, a bivariate probit model was used to test whether the error terms in the selection mechanism and the outcome decision (i.e probability of agreeing to the offered bid) are correlated. A simple likelihood ratio test shows that the correlation parameter (ρ) is not significantly different from zero. Thus, we can run a logit regression on the pooled sample of connected and non- connected households. The results of the logit regression for the Septage study are shown in Table 6. The negative and significant coefficient for the bid offer indicates that respondents are less likely to pay for higher bids. Younger respondents are also more likely to pay for the sanitation service. Besides this, ones’ knowledge on sanitation issues increases the probability of saying yes to the bid offer. In particular, knowledge regarding the provisions of the Sanitation Code of Dagupan City as well as general knowledge on wastewater hazards and sanitation can increase the probability of agreeing to a bid offer by 13 percentage points. The probability of agreeing to a bid is also positively related to the level of education. This implies that well-informed households tend to agree more to the offered bid, perhaps, because these households can comprehend the scenario better. We can see this because all the categorical variables for education (educ 2 to educ 8) are significant and exhibit positive signs. Another determinant of the probability of agreeing to an offered bid is the age of the house. The variable is negatively related with the probability of paying the bid offer. That is, households with older houses have a lower probability of supporting the Septage Management Plan. This is not surprising since old houses are prone to abandonment. The information obtained from the logit analysis is used to compute for the amount that households on average are willing to pay (WTP) for septage service improvements (Table 7). Calculations show that the mean WTP of Dagupan City residents is around PhP 73. Note that WTP increases with income. For instance, households with monthly income in excess of 35,000 would be willing to add PhP 121 to their monthly water bill to avail of the improved septage services. While households in the lowest monthly income bracket (below PhP 4,999) are willing to pay an additional PhP 67 per month. Furthermore, the WTP for each income bracket are significantly different from each other at 5% level of confidence. The variability in the WTP across income brackets is suggestive that a “socialized” pricing scheme can be implemented. However, not all of the WTP amounts shown in Table 8 would pass a referendum. Estimates show that for the whole sample and across income brackets, only 50% on average would vote for a referendum that will add the septage fee to the existing water bill. The calculated bid or proposed price and the probability of a yes vote are shown in Table 8. From the table, inorder to have at least 60% of households across income brackets agreeing to a new septage ordinance, the proposed septage fee should only be around PhP 46. The discussions to follow will revolve only on the bids that would pass a referendum, i.e. bids PhP 0.50 to PhP 73.00 Table 6. Results of Logit Regression for the Septage Program Variables Coefficients bid -0.0134*** Know (score) 0.6555* age -0.0196*** civil 0.0492 educ 2 1.4595* educ 3 1.8127** educ 4 1.5965** educ 5 2.0786** educ 6 1.9787** educ 8 2.2788** hseown -0.0014 agehouse -0.5830* ageown 0.0021 yrslive 0.0022 yrslive2 0.0093 diarr 0.0001 buss 0.3662 Income 2 0.2794 Income 3 -0.3054 Income 4 -0.0790 Income 5 0.2850 hh_all 0.4004 age_septic -0.0078 Class 2 -0.2574 Class 3 0.0269 Class 4 -0.3794 constant 0.1418 *** significant at 1%; ** significant at 5%, * significant at 10% Table 7. Calculated Willingness To Pay for the Septage Program, by Income Bracket and for the Whole Sample Income bracket (N=790) WTP/mo Income 1 67 Income 2 55 Income 3 73 Income 4 102 Income 5 121 Whole Sample 73 Table 8. Probability of Paying/Voting for Different Proposed Fees Probability of Paying (%) Proposed Fee (PhP/ mo) 30 133.80 40 101.60 50 73.00 60 46.00 70 23.00 80 7.00 90 0.50 99 0.00 The Efficient Price and Desludging Frequency We are now left with the question on the relevant price to charge households and the optimal desludging frequency. To answer this, we will use the Benefit Cost Analysis approach. BCA assumptions are outlined in Appendix Table 1. The welfare gains households derive from different fees and desludging frequencies are shown in Table 9. The figures are welfare gains in present value terms of different pricing schemes and household desludging frequency. The Status quo represents the present value of paying the average commercial desludging price of PhP 2,500. The first thing to notice is that the range of septage fees widens as the desludging frequency is shortened. For example, if the Dagupan LGU imposes that households desludge their septic tanks once in every 10 years, then households will only rationally participate in that scheme if Septage fees are between Php 0.50 and PhP 7.00. This is because they will be better off with the status quo of hiring a commercial desludger and paying him the average of PhP 2,500 (or PhP 1,060 in PV terms) every 10 years. On the other hand, if the mandated desludging is every three years, then the Septage Program will be more appealing to households. We expect, given the current cost assumptions that households will participate if the Septage fees are between PhP 0.50 to PhP 46.00. Fees falling between these ranges are shown to be superior to the status quo. What drives the results of the BCA analysis is the discounting or assumptions on households’ time preferences. Increasing the interval between desludging (i.e. reducing desludging frequency) favors the status quo because they are smaller in present value terms. Table 9. Welfare Gains of Households from Various Proposed Fee and Desludging Frequencies 3 Year Desludging 5 Year Desludging 10 Year Desludging Bid/ Price (PhP/ mo) PV (Proposed Septage) PV (Status Quo) Difference PV (Proposed Septage) PV (Status Quo) Difference PV (Proposed Septage) PV (Status Quo) Difference 73 2,636 2,066 -570 3,653 1,708 -1,945 5,921 1,060 -4,861 46 1,661 2,066 405 2,302 1,708 -594 3,731 1,060 -2,671 23 831 2,066 1,236 1,151 1,708 557 1,865 1,060 -805 7 253 2,066 1,813 350 1,708 1,357 568 1,060 492 0.5 18 2,066 2,048 25 1,708 1,683 41 1,060 1,020 An attendant question is which of these proposed septage fees would be feasible to implement? Earlier, we have answered the question on which fees and desludging frequency is rational for the household. We found out that the three-year interval will yield the largest gains for the households. Given this, we will analyze which of these Septage fees would be financially viable to implement along with the three-year desludging plan. Another BCA was conducted using the same assumptions in Appendix Table 1. Given the assumptions, results show that the financially viable Septage Fees are PhP 46/mo. and PhP 73/mo. By financial viability, we mean that the initial investment costs are recovered (fully paid) within 10 years and that yearly operation and maintenance (O&M) expenses are also recovered within the lifetime of the project, which is set at 25 years. These are the only Septage Fees that will result to a positive Net Present Value (NPV). Given this, the only rational and financially viable Septage Fee is PhP 46. Thus, with the current information at hand, a Septage Fee of PhP 46/ mo. is recommended. Households in turn will receive regular scheduled desludging of their septic tanks every three years. Table 10. Financial Viability of Different Proposed Septage Fees Bid/ Price (PhP/mo) NPV (3 year Desludging) 73 103,152,095.48 46 28,591,923 23 -34,922,297.86 7 -79,106,103.73 0.5 -97,055,775 Comparison of the Proposed Septage Program with Other Existing Programs The recommended Septage Program would charge PhP 46/ mo. or approximately Php 2/ cu. m. if the average household water consumption were 24 cu. m. per month. These charges are similar with the computed price by the LINAW project in Dumaguete CIty. In terms of monthly water bill, the proposed fee is close to the partial charges of Manila Water. It roughly constitutes 11% of the average household water bill of PhP 416.20 (Table 11). Table 11. Charges for Septage and Sewerage Programs in the Philippines Manila Water Sewerage and sanitation programs are funded partly by 10% environmental fee charged to all MWSS customers on the water bill. Households covered by sewerage are also assessed 50% of household water bill. This is not enough to cover O&M of sewerage (perhaps only 60-70%). There is also some cross- subsidization from water supply income Dumaguete City Dumaguete will cover its O&M costs for its septage treatment facility through a user fee of PhP 2/m3 on the water bill. This user fee will cover both capital and operating costs, as well as funding an environmental fee Zamboanga City Metro Zamboanga Water District sets its sewerage charges at 50% of the water bill, and has a 99% collection rate, allowing it to fully recover its O&M costs. Others Other cities charge a flat rate (or zero) tariffs, collect revenues lower than their O&M costs and, are dependent on subsidies from the LGU or, where managed by a Water District, on cross-subsidies from water supply income. Source: USAID (2007) All in all, the proposed Septage Management Plan is not far from existing and planned Septage Programs in the country. The innovation is that the numbers obtained came from a more rigorous methodology. Furthermore, we are sure that the recommendations are based on household stated behavior or preferences. 4.2.2 Implementing the Sewerage Plan 13 in Dagupan City For the Sewerage Plan, the useable sample for the analysis is 322 after protest screening (8% lower than the original sample). Respondent Characteristics Households who are willing to pay are relatively more mature in age, have older houses and septic tanks and have lived longer in their current residences relative to respondents who are not willing to support the Sewerage Plan. Both, however, have almost the same number of household members per family, number of families living in the house and number of houses owned within Dagupan City. For both subgroups, nearly three fourths are female and married. Respondents who are willing to pay are also more educated as two thirds have reached college. Thus, they may be more familiar with sanitation issues and can better appreciate the Sewerage Plan presented to them. A higher percentage of WTP respondents also had income under brackets 4 & 5 or income greater than Php20,000/month (28%) vis a vis those who are not willing to pay (16%). Bid Distribution Similar with the Septage Plan, the figure below exhibits a negative relationship between the offered bid and the percentage of respondents who said ‘yes’ to the Sewerage Plan. This observed trend 13 The CBD area is estimated to generate 2,772 cubic meters of wastewater per day, necessitating the need for a maximum design capacity of 6,00 cubic meters/day for the wastewater treatment plant. This takes into account population growth and the corresponding increase in wastewater generation. The projected cost of the STF would be around PhP 250-300 million. Nonetheless with additional construction of sewer lines/pipe networks, total investment could increase up to Php500-600 Million to fully address the sanitation problems of Dagupan City (SuSEA 2008) conforms to economic theory. 71.64 51.61 29.23 22.73 19.35 0 10 20 30 40 50 60 70 80 % y e s r e s p o n s e 50 100 150 200 250 Bids (Php) Fig 2. Bid Distribution of ‘Yes’ Responses, Sewerage Plan Table 12 also shows a positive relationship between income and willingness to pay. Table 12. Willingness to Pay for the Sewerage Plan by Income Brackets Willingness to Pay Income bracket No % no Yes % yes All Income 1 24 77.42 7 22.58 31 Income 2 57 65.52 30 34.48 87 Income 3 83 61.03 53 38.97 136 Income 4 23 52.27 21 47.73 44 Income 5 9 39.13 14 60.87 23 Total 196 61.06 125 38.94 321 Willingness to Pay for the Sewerage Plan Results of the logit regression for the Sewerage Program are shown in Table 13. For the final model used, only the bid, knowledge, and Income 4 estimates were found to significantly differ from zero. The bid variable shows a theoretically predicted sign. Specifically, the probability of a ‘yes vote’ decreases with the increasing bid offer. Income 4 is the only significant income bracket. This means that compared to the relatively poorest households in the CBD, households with income between Php20,000 to Php35,000 have higher probability of agreeing to the bid offer. All other income brackets have the same probability of agreeing to an offered bid relative to income 1. The knowledge variable is also significant but it does not have the predicted sign. This counterintuitive result indicates that households with more knowledge on sanitation issues are less likely to pay. One explanation would be that these households might have already been doing actions to improve their level of sanitation. Hence, they may be reluctant to pay more for a new program. Since, we have no information on the specific sanitation decisions of households other than what is associated with decisions affecting their septic tanks, we were not able to control for this in the regression model. In terms of the marginal effects of these variables, income bracket 4 has the largest effect. Households in this income bracket are 33 percentage points more likely to vote for an offered bid than those in other income brackets. On the other hand, increases in the offered bid by PhP 50 reduce the probability of saying yes by only 0.03 percentage points. The marginal effect of the knowledge variable is around 28 percentage points. Thus, the probability of agreeing to a bid for more knowledgeable households are lower by 28 percentage points compared to less knowledgeable households. The caveat in the previous discussion however, still holds in the analysis of the marginal effects for this variable. The computed WTP for the Sewerage Program for the whole sample and for each income bracket is shown in Table 14. Computed WTP per income bracket are significantly different from each other at 5% level. The mean WTP for the whole sample using the certainty and protest corrected model is PhP 102. What is interesting to note, is that unlike the trend observed for the WTP for Septage, the WTP for Sewerage increases with income for the first four brackets and then declines with further rise in income (bracket 5). Table 13. Results of Logit Regression for the Septage Program Variables Coefficient bid -0.013*** know -1.261** age 0.010 civil -0.187 educ 2 1.222 educ 3 0.442 educ 4 1.464 educ 5 1.504 educ 6 1.704 educ 8 1.883 hh_all -0.073 hseown -1.101 agehouse 0.106 ageown -0.111 yrslive 0.026 yrslive2 -0.000 diarr 0.499 buss -0.288 Income 2 0.492 Income 3 0.801 Income 4 1.391** Income 5 0.970 age_septic -0.001 class 2 0.223 constant 1.254 *** significant at 1%; ** significant at 5%, * significant at 10% Table 14. Calculated Willingness To Pay for the Sewerage Program, by Income Bracket and for the Whole Sample, in Php Income WTP/mo (Php) Income 1 48 Income 2 83 Income 3 103 Income 4 157 Income 5 137 Whole Sample 102 Going back to Table 14, the mean WTP for the whole sample has a 50% chance of being voted and is therefore expected to pass the referendum. Table 15 shows the proposed fee and the corresponding probability of payment. If 60% of households will vote for the Sewerage program, then the proposed legislation should charge a sewerage fee of PhP 73.00. The sewerage fees households would be willing to pay range from PhP2 to PhP102 per month. Note that these values are higher than the fees for the Septage Management Program. Table 15. Probability of Paying for Different Proposed Fees, Sewerage Program Probability of Paying (%) Proposed Fee(PhP) 30 164 40 129 50 102 60 73 70 45 80 18 90 2 Which of these referendum-passing fees would be financially viable? Given assumptions in Appendix Table 2, BCA results show that the prospects for an LGU financed Sewerage Program for Dagupan is less optimistic. Given the huge investment outlay 14 , none of the proposed fees are financially viable (Table 16). Table 16. NPV of Different Referendum Passing Fees, Sewerage Program Additional Fee/ month Price/ cu. m. NPV 102 4.3 -151,422,433 73 3 -187,779,355 45 1.9 -222,882,591 18 0.8 -256,732,139 2 0.1 -276,791,131 14 Note that we have conservative cost and slightly exaggerated revenue assumptions Comparison of WTP Estimates With Other Studies The bleak scenario for a self-financed Sewerage Program begs the question of whether the estimates are indeed far off from other studies. The only other study on household’s willingness to pay was conducted in 1993 15 . Table 17 shows that in general, the WTP for Sewerage is in the range of the values obtained by the study. However, even if we use the estimates for Dagupan from this 1993 study, it would not still be sufficient to cover the huge investment requirements for the proposed self-financed Sewerage Program. Table 17. Other WTP Estimates, Philippines Location Sewer Connection (PhP/mo/hh) Sewer Connection + Treatment (PhP/mo/hh) Calamba 124 103 Davao 62 92 Dagupan 169 207 Source: Philippine Environment Monitor (2003) 4.2.3 Household Perceptions on the Hypothetical Scenario for the Septage and Sewerage Plans Payment Vehicle Respondents were asked of their opinion regarding the hypothetical elements contained in the CV scenario, particularly the payment vehicle. Waterbill was chosen due to its wider coverage, mandatory nature (i.e. disconnection as effective sanction against non payment), credibility, and close link with sanitation services 16 . The Dagupan City Water District is also supportive of the Plan and is willing to collect sanitation charges to its customers. Overall, respondents from the two subgroups agree with the idea that they will pay for the septage and sewerage fees through their monthly waterbill. The major reason why some households did not agree to the current payment vehicle is the continuous increase in the water rates. We learned that the DCWD increased its rate per cubic meter months before the survey thereby creating a negative impact to respondents’ decision making. An alternative payment vehicle considered by some respondents includes a separate bill for septage and sewerage fees that will be handled by a private collector, so that the computations of the charges will be clear to them. Another suggestion would be to add these charges to the garbage fee (Php30/mo) collected by each barangay. Respondents find the barangay solid waste management program successful and the collection of fees efficient. The idea of collecting the sewerage and septage fees through community tax certificates was also forwarded (Table 18). Belief in the success of the Program/Plan Majority of the respondents believe that the Septage and Sewerage Management Plans, once implemented, will succeed in attaining the goal of reducing environmental pollution and health hazards associated with poor sanitation. In fact, some are already curious as to the design, location and actual start of the Project. Those that remained skeptic on the plausibility of the Plans gave three major reasons: a) distrust on the LGU due to possible corruption of funds; b) impact to others who may not be able to afford the fees and; c) failure of other people to cooperate (Table 19). 15 The date of the original study was cited in the Philippine Environment Monitor, 2003. However, there were no citations of the exact study in the publication. 16 However, the disadvantage of using this payment mechanism is the timing of increase in water rates before the survey period. Table 18. Respondent’s Perception to the Payment Vehicle & Alternative Schemes Septage Sewerage All N % N % N % Agreement of using waterbill as payment vehicle No 242 30.63 82 25.47 324 29.14 Yes 548 69.37 240 74.53 788 70.86 Reasons for not agreeing Water bill is continuously increasing 140 57.85 58 70.73 198 61.11 Not all have water connection 29 11.98 4 4.88 33 10.19 Don’t like mandatory payment 39 16.12 15 18.29 54 16.67 Prefers annual rather than monthly collection 10 4.13 1 1.22 11 3.40 Don’t like DCWD service 1 0.41 - - 1 0.31 Other people will not understand how the fee will be computed 6 2.48 4 4.88 10 3.09 All fees should be shouldered by the gov't 6 2.48 - - 6 1.85 Will migrate soon 2 0.83 - - 2 0.62 Separate bill instead of adding to water bill 9 3.72 - - 9 2.78 Alternative payment proposed Don't know 14 5.79 3 3.66 17 5.25 Monthly electric bill 5 2.07 2 2.44 7 2.16 Community tax 8 3.31 11 13.41 19 5.86 Land tax 1 0.41 3 3.66 4 1.23 Barangay collection 35 14.46 15 18.29 50 15.43 Garbage fee 20 8.26 11 13.41 31 9.57 Government fund 2 0.83 4 4.88 6 1.85 None 116 47.93 15 18.29 131 40.43 Own separate bill 41 16.94 18 21.95 59 18.21 Table 19. Respondent’s Belief in the Success of the Septage & Sewerage Plans Septage Sewerage ALL N % N % N % Belief in the success of the program No 166 21.01 63 19.57 229 20.59 Yes 624 78.99 259 80.43 883 79.41 Reasons for no response 1112 Money will be used for other purposes 1 18 7 11.11 8 3.85 Lack of trust in the LGU due to corruption 70 42.17 22 34.92 92 44.23 LGU lask technical skill to operate STF 4 2.41 2 3.17 6 2.88 Water district may not agree to collect payment 8 4.82 2 3.17 10 4.81 Project difficult to implement 10 6.02 8 12.7 18 8.65 Other people may not cooperate 19 11.45 12 19.05 31 14.90 Others cannot afford addt'l fee 32 19.28 11 17.46 43 20.67 Effect on Water Connection to DCWD Table 20 shows that 95% and 88% of sewerage and septage respondents respectively would retain their DCWD connection even if majority will vote for the septage/sewerage fee to be added to the waterbill. One of the main reasons for keeping their pipe water connection is the difficulty of finding alternative water sources in their area. On aggregate, 5% of those who choose to disconnect to the DCWD opt to get water from handpumps, neighbors with pipe connection and from water vendors. 4.0 CONCLUSIONS AND POLICY DIRECTIONS This study has looked at the stated preference for two hypothetical programs namely: a) Septage Management Plan that would affect households in Dagupan City; and b) Sewerage Program for households in the CBD area of Dagupan City. Several quantitative methodologies were employed to formulate policy recommendation that would be helpful in crafting a Sanitation Management Plan for the City. The empirical evidence provided in this study serves as a relevant input into the regulatory process of establishing fees for septage management and sewerage services based on consumer demand. Specifically, the results point to the following: 1. Income plays an important role on respondents’ willingness to pay. For the septage and sewerage studies, the variability of WTP across income groups indicates the plausibility of implementing a socialized pricing scheme for the fees. However, actual implementation of these fees would require a strong public consultation for the target to be met. Another important implication of the results is that increased demand for sanitation facilities would only take place as the general income levels of Dagupan City improves. Similar with the findings in General Santos City, the success of any sanitation initiative goes hand in hand with improvements in income levels and distribution. Table 20. Decision to Stay Connected to DCWD & Alternative Water Sources Septage Sewerage All N % N % N % Connection decision if waterbill will increase Don't know 38 6.42 8 2.48 46 5.03 Keep connected 521 88.01 305 94.72 826 90.37 Disconnect 33 5.57 9 2.8 42 4.60 Sure of keeping connection to DCWD Very unsure 3 0.54 - - 3 0.35 Unsure 60 10.83 41 13.06 101 11.64 Sure 373 67.33 189 60.19 562 64.75 Very sure 118 21.3 84 26.75 202 23.27 If disconnected where to obtain water? Neighbor with pipe connection 7 21.21 1 11.11 8 19.05 Public tap 1 3.03 1 11.11 2 4.76 Water vendor - - 7 77.78 7 16.67 Handpump 23 69.7 - - 23 54.76 River 1 3.03 - - 1 2.38 Deep well 1 3.03 - - 1 2.38 2. The individually rational and financially viable Septage Service Fee, given the assumptions outlined in Appendix Table 1, is around PhP 46.00/ mo. This corresponds to the optimal desludging frequency of three years. The septage fees are found to be similar to existing arrangements elsewhere in the country. This Septage fee further allows recovery of investment costs within 10 years as well as cover operating and maintenance (O & M) expenses of the Septage Program within a project lifetime of 25 years. This means that a self-financed Septage Program is possible for the city. 3. The story for the Sewerage Program, however, is not as optimistic. There is a range of bids or prices that households would agree to pay (with average of Php102/month for the whole sample). These bids are slightly higher than that of the WTP for the Septage Program. However, none of these can cover the huge investment costs associated with a self-financed Sewerage infrastructure based on BCA assumptions outlined in Appendix Table 2. Thus, it would be prudent for the LGU to look for other funds. 4. To implement the septage management and sewerage plan, waterbill can be used as a payment vehicle since its mandatory nature allows more efficient collection of fees unlike voluntary payment schemes such as community tax certificate and garbage fees. However, the computation of fees must be made transparent on the bill to allay the fear of customers of any hidden charges. If a private contractor will make a separate collection of septage and sewerage fees from the water bill, this needs to be studied carefully for cost implications. References ADB. 2007. Asian Water Development Outlook: Country Paper- Philippines. Manila: Philippines. Choe, K., D. Whittington and D. T. Lauria. 1996. “The Economic Benefits of Surface Water Quality Improvements in Developing Countries: A Case Study of Davao, Philippines.” Land Economics 72(4):519–37. Cochran, W. G. 1977. Sampling Techniques. 3rd ed. New York: John Wiley. Lauria, D. T., D. Whittington, K. A. Choe, C. Turingan, and V. Abiad. 1999. “Household Demand for Improved Sanitation Services: A Case Study of Calamba, the Philippines.” In Bateman, I. and K. Willies, eds., Valuing Environmental Preferences: Theory and Practice of the Contingent Valuation Method in the US, EU, and Developing Countries. Oxford: Oxford University Press. SUSEA. 2008. Draft Report of the Local Sustainable Plan of Dagupan City. USAID. 2007. Septage Management in the Philippines: Current Practices and Lessons Learned. World Bank. 2003. Philippine Environment Monitor: Water Quality Appendix Table 1. BCA Assumptions for Septage Management Plan Assumed Value Source Remarks Initial Investment PhP 60 M SUSEA Baseline Study Report for Dagupan City, 2008 Financing Costs 10 years to pay with 3 year grace period; Interest rate set at 10% per year Desludging and Treatment Costs Php 350/ cu. m. desludging costs; PhP 55/ cu. m. treatment costs ADB, 2007 These figures are based on calculation of Manila Water operations. Note that we have excluded the Bioapplication Costs which amounts to PhP 644/ cu. m. Salaries 960,000 Repairs 300,000 Supplies 300,000 Analysis 30,000 Annual Operating and Maintenance (O and M) Expenses Monitoring 120,000 USAID 2007 These estimates were based on the LINAW project in Dumaguete. The project is expected to serve 22,000 households and 2,500 business establishments in Dumaguete City. Since Dagupan is much larger these figures were increased by 39% for the analysis Revenue Base 75% of households have water connection; This comprises the paying population SUSEA Baseline Study Report for Dagupan City, 2008 Serviceable Households 97% of households have septic tanks will be eligible for the program SUSEA Baseline Study Report for Dagupan City, 2008 Frequency of Service 1/3 of eligible households will be serviced every year within the three year interval Average Size of Septic Tank 10.8 cu. m. SUSEA Baseline Study Report for Dagupan City, 2008 Corresponds to a septic tank with a 2x3x1.8 dimension; Only 1/3 of this amount (i.e. 3.2 cu. m.) will be left undesludged as per Sanitation Code of the Philippines Other assumptions: 1. Planning Horizon of 25 years 2. Discount rate of 10% 3. Annual Household Growth rate of 1.2% Appendix Table 2. BCA Assumptions for Sewerage Plan Assumed Value Source Initial Investment Php 500 M SUSEA Baseline Study Report for Dagupan City, 2008 Financing Costs 10 years to pay with 3 year grace period; Interest rate set at 10% per year Other assumptions: 1 Planning Horizon of 25 years 2 Discount rate of 10% 3 Annual Household Growth rate of 1.2%