household demand for waste recycling services
Content
PROFILE Household Demand for Waste Recycling Services
RUSLANA PALATNIK OFIRA AYALON* MORDECHAI SHECHTER Natural Resource & Environmental Research Center University of Haifa Haifa, 31905, Israel
DOI: 10.1007/s00267-004-0044-7
ABSTRACT / Municipalities everywhere are coping with increasing amounts of solid waste and need urgently to formulate efficient and sustainable solutions to the problem. This study examines the use of economic incentives in municipal waste management. Specifically, we address the issue of recycling, if and when this waste management option is—on social welfare grounds—a preferred solution. A number of studies have recently assessed the monetary value of the externalities of alternative solid waste management options. In the present context, these subsidies could be interpreted as the implicit value of the benefits from reducing environmental externalities associated with landfilling as perceived by local government authorities. We surmise that the difference between mean householdsÕ willingness to pay (WTP) for recycling services, via the
purchase of a subsidized waste disposal facility, and the above (proxy) value of externalities reflects the difference between private and public perception regarding the negative externality associated with landfilling. We believe that this information is useful in determining the level of subsidization needed (if at all) to sustain any recycling program. The study is unique in the sense that its conclusions are based on revealed household behavior when faced with increased disposal costs, as well as information on WTP responses in hypothetical but related (and, therefore, familiar) scenarios. The article also explores the influence of the subsidization schemes on recycling rates. It was found that with low levels of effort needed to participate in a curbside recycling program, householdsÕ participation rates are mainly influenced by economic variables and age, and households are willing to pay a higher price for the recycling scheme. When the required effort level is relatively high, however, households would pay a lower price, and the rate is influenced mainly by their environmental commitment and by economic considerations. We found that in both cases a subsidy would be required in order to achieve an efficient level of recycling. The median price that households are willing to pay for recycling devices is found to be about NIS 370 (New Israeli Shekel, approximately $90).
Municipal solid waste (MSW) is a combined private and public bad, which generates negative externalities when disposed of in certain manners. In Israel, as elsewhere, the collection, transportation, and disposal is provided by local authorities and financed through municipal taxes that is by-and-large unrelated to the volume or weight of waste (and externalities) generated by a household or a firm. Thus, for example, a study commissioned by Israel Ministries of the Environment and the Interior (Biotech 1995) showed that each person produces on average 1.14 kg of waste, or 1.73 kg including yard and construction waste, daily, and the trend has been upward due to population and income growth as well as changes in lifestyle; consequently, greater than optimal amounts of waste are produced. Understandably, authorities seek alternaKEY WORDS: Municipal solid waste (MSW); Willingness to pay (WTP); Recycling, Composting; Economic incentives Published online February 16, 2005. *Author to whom correspondence should be addressed; email: [email protected]
tive, cost-saving solutions to deal with the problem of the mounting volume and weight of MSW, recycling being a politically favored option. Clearly, the higher the level of the householdsÕ participation in recycling efforts, the more cost-efficient this alternative becomes. Households, however, are not always willing to participate, especially if it involves large investments of time, space to store recyclables, and out-of-pocket expenses. The Israeli situation is compounded because of the composition of MSW. Forty-eight percent of the waste (by weight) is comprised of organic matter, yard waste, and disposable diapers (Table 1), a much higher percentage than typical to most Western countries. Consequently, the separation of these organic materials from the general MSW flow will reduce the costs of transportation and the need for scarce landfill space. It will also reduce negative externalities at the landfill, such as contamination of groundwater, release of greenhouse gases [according to Ayalon (1999), 12% of total greenhouse gases in Israel are related to methane emitted from landfills] and other noxious gases into the environment, and the spread of diseases by animals and birds feeding at the landfill site. Furthermore, the
ª 2005 Springer Science+Business Media, Inc.
Environmental Management Vol. 35, No. 2, pp. 121–129
122
R. Palatnik and others
Table 1. Composition of solid waste in Israel, 1995
Component Organic fraction Paper and cardboard Plastic Yard waste Metals Glass Disposable diapers Textile Other (rubber, leather)
Source: Biotech (1995).
Percent (by weight) 38 22 14 6 4 3 4 3 6
biodegradable materials can be turned into compost, a valuable material for agriculture in semiarid areas; it enriches soil quality and increases yields (Avnimelech 1997). For these reasons, recycling (at least the organic—or ‘‘wet’’—components of MSW), in countries like Israel, could serve as a highly viable, economically efficient policy option for MSW management (Artzi and Aharon 2001). The aim of this study was to examine the efficacy of economic instruments in promoting waste recycling schemes in Israel. To date, effectively very little use, if at all, has been made in Israel of this option in either MSW management or other environmental policy spheres. As Ayalon and others (1999) have shown, the use of such an incentive program in Israel (as elsewhere) can help reduce costs of source separation and recycling programs. In order to examine the potential benefits of recycling programs, we have decided to focus on two ongoing programs of source separation of waste and recycling, targeting separation of clean organic waste (including yard waste and diapers), in two Israeli municipalities. TivÕon is a town of 12,700 residents generating about 14,000 tons of MSW annually. Most of the dwellings are single-family homes. A voluntary policy of source separation of waste was introduced early in 2000, offering homeowners 500-L concrete bins, priced at NIS 420 (New Israeli Shekel equivalent to about 0.25 US$ at the time of the study) each, and ideally suited for separating organic ‘‘wet’’ material (food waste, disposable diapers, and light garden trimming) from ‘‘dry’’ material—the rest of the waste stream. The price embodied a 50% subsidy. Under an earlier, mandatory system, 90-L bins were posted outside each home. They were designed to accommodate ‘‘nonrecyclable’’ waste. In addition, residents were required to store certain recyclables (paper, cardboard, plastic containers, etc.) in their homes and put them out once a week at the curbside for collection. The new system made easier
demands on the residents in terms of separation efforts, because they were no longer required to store the dry waste in their homes for more than a week, but directly place the separated waste in the new containers. By March 2000, 564 out of 1864 eligible families purchased the new separation container (SC). TivÕon can be referred as a representative of 54 similar communities in Israel (by size of the population and its socioeconomic parameters), which are about 7% of the whole population (Israeli Central Bureau of Statistics, 2003). Misgav is a township with a population of 2600, living in 30 separate, nonagricultural communities. It generated 5900 tons of solid waste annually. Here, subsidized (again, at a rate of 50% of cost) home composters (HCs) were offered for sale to residents. In order to produce high-quality compost for use in their own gardens, residents must carefully separate organic material from metals, plastic, glass, and so forth. The compost must be mixed from time to time and watered in order to speed up the biodegradation process. In addition to the individual household subsidy, the regional authority has promised that each community in which at least 80% of the homes purchased HCs, would enjoy a NIS 40 rebate per household on the taxes it charges each community in Misgav for environmental services. As of May 2000, 502 HCs were purchased at prices ranging from NIS 120 to 240 (after a 50% subsidy of the full cost of HCs, which ranged between NIS 240 and 480). The price reflected the differences in size and capacity of HCs offered by the Misgav regional authority. Misgav can be considered representative of 39 similar townships in Israel (by size of the population and its socioeconomic parameters), which are about 8% of the whole population (Israeli Central Bureau of Statistics, 2003).
Related Studies
Rich literature is found on the issue of environmental policy toward reducing disposal and increasing recycling of household solid waste using economic incentives. Kinnaman and Fullerton (2000/2001) provided a good summary of most of the related researches on the subject. However, only few works were found in which householdsÕ willing to pay (WTP) for recycling was estimated using observations of actual behavior, as did the present study. Specifically, a couple of recent studies have employed a somewhat similar approach to the one employed in this study to analyze the issue of recycling in MSW management. Sterner and Bartelings (1999) analyzed the cost of waste disposal, recycling, and composting in municipalities in southern Sweden,
Household Demand for Waste Recycling Services
123
Table 2. Comparison of relevant studies
Present study 1. Subsidy on waste collection device; curbside recycling and composting 2. Based on actual WTP and survey data Swedish study US study 1. Unit pricing on waste removed; curbside recycling 2. Based on actual percentage waste recycled and survey data 3. Jenkins ‘‘Household Production Model’’ (1993) framework 4. Comparison between unit pricing and curbside collection as incentive to waste recycling
1. Tax on waste removal; source separation and composting 2. Based on actual composition and quantity of waste and survey data 3. Jenkins ‘‘Household Production Model’’ 3. Jenkins ‘‘Household Production Model’’ (1993) framework (1993) framework 4. Analysis of waste removal, recycling, and 4. Analysis of waste removal, recycling and composting composting 5. Estimation of WTP also through CVM 5. Estimation of WTP
where a unit tax system based on the weight of waste generated by each household was introduced in 1994. At the same time, the municipalities initiated an advertising campaign on the environmental merits of the program. This policy caused a significant decrease in the amount of waste collected and an increase in recycling. The researchers had access to data measuring the actual amount of waste disposed of, in addition to household characteristics from a Contingent Valuation Method (CVM) survey. They found that the important determinants of the householdsÕ behavior and waste composition are the relative difficulty of composting different materials, living space, and age. WTP was measured and compared for three different municipalities. It was found that composting had a statistically significant negative correlation with waste generation: Composting reduced the flow of waste for general disposal. Age also was a significant factor in explaining the volume of waste generated; elderly people were willing to invest more time in recycling and composting. Economic incentives were also found to be important in promoting recycling. A study by Jenkins and others (1999, 2003) falls between our study of curbside recycling, as detailed below, and the Swedish study on composting and unit pricing of waste. It analyzed two options for reducing MSW: (1) unit pricing and (2) curbside recycling of five types of waste: glass bottles, plastic bottles, aluminum, newspaper, and yard waste. The data analyzed were unique in terms of the information level on US households and included information on (1) the percentage of the five types of waste recycled, (2) characteristics of the recycling program (curbside recycling versus removal on a voluntary or mandatory basis), (3) socioeconomic characteristics of households, and (4) fees collected for removal in municipalities with a unit pricing. It was found that the access to curbside recycling and the time that passed from the introduction of the program were the two main positive and statistically significant influences on the percentage of recycling in
all five categories of waste, although the scale of the influence varied from material to material. The unit pricing option influenced only the effort to separate and recycle newspapers. Household income had a positive, statistically significant influence on the recycling effort for newspaper and plastic bottles. Generally, the higher the householdsÕ income, the larger the amount of these materials recycled. From the above comparison, it is clear that the Swedish and American studies complement each other with regard to the type of incentives analyzed by them. Both of them, as well as the study presented below, were based on a combination of observed data and observed behavior. However, only the present work specifically employs data on WTP obtained from a householdÕs stated preferences (CVM) survey and actual behavior (purchasing SC/HC at the authority proposed price). Table 2 presents and compares the three studies.
Data Description
The present study is unique in the sense that the analysis is based on combining responses from a CVM survey with observed household behavior when faced with actual choice of payment for recycling services. The database included information obtained from the municipalities on the purchases of separation or composting container (depending on the municipality) and on levels of waste removal and composting in the two municipalities. A CVM telephone survey of samples of residents in the two municipalities gathered data on several aspects of the extent of the householdsÕ awareness regarding recycling, familiarity with a variety of characteristics of the relevant waste separation system, and socioeconomic background. Respondents who stated that they owned or definitely intended to buy the relevant container were queried whether they would be willing to pay a higher price, which was randomly chosen out of five (TivÕon) and
124
R. Palatnik and others
three levels (Misgav) bid levels. In each case, the maximum bid was equal to the actual cost of the container (i.e., without any subsidy). Subjects who indicated that they would not buy the container at the actual price were asked for the reason. If it was income related, they were queried whether they would be willing to pay a lower price (randomly selected from among the options). However, in both municipalities, relatively few respondents indicated that the reason for nonacquisition was economic (e.g., in TivÕon 18 out of 131 respondents). TivÕon The sample consisted of 445 households, out of which 271 households owned SCs and 43 indicated that they intend to acquire one. Thus, the data from the 445 households were used to estimate the WTP of the total relevant population in TivÕon (denoted WTPt) and data from 314 (271 + 43) households were used to estimate the WTP of those who actually bought the container (WTPb). Misgav The sample consisted of 575 homeowners, of whom 296 purchased HCs and 19 declared their intention to buy one. Data from the 575 homeowners served to estimate the appropriate WTPt and from 315 (296 + 19) to estimate WTPb.
Figure 1. Cumulative distribution function of probability to buy at price A.
Methods for Estimating WTP for Recycling Services
The Random Utility Model (RUM) (Hanemann 1984) underlies the estimation of the WTP for recycling services function. The model posits the point at which a household is indifferent between paying to participate in the new recycling program, and not paying. This point is indicated by E in Figure 1 on the function, GE(A), depicting the cumulative distribution function (CDF) of the buying decision for any given price A (of a HC or SC). The indifference point is derived from data on individuals who did purchase the SC/HC, supplemented by information from the CVM survey of households that declined to buy at the stipulated price A. Essentially, the CDF of the purchase decision represents the household demand function for SC or HC, and, by association, the demand for recycling services. Given the high proportion of nonbuyers, we supplemented the analysis by employing Reiser and ShechterÕs (1999) spike model method for incorporating zero CVM responses.
The estimated regressions for the WTP for recycling services in TivÕon and Misgav served two objectives. The first was to identify key parameters in the characterization of communities that affect the demand for waste separation services under the two recycling programs. The second objective was to calculate the median WTP in each case [i.e., the price at which households are indifferent between purchasing a SC/HC (and thereby ensure their participation in the recycling program), and not purchasing]. To carry out the first objective, a regression for theentire population of homeowners (WTPt) was estimated; therefore, it was necessary to weigh differently purchasers and nonpurchasers in the sample according to their actual representation in the entire population. For example, 564 households in TivÕon owned SCc out of 1864 private homeowners at the time of the survey. Out of 445 private homeowners participating in the survey, 314 owned SCs. In order to obtain results that reflected the actual situation, in each run of the regression we used a weighted regression that matched the responses of owners of SC/HC with their real weight in the population. The dependent variable in the estimated regression equation is BUY, the probability that WTP is greater than zero for a household with given characteristics (i.e., the probability that the household would purchase a SC/HC at a stipulated price). In the case of TivÕon, BUY stands for the probability that a SC would be purchased at a price of NIS 420. In the case of Misgav, BUY stands for the probability that a HC would be purchased at a price ranging from NIS 120 to NIS 240. To deal with the second objective, an additional regression was estimated using responses of only those households that actually bought or intended to buy the SC/HC (regression of WTPb). In this case, the
Household Demand for Waste Recycling Services
125
dependent variable BUY is the probability of purchasing the SC/HC at a higher price than the actual stated in the explanatory variable ln(bid). For TivÕon, ln(bid) was randomly selected from five bids: NIS 500, 600, 700, 800, and 920 (equal to the full cost of the container). For Misgav, ln(bid) applied to respondents who had paid NIS 120 for the HC, selected randomly from the following four bids: NIS 150, 170, 200, and 240 (the full cost of the smallest HCs). As we noted earlier, the price of HCs in Misgav jumped from NIS 120 to 240, so for those who paid the higher price, the following three bids were used: NIS 300, 400, and 480 (the full price of the largest HC). The regression equations were estimated using the forward estimation procedure. According to Reiser and ShechterÕs (1999) spike model method, estimations of the dependent variable BUY from WTPt regression and estimated parameters of independent variables from WTPb regression were employed to derive the household demand function for SCs or HCs.
Results and Discussion
Table 3 summarizes the results of the four regressions (WTPt and WTPb for TivÕon and Misgav, respectively). All information gathered in the surveys was used in the form of explanatory variables. Maximum likelihood estimation (MLE) using the forward logistic procedure estimated the vector of coefficients. The value of each of the statistically significant variables, as they appeared in each regression, is shown in Table 3. The level of confidence of all significant variables was 95% or greater. The variables that were highly insignificant (probability lower than 0.9) were omitted from the analysis. The variable ln(bid) was used only in the regression for the WTPb; it is not relevant for the WTPt regression.
household. All socioeconomic variables had a positive effect on the likelihood of purchasing a SC. The likelihood increases if the respondent or his/her spouse is self-employed. The same also holds true for retirees. A possible explanation is that the shadow value of time of pensioners is relatively low. Our findings with respect to the age variable conform to other studies (Jenkins and others 2003; Sterner and Bartelings 1999; Ayalon and others 1999; Fullerton and Kinnaman 1995). In general, our findings regarding the positive effect of economic status variables correspond to those of other works dealing with recycling effort (Jenkins and others 2003; Ayalon and others 1999; Callan and Thomas 1999). In the regression of the WTP for HCs in Misgav, none of the socioeconomic variables was found statistically significant (Table 3). All of the explanatory variables found significant in the regression were related to the awareness of the economic and environmental benefits of recycling. As the level of awareness about recycling and composting rose, the willingness of Misgav households to participate in the recycling program grew. The difference of found influences on decision of Misgav households compared to TivÕon households might be explained by the following factors: 1. The diverse nature of communities and households represented in our dataset: Socioeconomic status of households in Misgav varied much less than it did in TivÕon (most of the family heads are about 40 years old, employed, have three children, whereas TivÕonÕs population were much more diverse). 2. Differences in the price of the relevant commodities: The SC in TivÕon costs almost twice as much as the most expensive HC available in Misgav (NIS 420 compared to NIS 240 and almost four times the price of the least expensive—NIS 120). 3. Differences in the amount of time and effort required under the two systems: In TivÕon, the citizens are just required to separate the waste, whereas in Misgav, residents need to invest time and effort in periodically watering and turning the compost; only after a year could they use the resulting compost on their gardens. Therefore, the purchase of an HC might be interpreted as a demand for additional effort. 4. Differences in the ‘‘lifetime’’ of the commodity: Misgav HCs are made of plastic and, as such, have an estimated life of 5 years. In contrast, the concrete SC proposed for TivÕon has an estimated life of 15 years or more.
WTP Regressions
WTPt The econometric results reported in Table 3 indicate that two categories of variables influenced the decision of TivÕon households to purchase SCs at a price of NIS 420. One involves household socioeconomic and demographic characteristics and the other is associated with attitude, such as environmental awareness and a positive impression regarding the design of the SC. Statistically significant socioeconomic variables in the model were type of employment of the respondent, type of employment of the respondentÕs spouse, age, and the number of cars owned by the
126
R. Palatnik and others
Table 3. Regression estimates, logit model
TivÕon WTPt Sample size (N) of private homeowners Explanatory variable Constant Employee Self-employed Pensioner Age (in years) Number of cars Participate frequently in the scheme Finds the container practical Reason for recycling is environmental Reason for recycling is economic Reason for recycling is local Babies in the family Municipal information actions are clear Separate organic fraction to save costs Separate organic fraction to recycle more Separate organic fraction to get compost Spouse is employed Spouse is self-employed Partner is a pensioner Natural log of the proposed price Number of rooms in house Income above average Received information from municipality Education (in years) Number of family members Thinks neighbors participate frequently Knows the separation rule Heard about the program Score for the municipality waste treatment (1–10) Participates for local reasons Participates to get compost Participates for ecological reason Knows that organic fraction has to be clean 445 )6.498** (0.85)*** 0.057 (0.405) 0.006 (0.471) 1.2** (0.46) 0.029** (0.29) 0.417** (0.161) 3.034* (0.5) 1.191* (0.247) — — — — — — — — 0.734** (0.376) 0.277 (0.443) 0.18 (0.467) Not used — — — — — — — — — — — — — R2 = 36%
Note. Values in parentheses are standard deviations of the variables. *a = 0.05 **a = 0.01
Misgav WTPb 314 575 )3.495** (0.282) — — — — — — — 4.645 (0.84) 3.096** (0.7) 9.501** (15.8) — 2.178** (0.384) — — — — — — Not used — — — — — — — — — — — — — R2 = 72% WTPt 315 WTPb
12.756** (3.66) — — — — 0.446* (0.207) — — — 0.493** (0.244) — — — — — — 0.945** (0.311) 0.724* (0.441) 0.708* (0.372) )2.19** (0.562) — — — — — — — — — — — — — R2 = 16%
12.87** (2.313) — — — — — — — — 0.736** (0.313) — 0.859** (0.342) — 1.13* (0.445) — — — — — )2.415** (0.429) — — — — — — — — — — — — — R2 = 24%
Household Demand for Waste Recycling Services
127
WTPb The WTPb analysis refers to 314 TivÕon households that had already purchased a SC or indicated a definite intention to do so at the time of the survey. Consistent with expectations, as the suggested price of a container rises, the probability to buy one falls (Table 3). The analysis also indicates that a higher economic status of the household increases the probability of purchasing the container at the suggested price. In addition, those households show more awareness of recycling and, especially, that source separation might help decrease costs and, therefore, increase the efficacy of waste management. Willingness to pay analysis for Misgav concerned only 315 households in the sample that had already purchased a HC or had indicated the definite intention to do so. As indicated earlier, some of the HCs in Misgav were purchased for NIS 240 and some for NIS 120. This fact might have influenced the willingness of households to participate in the recycling program and the price that they were willing to pay. Therefore, a dummy variable, LowPrice, was added to the analysis. It had value 1 if the household had purchased a HC for NIS 120 and 0 otherwise. The effect of this variable on the WTP was not found to be statistically significant. As hypothesized, an inverse relationship was found between the price of the HC and the WTP for it: The higher the proposed price, the lower the WTP [negative coefficient on log(Bid)]. Additional variables influencing the actual buyersÕ WTP, as in the general population of Misgav, were awareness of recycling and its economic and environmental advantages. Positive coefficients on the variables indicate that higher levels of awareness generally, and of the economic advantages in particular, led to a higher WTP. A new variable, appearing for the first time as significant in the analysis, was the presence of a baby in the household. The effect on WTP in this case was negative, indicating that individuals taking care of a baby were less inclined to acquire a HC and shoulder its corresponding requirements in investment of time and effort. This variable can be interpreted as a proxy for free time. The Demand for Recycling Services After accounting for zero responses using the procedure suggested by Reiser and Shechter (1999), we applied the Hanneman RUM to derive the median price on the demand schedule. It was found that NIS 369 is the price at which the TivÕon householdsÕ utility from purchasing the container and participation in the new program was equal to the utility that would be re-
Figure 2. Cumulative distribution function of the probability of purchasing a container at price A: TivÕon.
ceived from not buying the container. In other words, the median price is the point at which 50% of homeowners were willing to pay a positive amount for the SC. Figure 2 depicts the cumulative distribution function of purchasing at a given price A for TivÕon residents, including standard errors. The indicated price of NIS 369 is a relatively high one; yet, it is lower than the NIS 420, which TivÕon residents had to pay, and much lower than the full cost of the container (NIS 920). As Figure 2 indicates, 38% of homeowners in TivÕon would purchase a SC for NIS 420. In fact, by October 2002, more than half of TivÕonÕs homeowners had bought a SC. It should be noted that during the time the survey was carried out, the project was based on the voluntary will of the residents to buy the SC. However, in 2002, the municipality required those who needed to replace the old containers to purchase the new SC. We note that the estimated median price is very close to the subsidized price that was actually charged. Although households that purchased the container were prepared to pay a higher price than the median, our findings indicate that the new recycling program in TivÕon could not have succeeded without the subsidy. The demand schedule for SC can be derived from the cumulative distribution function, because the independent variable is the price and the dependent variable is the probability of buying the SC. Upon inversion of the axes of the CDF of the probability of purchasing a SC at price A, found earlier, the demand function for SC for private homeowners in TivÕon was derived (see Figure 3). We can now estimate the amount of additional recycling at the TivÕon municipality after the introduction of a subsidized SC program. The demand function for SC indicates the extent of waste recycling in response to a
128
R. Palatnik and others
Figure 3. Demand curve for private homeowners in TivÕon.
change in the subsidized price. The external costs of landfilled waste in Israel had been estimated to approximate NIS 23 per ton (Goren 1997; Enosh 1996), implying that the capitalized external cost of landfilled waste produced by a typical TivÕon household over a 15-year period (the life of a SC) amounts to NIS 549. According to the estimated demand function, 25% of homeowners in TivÕon (455 out of 1864 homes) would acquire the SC at that price (NIS 549). Hence, a subsidy of NIS 180 that would lower the cost to NIS 369 (549-180) would raise participation so that 50% of homeowners would buy one (932 homes in the case of TivÕon). Because the average household in TivÕon produces 4.56 kg of solid waste daily, acquiring a SC would increase the amount of waste recycled from 20% to 70% (i.e., from 0.91 to 3.19 kg. per day per household). A NIS 180 subsidy would therefore raise the amount of recycled waste in TivÕon by about 1 ton per day or 378 tons annually over a 15-year period. In a similar vein, one can derive the WTP per kilogram waste recycled. Given the TivÕon projectÕs price of NIS 369 per SC, the corresponding annualized sum is NIS 31 (for a device life of 15 years and at 3% interest). This sum can be interpreted as the annual payment that a representative household would be willing to pay for waste collection and recycling in TivÕon. Because Israeli households, on average, produce 2 tons of solid waste annually (Biotech 1995), TivÕon households would, accordingly, be willing to pay NIS 0.015 for recycling 1 kg of waste, over and above the charge set by the municipality for waste disposal (i.e., 10% above the current flat tax). The above analysis refers to TivÕonsÕ data only; similar analysis for Misgav can be found in Palatnik (2002).
This amounts to a marginal cost of zero on waste disposal and creates an incentive to dispose of more waste than is socially efficient. Municipalities are striving to make households aware of the cost related to waste, namely the cost of treatment and related external costs. This article analyzed the experience of a local authority in carrying out a program of source separation and recycling combined with financial participation and investment of time by residents. We found the following: (1) As the price that the household is asked to pay for waste disposal services rises, the impact of the householdÕs socioeconomic characteristics on the decision to purchase a separation container increases. (2) As the amount of effort required for participation rises, the willingness to participate falls, as well as the price that the household is willing to pay. (3) As the effort required for participation rises, environmental awareness exerts a much greater impact upon the decision to participate in the recycling program. Although it seems that households tend to state that economic reasons did not influence their decision not to take part in the program, there is definitely a significant correlation between the price of the SC/HC and the willingness to purchase the device. Households are willing to pay for a weight-base waste recycling program. Thus, TivÕon residents are willing to pay a 10% premium for waste separation. However, we found that the median WTP was significantly lower than the full price of the SC/HC. It thus seems fair to conclude that without the subsidy, the recycling programs (in both TivÕon and Misgav communities) would not have succeeded to the extent that they did.
Acknowledgment
We wish to acknowledge the generous support of Life-3rd Countries project no. LIFETCY 97/IL/ 048.
Literature Cited
Artzi, A., and Aharon. 2001. Waste recycling in Israel—Findings and targets. Israel Union for Environmental Defense. Available at http://www.iued.org.il (in Hebrew). Avnimelech, Y. 1997. Land application of composted municipal wastes. Pages 551–570 in P. N. Chremisinoff (ed.), Ecological issues & environmental impact assessment. Advances in environmental control technology. Gulf Publishing, Houston, Texas. Ayalon, O. 1999. Priorities in municipal solid waste treatment. Pages 8/1–8/17 in Y. Avnimelech (ed.), National priorities in environmental quality in Israel. The Neaman Institute Publishing, Haifa, Israel (in Hebrew).
Concluding Remarks
Waste is a nonmarket commodity. Households pay a flat fee unrelated to the amount of waste they produce.
Household Demand for Waste Recycling Services
129
Ayalon, O., Y. Avnimelech, and M. Shechter. 1999. Issues in designing an effective solid waste policy: The Israeli experience. Pages 389–406 in T. Sterner (ed.), The market and the environment: The effectiveness of market based in reply to instruments for environmental reform. Edward Elgar, Cheltenham, Cheltenham, UK. Biotec. 1995. Composition of solid waste in Israel. Survey for the Ministries of the Environment and Interior (in Hebrew). Callan, S. J., and J. M. Thomas. 1999. Adopting a unit pricing system for municipal solid waste: Policy and socioeconomic determinants. Environmental and Resource Economics 14:503– 518. Central Bureau of Statistics, 2003. Real interest rate of govermental bonds, Israel. (Available at http://www.cbs.gov.il/). Enosh Consultants Ltd. 1996. Examination of external costs of waste disposal. Ministry of the Environment, the Waste Department (in Hebrew). Fullerton, D., and T. Kinnaman. 1995. Garbage, recycling, and illicit burning or dumping. Journal of Environmental Economics and Management 29(1):78–91, July. Goren, T. 1997. Assessment of external costs in waste disposal sites. Final paper for the degree of Master of Agriculture Sciences, Hebrew University of Jerusalem (in Hebrew). Hanemann, M. 1984. Welfare evaluations in contingent valuation experiments with discrete responses. American Journal of Agricultural Economics 66:332–342.
Jenkins, R. J., S. A. Martinez, K. Palmer, and M. J.Podolsky, 1999. The determinants of household recycling: A materialspecific analysis of recycling program features and unit pricing. Draft presented in 9th Annual Meeting of the European Association of Environmental and Resource Economists, Oslo. Jenkins, R. J., S. A. Martinez, K. Palmer, and M. J. Podolsky. 2003. The determinants of household recycling: A materialspecific analysis of recycling program features and unit pricing. Journal of Environmental Economics and Management 45:294–318. Kinnaman, T. C., and D. Fullerton (2000/2001). The economics of residential solid waste management. In H. Folmer and T. Tietenberg, The international year book of environmental and resource economics 2001/2002, Edward Elger, Cheltenham, UK, pp. 100–147. Palatnik, R. 2002. Assessment of demand for recycling services of household solid waste. Masters thesis, Haifa University, The Faculty of Economics (in Hebrew). Reiser, B., and M. Shechter. 1999. Incorporating zero values in the economic valuation of environmental program benefits. Environmetrics 10:87–101. Sterner, T., and H. Bartelings. 1999. Household waste management in a Swedish municipality: Determinants of disposal, recycling and composting. Environmental and Resource Economics 13:473–491.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RUSLANA PALATNIK OFIRA AYALON* MORDECHAI SHECHTER Natural Resource & Environmental Research Center University of Haifa Haifa, 31905, Israel
DOI: 10.1007/s00267-004-0044-7
ABSTRACT / Municipalities everywhere are coping with increasing amounts of solid waste and need urgently to formulate efficient and sustainable solutions to the problem. This study examines the use of economic incentives in municipal waste management. Specifically, we address the issue of recycling, if and when this waste management option is—on social welfare grounds—a preferred solution. A number of studies have recently assessed the monetary value of the externalities of alternative solid waste management options. In the present context, these subsidies could be interpreted as the implicit value of the benefits from reducing environmental externalities associated with landfilling as perceived by local government authorities. We surmise that the difference between mean householdsÕ willingness to pay (WTP) for recycling services, via the
purchase of a subsidized waste disposal facility, and the above (proxy) value of externalities reflects the difference between private and public perception regarding the negative externality associated with landfilling. We believe that this information is useful in determining the level of subsidization needed (if at all) to sustain any recycling program. The study is unique in the sense that its conclusions are based on revealed household behavior when faced with increased disposal costs, as well as information on WTP responses in hypothetical but related (and, therefore, familiar) scenarios. The article also explores the influence of the subsidization schemes on recycling rates. It was found that with low levels of effort needed to participate in a curbside recycling program, householdsÕ participation rates are mainly influenced by economic variables and age, and households are willing to pay a higher price for the recycling scheme. When the required effort level is relatively high, however, households would pay a lower price, and the rate is influenced mainly by their environmental commitment and by economic considerations. We found that in both cases a subsidy would be required in order to achieve an efficient level of recycling. The median price that households are willing to pay for recycling devices is found to be about NIS 370 (New Israeli Shekel, approximately $90).
Municipal solid waste (MSW) is a combined private and public bad, which generates negative externalities when disposed of in certain manners. In Israel, as elsewhere, the collection, transportation, and disposal is provided by local authorities and financed through municipal taxes that is by-and-large unrelated to the volume or weight of waste (and externalities) generated by a household or a firm. Thus, for example, a study commissioned by Israel Ministries of the Environment and the Interior (Biotech 1995) showed that each person produces on average 1.14 kg of waste, or 1.73 kg including yard and construction waste, daily, and the trend has been upward due to population and income growth as well as changes in lifestyle; consequently, greater than optimal amounts of waste are produced. Understandably, authorities seek alternaKEY WORDS: Municipal solid waste (MSW); Willingness to pay (WTP); Recycling, Composting; Economic incentives Published online February 16, 2005. *Author to whom correspondence should be addressed; email: [email protected]
tive, cost-saving solutions to deal with the problem of the mounting volume and weight of MSW, recycling being a politically favored option. Clearly, the higher the level of the householdsÕ participation in recycling efforts, the more cost-efficient this alternative becomes. Households, however, are not always willing to participate, especially if it involves large investments of time, space to store recyclables, and out-of-pocket expenses. The Israeli situation is compounded because of the composition of MSW. Forty-eight percent of the waste (by weight) is comprised of organic matter, yard waste, and disposable diapers (Table 1), a much higher percentage than typical to most Western countries. Consequently, the separation of these organic materials from the general MSW flow will reduce the costs of transportation and the need for scarce landfill space. It will also reduce negative externalities at the landfill, such as contamination of groundwater, release of greenhouse gases [according to Ayalon (1999), 12% of total greenhouse gases in Israel are related to methane emitted from landfills] and other noxious gases into the environment, and the spread of diseases by animals and birds feeding at the landfill site. Furthermore, the
ª 2005 Springer Science+Business Media, Inc.
Environmental Management Vol. 35, No. 2, pp. 121–129
122
R. Palatnik and others
Table 1. Composition of solid waste in Israel, 1995
Component Organic fraction Paper and cardboard Plastic Yard waste Metals Glass Disposable diapers Textile Other (rubber, leather)
Source: Biotech (1995).
Percent (by weight) 38 22 14 6 4 3 4 3 6
biodegradable materials can be turned into compost, a valuable material for agriculture in semiarid areas; it enriches soil quality and increases yields (Avnimelech 1997). For these reasons, recycling (at least the organic—or ‘‘wet’’—components of MSW), in countries like Israel, could serve as a highly viable, economically efficient policy option for MSW management (Artzi and Aharon 2001). The aim of this study was to examine the efficacy of economic instruments in promoting waste recycling schemes in Israel. To date, effectively very little use, if at all, has been made in Israel of this option in either MSW management or other environmental policy spheres. As Ayalon and others (1999) have shown, the use of such an incentive program in Israel (as elsewhere) can help reduce costs of source separation and recycling programs. In order to examine the potential benefits of recycling programs, we have decided to focus on two ongoing programs of source separation of waste and recycling, targeting separation of clean organic waste (including yard waste and diapers), in two Israeli municipalities. TivÕon is a town of 12,700 residents generating about 14,000 tons of MSW annually. Most of the dwellings are single-family homes. A voluntary policy of source separation of waste was introduced early in 2000, offering homeowners 500-L concrete bins, priced at NIS 420 (New Israeli Shekel equivalent to about 0.25 US$ at the time of the study) each, and ideally suited for separating organic ‘‘wet’’ material (food waste, disposable diapers, and light garden trimming) from ‘‘dry’’ material—the rest of the waste stream. The price embodied a 50% subsidy. Under an earlier, mandatory system, 90-L bins were posted outside each home. They were designed to accommodate ‘‘nonrecyclable’’ waste. In addition, residents were required to store certain recyclables (paper, cardboard, plastic containers, etc.) in their homes and put them out once a week at the curbside for collection. The new system made easier
demands on the residents in terms of separation efforts, because they were no longer required to store the dry waste in their homes for more than a week, but directly place the separated waste in the new containers. By March 2000, 564 out of 1864 eligible families purchased the new separation container (SC). TivÕon can be referred as a representative of 54 similar communities in Israel (by size of the population and its socioeconomic parameters), which are about 7% of the whole population (Israeli Central Bureau of Statistics, 2003). Misgav is a township with a population of 2600, living in 30 separate, nonagricultural communities. It generated 5900 tons of solid waste annually. Here, subsidized (again, at a rate of 50% of cost) home composters (HCs) were offered for sale to residents. In order to produce high-quality compost for use in their own gardens, residents must carefully separate organic material from metals, plastic, glass, and so forth. The compost must be mixed from time to time and watered in order to speed up the biodegradation process. In addition to the individual household subsidy, the regional authority has promised that each community in which at least 80% of the homes purchased HCs, would enjoy a NIS 40 rebate per household on the taxes it charges each community in Misgav for environmental services. As of May 2000, 502 HCs were purchased at prices ranging from NIS 120 to 240 (after a 50% subsidy of the full cost of HCs, which ranged between NIS 240 and 480). The price reflected the differences in size and capacity of HCs offered by the Misgav regional authority. Misgav can be considered representative of 39 similar townships in Israel (by size of the population and its socioeconomic parameters), which are about 8% of the whole population (Israeli Central Bureau of Statistics, 2003).
Related Studies
Rich literature is found on the issue of environmental policy toward reducing disposal and increasing recycling of household solid waste using economic incentives. Kinnaman and Fullerton (2000/2001) provided a good summary of most of the related researches on the subject. However, only few works were found in which householdsÕ willing to pay (WTP) for recycling was estimated using observations of actual behavior, as did the present study. Specifically, a couple of recent studies have employed a somewhat similar approach to the one employed in this study to analyze the issue of recycling in MSW management. Sterner and Bartelings (1999) analyzed the cost of waste disposal, recycling, and composting in municipalities in southern Sweden,
Household Demand for Waste Recycling Services
123
Table 2. Comparison of relevant studies
Present study 1. Subsidy on waste collection device; curbside recycling and composting 2. Based on actual WTP and survey data Swedish study US study 1. Unit pricing on waste removed; curbside recycling 2. Based on actual percentage waste recycled and survey data 3. Jenkins ‘‘Household Production Model’’ (1993) framework 4. Comparison between unit pricing and curbside collection as incentive to waste recycling
1. Tax on waste removal; source separation and composting 2. Based on actual composition and quantity of waste and survey data 3. Jenkins ‘‘Household Production Model’’ 3. Jenkins ‘‘Household Production Model’’ (1993) framework (1993) framework 4. Analysis of waste removal, recycling, and 4. Analysis of waste removal, recycling and composting composting 5. Estimation of WTP also through CVM 5. Estimation of WTP
where a unit tax system based on the weight of waste generated by each household was introduced in 1994. At the same time, the municipalities initiated an advertising campaign on the environmental merits of the program. This policy caused a significant decrease in the amount of waste collected and an increase in recycling. The researchers had access to data measuring the actual amount of waste disposed of, in addition to household characteristics from a Contingent Valuation Method (CVM) survey. They found that the important determinants of the householdsÕ behavior and waste composition are the relative difficulty of composting different materials, living space, and age. WTP was measured and compared for three different municipalities. It was found that composting had a statistically significant negative correlation with waste generation: Composting reduced the flow of waste for general disposal. Age also was a significant factor in explaining the volume of waste generated; elderly people were willing to invest more time in recycling and composting. Economic incentives were also found to be important in promoting recycling. A study by Jenkins and others (1999, 2003) falls between our study of curbside recycling, as detailed below, and the Swedish study on composting and unit pricing of waste. It analyzed two options for reducing MSW: (1) unit pricing and (2) curbside recycling of five types of waste: glass bottles, plastic bottles, aluminum, newspaper, and yard waste. The data analyzed were unique in terms of the information level on US households and included information on (1) the percentage of the five types of waste recycled, (2) characteristics of the recycling program (curbside recycling versus removal on a voluntary or mandatory basis), (3) socioeconomic characteristics of households, and (4) fees collected for removal in municipalities with a unit pricing. It was found that the access to curbside recycling and the time that passed from the introduction of the program were the two main positive and statistically significant influences on the percentage of recycling in
all five categories of waste, although the scale of the influence varied from material to material. The unit pricing option influenced only the effort to separate and recycle newspapers. Household income had a positive, statistically significant influence on the recycling effort for newspaper and plastic bottles. Generally, the higher the householdsÕ income, the larger the amount of these materials recycled. From the above comparison, it is clear that the Swedish and American studies complement each other with regard to the type of incentives analyzed by them. Both of them, as well as the study presented below, were based on a combination of observed data and observed behavior. However, only the present work specifically employs data on WTP obtained from a householdÕs stated preferences (CVM) survey and actual behavior (purchasing SC/HC at the authority proposed price). Table 2 presents and compares the three studies.
Data Description
The present study is unique in the sense that the analysis is based on combining responses from a CVM survey with observed household behavior when faced with actual choice of payment for recycling services. The database included information obtained from the municipalities on the purchases of separation or composting container (depending on the municipality) and on levels of waste removal and composting in the two municipalities. A CVM telephone survey of samples of residents in the two municipalities gathered data on several aspects of the extent of the householdsÕ awareness regarding recycling, familiarity with a variety of characteristics of the relevant waste separation system, and socioeconomic background. Respondents who stated that they owned or definitely intended to buy the relevant container were queried whether they would be willing to pay a higher price, which was randomly chosen out of five (TivÕon) and
124
R. Palatnik and others
three levels (Misgav) bid levels. In each case, the maximum bid was equal to the actual cost of the container (i.e., without any subsidy). Subjects who indicated that they would not buy the container at the actual price were asked for the reason. If it was income related, they were queried whether they would be willing to pay a lower price (randomly selected from among the options). However, in both municipalities, relatively few respondents indicated that the reason for nonacquisition was economic (e.g., in TivÕon 18 out of 131 respondents). TivÕon The sample consisted of 445 households, out of which 271 households owned SCs and 43 indicated that they intend to acquire one. Thus, the data from the 445 households were used to estimate the WTP of the total relevant population in TivÕon (denoted WTPt) and data from 314 (271 + 43) households were used to estimate the WTP of those who actually bought the container (WTPb). Misgav The sample consisted of 575 homeowners, of whom 296 purchased HCs and 19 declared their intention to buy one. Data from the 575 homeowners served to estimate the appropriate WTPt and from 315 (296 + 19) to estimate WTPb.
Figure 1. Cumulative distribution function of probability to buy at price A.
Methods for Estimating WTP for Recycling Services
The Random Utility Model (RUM) (Hanemann 1984) underlies the estimation of the WTP for recycling services function. The model posits the point at which a household is indifferent between paying to participate in the new recycling program, and not paying. This point is indicated by E in Figure 1 on the function, GE(A), depicting the cumulative distribution function (CDF) of the buying decision for any given price A (of a HC or SC). The indifference point is derived from data on individuals who did purchase the SC/HC, supplemented by information from the CVM survey of households that declined to buy at the stipulated price A. Essentially, the CDF of the purchase decision represents the household demand function for SC or HC, and, by association, the demand for recycling services. Given the high proportion of nonbuyers, we supplemented the analysis by employing Reiser and ShechterÕs (1999) spike model method for incorporating zero CVM responses.
The estimated regressions for the WTP for recycling services in TivÕon and Misgav served two objectives. The first was to identify key parameters in the characterization of communities that affect the demand for waste separation services under the two recycling programs. The second objective was to calculate the median WTP in each case [i.e., the price at which households are indifferent between purchasing a SC/HC (and thereby ensure their participation in the recycling program), and not purchasing]. To carry out the first objective, a regression for theentire population of homeowners (WTPt) was estimated; therefore, it was necessary to weigh differently purchasers and nonpurchasers in the sample according to their actual representation in the entire population. For example, 564 households in TivÕon owned SCc out of 1864 private homeowners at the time of the survey. Out of 445 private homeowners participating in the survey, 314 owned SCs. In order to obtain results that reflected the actual situation, in each run of the regression we used a weighted regression that matched the responses of owners of SC/HC with their real weight in the population. The dependent variable in the estimated regression equation is BUY, the probability that WTP is greater than zero for a household with given characteristics (i.e., the probability that the household would purchase a SC/HC at a stipulated price). In the case of TivÕon, BUY stands for the probability that a SC would be purchased at a price of NIS 420. In the case of Misgav, BUY stands for the probability that a HC would be purchased at a price ranging from NIS 120 to NIS 240. To deal with the second objective, an additional regression was estimated using responses of only those households that actually bought or intended to buy the SC/HC (regression of WTPb). In this case, the
Household Demand for Waste Recycling Services
125
dependent variable BUY is the probability of purchasing the SC/HC at a higher price than the actual stated in the explanatory variable ln(bid). For TivÕon, ln(bid) was randomly selected from five bids: NIS 500, 600, 700, 800, and 920 (equal to the full cost of the container). For Misgav, ln(bid) applied to respondents who had paid NIS 120 for the HC, selected randomly from the following four bids: NIS 150, 170, 200, and 240 (the full cost of the smallest HCs). As we noted earlier, the price of HCs in Misgav jumped from NIS 120 to 240, so for those who paid the higher price, the following three bids were used: NIS 300, 400, and 480 (the full price of the largest HC). The regression equations were estimated using the forward estimation procedure. According to Reiser and ShechterÕs (1999) spike model method, estimations of the dependent variable BUY from WTPt regression and estimated parameters of independent variables from WTPb regression were employed to derive the household demand function for SCs or HCs.
Results and Discussion
Table 3 summarizes the results of the four regressions (WTPt and WTPb for TivÕon and Misgav, respectively). All information gathered in the surveys was used in the form of explanatory variables. Maximum likelihood estimation (MLE) using the forward logistic procedure estimated the vector of coefficients. The value of each of the statistically significant variables, as they appeared in each regression, is shown in Table 3. The level of confidence of all significant variables was 95% or greater. The variables that were highly insignificant (probability lower than 0.9) were omitted from the analysis. The variable ln(bid) was used only in the regression for the WTPb; it is not relevant for the WTPt regression.
household. All socioeconomic variables had a positive effect on the likelihood of purchasing a SC. The likelihood increases if the respondent or his/her spouse is self-employed. The same also holds true for retirees. A possible explanation is that the shadow value of time of pensioners is relatively low. Our findings with respect to the age variable conform to other studies (Jenkins and others 2003; Sterner and Bartelings 1999; Ayalon and others 1999; Fullerton and Kinnaman 1995). In general, our findings regarding the positive effect of economic status variables correspond to those of other works dealing with recycling effort (Jenkins and others 2003; Ayalon and others 1999; Callan and Thomas 1999). In the regression of the WTP for HCs in Misgav, none of the socioeconomic variables was found statistically significant (Table 3). All of the explanatory variables found significant in the regression were related to the awareness of the economic and environmental benefits of recycling. As the level of awareness about recycling and composting rose, the willingness of Misgav households to participate in the recycling program grew. The difference of found influences on decision of Misgav households compared to TivÕon households might be explained by the following factors: 1. The diverse nature of communities and households represented in our dataset: Socioeconomic status of households in Misgav varied much less than it did in TivÕon (most of the family heads are about 40 years old, employed, have three children, whereas TivÕonÕs population were much more diverse). 2. Differences in the price of the relevant commodities: The SC in TivÕon costs almost twice as much as the most expensive HC available in Misgav (NIS 420 compared to NIS 240 and almost four times the price of the least expensive—NIS 120). 3. Differences in the amount of time and effort required under the two systems: In TivÕon, the citizens are just required to separate the waste, whereas in Misgav, residents need to invest time and effort in periodically watering and turning the compost; only after a year could they use the resulting compost on their gardens. Therefore, the purchase of an HC might be interpreted as a demand for additional effort. 4. Differences in the ‘‘lifetime’’ of the commodity: Misgav HCs are made of plastic and, as such, have an estimated life of 5 years. In contrast, the concrete SC proposed for TivÕon has an estimated life of 15 years or more.
WTP Regressions
WTPt The econometric results reported in Table 3 indicate that two categories of variables influenced the decision of TivÕon households to purchase SCs at a price of NIS 420. One involves household socioeconomic and demographic characteristics and the other is associated with attitude, such as environmental awareness and a positive impression regarding the design of the SC. Statistically significant socioeconomic variables in the model were type of employment of the respondent, type of employment of the respondentÕs spouse, age, and the number of cars owned by the
126
R. Palatnik and others
Table 3. Regression estimates, logit model
TivÕon WTPt Sample size (N) of private homeowners Explanatory variable Constant Employee Self-employed Pensioner Age (in years) Number of cars Participate frequently in the scheme Finds the container practical Reason for recycling is environmental Reason for recycling is economic Reason for recycling is local Babies in the family Municipal information actions are clear Separate organic fraction to save costs Separate organic fraction to recycle more Separate organic fraction to get compost Spouse is employed Spouse is self-employed Partner is a pensioner Natural log of the proposed price Number of rooms in house Income above average Received information from municipality Education (in years) Number of family members Thinks neighbors participate frequently Knows the separation rule Heard about the program Score for the municipality waste treatment (1–10) Participates for local reasons Participates to get compost Participates for ecological reason Knows that organic fraction has to be clean 445 )6.498** (0.85)*** 0.057 (0.405) 0.006 (0.471) 1.2** (0.46) 0.029** (0.29) 0.417** (0.161) 3.034* (0.5) 1.191* (0.247) — — — — — — — — 0.734** (0.376) 0.277 (0.443) 0.18 (0.467) Not used — — — — — — — — — — — — — R2 = 36%
Note. Values in parentheses are standard deviations of the variables. *a = 0.05 **a = 0.01
Misgav WTPb 314 575 )3.495** (0.282) — — — — — — — 4.645 (0.84) 3.096** (0.7) 9.501** (15.8) — 2.178** (0.384) — — — — — — Not used — — — — — — — — — — — — — R2 = 72% WTPt 315 WTPb
12.756** (3.66) — — — — 0.446* (0.207) — — — 0.493** (0.244) — — — — — — 0.945** (0.311) 0.724* (0.441) 0.708* (0.372) )2.19** (0.562) — — — — — — — — — — — — — R2 = 16%
12.87** (2.313) — — — — — — — — 0.736** (0.313) — 0.859** (0.342) — 1.13* (0.445) — — — — — )2.415** (0.429) — — — — — — — — — — — — — R2 = 24%
Household Demand for Waste Recycling Services
127
WTPb The WTPb analysis refers to 314 TivÕon households that had already purchased a SC or indicated a definite intention to do so at the time of the survey. Consistent with expectations, as the suggested price of a container rises, the probability to buy one falls (Table 3). The analysis also indicates that a higher economic status of the household increases the probability of purchasing the container at the suggested price. In addition, those households show more awareness of recycling and, especially, that source separation might help decrease costs and, therefore, increase the efficacy of waste management. Willingness to pay analysis for Misgav concerned only 315 households in the sample that had already purchased a HC or had indicated the definite intention to do so. As indicated earlier, some of the HCs in Misgav were purchased for NIS 240 and some for NIS 120. This fact might have influenced the willingness of households to participate in the recycling program and the price that they were willing to pay. Therefore, a dummy variable, LowPrice, was added to the analysis. It had value 1 if the household had purchased a HC for NIS 120 and 0 otherwise. The effect of this variable on the WTP was not found to be statistically significant. As hypothesized, an inverse relationship was found between the price of the HC and the WTP for it: The higher the proposed price, the lower the WTP [negative coefficient on log(Bid)]. Additional variables influencing the actual buyersÕ WTP, as in the general population of Misgav, were awareness of recycling and its economic and environmental advantages. Positive coefficients on the variables indicate that higher levels of awareness generally, and of the economic advantages in particular, led to a higher WTP. A new variable, appearing for the first time as significant in the analysis, was the presence of a baby in the household. The effect on WTP in this case was negative, indicating that individuals taking care of a baby were less inclined to acquire a HC and shoulder its corresponding requirements in investment of time and effort. This variable can be interpreted as a proxy for free time. The Demand for Recycling Services After accounting for zero responses using the procedure suggested by Reiser and Shechter (1999), we applied the Hanneman RUM to derive the median price on the demand schedule. It was found that NIS 369 is the price at which the TivÕon householdsÕ utility from purchasing the container and participation in the new program was equal to the utility that would be re-
Figure 2. Cumulative distribution function of the probability of purchasing a container at price A: TivÕon.
ceived from not buying the container. In other words, the median price is the point at which 50% of homeowners were willing to pay a positive amount for the SC. Figure 2 depicts the cumulative distribution function of purchasing at a given price A for TivÕon residents, including standard errors. The indicated price of NIS 369 is a relatively high one; yet, it is lower than the NIS 420, which TivÕon residents had to pay, and much lower than the full cost of the container (NIS 920). As Figure 2 indicates, 38% of homeowners in TivÕon would purchase a SC for NIS 420. In fact, by October 2002, more than half of TivÕonÕs homeowners had bought a SC. It should be noted that during the time the survey was carried out, the project was based on the voluntary will of the residents to buy the SC. However, in 2002, the municipality required those who needed to replace the old containers to purchase the new SC. We note that the estimated median price is very close to the subsidized price that was actually charged. Although households that purchased the container were prepared to pay a higher price than the median, our findings indicate that the new recycling program in TivÕon could not have succeeded without the subsidy. The demand schedule for SC can be derived from the cumulative distribution function, because the independent variable is the price and the dependent variable is the probability of buying the SC. Upon inversion of the axes of the CDF of the probability of purchasing a SC at price A, found earlier, the demand function for SC for private homeowners in TivÕon was derived (see Figure 3). We can now estimate the amount of additional recycling at the TivÕon municipality after the introduction of a subsidized SC program. The demand function for SC indicates the extent of waste recycling in response to a
128
R. Palatnik and others
Figure 3. Demand curve for private homeowners in TivÕon.
change in the subsidized price. The external costs of landfilled waste in Israel had been estimated to approximate NIS 23 per ton (Goren 1997; Enosh 1996), implying that the capitalized external cost of landfilled waste produced by a typical TivÕon household over a 15-year period (the life of a SC) amounts to NIS 549. According to the estimated demand function, 25% of homeowners in TivÕon (455 out of 1864 homes) would acquire the SC at that price (NIS 549). Hence, a subsidy of NIS 180 that would lower the cost to NIS 369 (549-180) would raise participation so that 50% of homeowners would buy one (932 homes in the case of TivÕon). Because the average household in TivÕon produces 4.56 kg of solid waste daily, acquiring a SC would increase the amount of waste recycled from 20% to 70% (i.e., from 0.91 to 3.19 kg. per day per household). A NIS 180 subsidy would therefore raise the amount of recycled waste in TivÕon by about 1 ton per day or 378 tons annually over a 15-year period. In a similar vein, one can derive the WTP per kilogram waste recycled. Given the TivÕon projectÕs price of NIS 369 per SC, the corresponding annualized sum is NIS 31 (for a device life of 15 years and at 3% interest). This sum can be interpreted as the annual payment that a representative household would be willing to pay for waste collection and recycling in TivÕon. Because Israeli households, on average, produce 2 tons of solid waste annually (Biotech 1995), TivÕon households would, accordingly, be willing to pay NIS 0.015 for recycling 1 kg of waste, over and above the charge set by the municipality for waste disposal (i.e., 10% above the current flat tax). The above analysis refers to TivÕonsÕ data only; similar analysis for Misgav can be found in Palatnik (2002).
This amounts to a marginal cost of zero on waste disposal and creates an incentive to dispose of more waste than is socially efficient. Municipalities are striving to make households aware of the cost related to waste, namely the cost of treatment and related external costs. This article analyzed the experience of a local authority in carrying out a program of source separation and recycling combined with financial participation and investment of time by residents. We found the following: (1) As the price that the household is asked to pay for waste disposal services rises, the impact of the householdÕs socioeconomic characteristics on the decision to purchase a separation container increases. (2) As the amount of effort required for participation rises, the willingness to participate falls, as well as the price that the household is willing to pay. (3) As the effort required for participation rises, environmental awareness exerts a much greater impact upon the decision to participate in the recycling program. Although it seems that households tend to state that economic reasons did not influence their decision not to take part in the program, there is definitely a significant correlation between the price of the SC/HC and the willingness to purchase the device. Households are willing to pay for a weight-base waste recycling program. Thus, TivÕon residents are willing to pay a 10% premium for waste separation. However, we found that the median WTP was significantly lower than the full price of the SC/HC. It thus seems fair to conclude that without the subsidy, the recycling programs (in both TivÕon and Misgav communities) would not have succeeded to the extent that they did.
Acknowledgment
We wish to acknowledge the generous support of Life-3rd Countries project no. LIFETCY 97/IL/ 048.
Literature Cited
Artzi, A., and Aharon. 2001. Waste recycling in Israel—Findings and targets. Israel Union for Environmental Defense. Available at http://www.iued.org.il (in Hebrew). Avnimelech, Y. 1997. Land application of composted municipal wastes. Pages 551–570 in P. N. Chremisinoff (ed.), Ecological issues & environmental impact assessment. Advances in environmental control technology. Gulf Publishing, Houston, Texas. Ayalon, O. 1999. Priorities in municipal solid waste treatment. Pages 8/1–8/17 in Y. Avnimelech (ed.), National priorities in environmental quality in Israel. The Neaman Institute Publishing, Haifa, Israel (in Hebrew).
Concluding Remarks
Waste is a nonmarket commodity. Households pay a flat fee unrelated to the amount of waste they produce.
Household Demand for Waste Recycling Services
129
Ayalon, O., Y. Avnimelech, and M. Shechter. 1999. Issues in designing an effective solid waste policy: The Israeli experience. Pages 389–406 in T. Sterner (ed.), The market and the environment: The effectiveness of market based in reply to instruments for environmental reform. Edward Elgar, Cheltenham, Cheltenham, UK. Biotec. 1995. Composition of solid waste in Israel. Survey for the Ministries of the Environment and Interior (in Hebrew). Callan, S. J., and J. M. Thomas. 1999. Adopting a unit pricing system for municipal solid waste: Policy and socioeconomic determinants. Environmental and Resource Economics 14:503– 518. Central Bureau of Statistics, 2003. Real interest rate of govermental bonds, Israel. (Available at http://www.cbs.gov.il/). Enosh Consultants Ltd. 1996. Examination of external costs of waste disposal. Ministry of the Environment, the Waste Department (in Hebrew). Fullerton, D., and T. Kinnaman. 1995. Garbage, recycling, and illicit burning or dumping. Journal of Environmental Economics and Management 29(1):78–91, July. Goren, T. 1997. Assessment of external costs in waste disposal sites. Final paper for the degree of Master of Agriculture Sciences, Hebrew University of Jerusalem (in Hebrew). Hanemann, M. 1984. Welfare evaluations in contingent valuation experiments with discrete responses. American Journal of Agricultural Economics 66:332–342.
Jenkins, R. J., S. A. Martinez, K. Palmer, and M. J.Podolsky, 1999. The determinants of household recycling: A materialspecific analysis of recycling program features and unit pricing. Draft presented in 9th Annual Meeting of the European Association of Environmental and Resource Economists, Oslo. Jenkins, R. J., S. A. Martinez, K. Palmer, and M. J. Podolsky. 2003. The determinants of household recycling: A materialspecific analysis of recycling program features and unit pricing. Journal of Environmental Economics and Management 45:294–318. Kinnaman, T. C., and D. Fullerton (2000/2001). The economics of residential solid waste management. In H. Folmer and T. Tietenberg, The international year book of environmental and resource economics 2001/2002, Edward Elger, Cheltenham, UK, pp. 100–147. Palatnik, R. 2002. Assessment of demand for recycling services of household solid waste. Masters thesis, Haifa University, The Faculty of Economics (in Hebrew). Reiser, B., and M. Shechter. 1999. Incorporating zero values in the economic valuation of environmental program benefits. Environmetrics 10:87–101. Sterner, T., and H. Bartelings. 1999. Household waste management in a Swedish municipality: Determinants of disposal, recycling and composting. Environmental and Resource Economics 13:473–491.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
