Issues in Designing an Effective Solid Waste Policy: the Israeli Experience

Disposal of municipal solid waste entails severe environmental burdens especially on small, densely populated regions and countries. Israel is grappling with the task of designing an efficient (and politically acceptable!) solid-waste policy, taking into consideration externalities associated with alternative disposal options and the pervasive NIMBY (Not In My Back Yard) syndrome. This paper analyzes alternative waste management options for the country, including collection, transportation, and processing costs, landfill tipping fees as well as environmental externalities, for a representative Israeli town. Externality costs may increase disposal costs by about 10 per cent. A source separation option, namely, sorting of the waste into wet, clean compostable material and dry fractions was found to be the least costly option when landfill tipping fees are higher than 15 US$. This is due to the fact that wet organic material (mostly kitchen waste) comprises about 50 per cent of municipal solid waste in Israel. Presently, neither direct nor indirect incentives for individual households to reduce, reuse and recycle waste are built into the system; households pay a fixed, uniform fee for garbage collection with no incentives for either reducing the volume of garbage or separating it at source. In order to evaluate the feasibility of an economic incentive scheme, we carried out telephone surveys in two mediumsized communities to assess the acceptability of alternative incentive schemes, and gather data on households’ willingness to pay (WTP), using a contingent valuation approach (CV), for expanded municipal waste services which would encourage recycling activity. The results indicate that the most feasible waste management scheme in these cases (probably true for the rest of the urban areas in the country) woud entail some form of a “pay-per-bag” system, assuming “backyard” dumping can be avoided. The study suggests imposing an a uniform charge of 0.6 NIS (New Israeli Shekel, ca. 0.17 US$) per waste bag, to serve as an incentive for sorting waste at home. 1 The study was partially funded by a grant from the Chief Scientist Office at the Ministry of the Environment. It does not in any way reflect the official policy of the ministry. We wish to thank an anonymous referee for helpful comments and suggestions. 1. Background The aim of this study was to analyze alternative options for municipal solid waste (MSW) disposal in order to delineate economically-efficient and environmentally-sound MSW policies, taking into consideration political constraints involved in implementation (such as local political pressures which might block rational choices), the level of environmental awareness of the citizenry, and the physical characteristics of the system (such as the composition of waste). A rational solid waste policy should attempt to select an option which would minimize total direct and indirect (i.e., external) costs. Israel, is a densely populated country (average population density, excluding the sparsely populated, arid Negev region, is over 550 people per km, compared with 350 in the Netherlands, Japan and Belgium). The country is undergoing a gradual closure of poorly regulated or inappropriately-sited dump sites of MSW and the opening of new sites that conform to strict environmental standards. This adjustment process is aggravated by the familiar NIMBY problem. In 1993 the Israeli government mandated the closure of most existing dumps and their replacement with five national and 14 smaller regional, state-of-the-art landfills. According to the Ministry of the Environment (MoE), (Israel Environment Bulletin, 1997), recycling rates of post-consumer domestic waste leaped from 4 per cent to 10.5 per cent. These figures are due to increased costs of waste transportation to more remote, approved landfills; higher tipping fees at the new sites, opening of new sorting plants and local initiatives for collection and recycling of specific wastes and public awareness. Although we have reason to believe that these figures are somewhat overestimated, they nevertheless indicate a degree of price-responsiveness on part of consumers and other waste generators, and suggest that the introduction of market incentives might move the system towards economic efficiency. In order to promote recycling, the MoE has advocated a mixture of Command and Control (CAC) and financial instruments: regulations requiring municipalities and businesses to meet specific recycling and waste reduction goals; restriction or prohibition of import of cheaper recycled waste materials (e.g., plastics); subsidies to recycling plants; a charge for 0.25 per cent of total package cost on producers of one-trip soft drinks containers. On the other hand, due to political pressures of local municipalities, MoE has subsidized the additional costs incurred because municipal wastes need to be moved to the new officially-approved, better-managed but more remote landfills. This, of course, enables the affected towns to avoid raising municipal 2 For example, the city of Rishon Le’Zion enforced the use of cardboard compactors in medium and large commercial businesses. If compactors are used, the business gains a reduction in municipal taxes. The cardboard is collected by a private hauler for a lower price than the rest of the waste, which is collected by the municipality. The cardboard is recycled in a paper mill. A reduction of more than 15% (by weight) of MSW transferred to the landfill taxes. Clearly, this policy in turn would render landfilling the least-cost option for waste management, and acts to limit the development of new enterprises and other disposal options. Following the selection of an efficient option, an appropriate mix of economic incentives and regulations should accompany its implementation. Studies from a number of countries on the use of market incentives (mainly bag-based fees) have shown a positive response to the fee. Thus, a volume-based fee was followed by a 37 per cent decrease by volume (14 per cent by weight) in Charlottesville, Virginia (Kinnaman & Fullerton, 1994). Similar results were reported for other US municipalities (Cuthbert, 1994): a 10 per cent increase in MSW tax was followed by a 2 per cent decrease in the amount of household waste. Several studies have shown price elasticity of demand for MSW management services to be on the order of -0.12 to -0.6, depending on covariates such as income, perceived benefits from recycling, size of household, degree of urbanization and weather conditions (i.e, McFarland et al, 1972; Wertz, 1976; Jenkins, 1993). One should note, however, that the introduction of a fee-based system is more appropriate and more easily implemented in neighborhoods with single-family units; in cities of multi-family housing (and with the a populace not strongly motivated towards environmental issues), the problem of “free riding”, illegal dumping and burning could pose a serious obstacle (Kinnaman & Fullerton, 1994), creating social costs which cannot be overlooked. 2. Assumptions and Methods 2.1 Delineating Waste Management Options We have analyzed in this study what we believe to constitute four “viable” options for MSW policy: (1) landfilling of the total amount of the waste; (2) sorting mixed waste at a MRF (material recovery facility); (3) curbside collection of recyclables; and (4) wet/dry separation at home. A detailed description of these options is given in appendix 1. The choice of an optimal MSW policy should be based upon considerations of direct disposal costs as well as the indirect environmental costs (but should not overlook public sensitivities with regard to issues such as siting). In Israel, a flat, uniform waste disposal charge is included in the monthly municipal tax bill (not as a separate item); therefore, there is a complete absence of any incentive for the households to separate or reduce the waste stream. One of the intricate policy problem is to design an efficient incentive system in order to encourage households to initiate or improve source separation (if that were shown to be the optimal option). Accordingly, a number of important facts should be taken into account in deappropriate in the case of countries like Israel. For one, each person produces on average 1.14 kg of garbage per day. Population growth combined with rising incomes and living standards about half of the Israeli MSW is composed of wet food waste (including yard waste and disposable diapers), due to a diet rich in vegetables and fruits (coupled with lack of sink garbage disposal devices), compared with ca. 7 per cent in the USA and 30 per cent in Europe (Table 1). On the other hand, the “classic” recyclables (paper, glass, plastics and metals as secondary raw materials) comprise less than 30 per cent of total MSW. Thus, even under ideal conditions regarding participation rates, collection, sorting and recycling of these recyclables, this component would not be expected to divert more than 20 per cent of the waste stream – much lower than in western Europe or the US. On top of all of this, population density along the urban coastal plain (on top of the coastal aquifer which supplies about 1/5 of the country’s fresh water), where most of the population resides, is among the world’s highest. Consequently, the availability of land for landfills (except in certain areas of the southern Negev desert) is severely limited. Table 1 2.2 Data 2.2.1 Cost of Alternative MSW Options To compare direct and indirect costs across management options, we assembled and analyzed technical and cost data from recycling companies, municipalities, and other relevant sources, as well as data on potential environmental impacts and the associated external costs of the different MSW management options. We next compared costs of the different solid waste management scenarios for a “representative” town, generating 500 ton per day (tpd) of MSW. Apart from the customary collection and hauling costs, in the case of the separated recyclables collection, an additional charge of 5 per cent was added, and in the case of the wet/dry option the cost of the additional bins needed for collecting the separated waste was incorporated. Under all options, the opportunity cost of households’ time devoted to waste separation was not taken into account. We surmise that in any case the time needed for wet/dry separation is less than that needed for sorting recyclables. 3 To be soon prohibited by law in order to conserve water and prevent over-loading of sewage treatment plants. 4 Even in this region a good portion of it could not accommodate landfills because of potential groundwater pollution or military needs. 5 Economic evaluations of environmental impacts were obtained from a research conducted by EMC Ltd. for the Israeli Ministry of the Environment in 1996. Evaluations of external costs from air pollution (caused by transportation and landfill gas) and leachates from landfills were obtained from DoE report (1993). 6 Some of the data were obtained from private firms engaged in recycling as “black boxes” i.e., “bottom-line” 2.2.2 Survey Data Applying economic incentive schemes in the context of MSW management must take cognizance of the aforementioned fact that household have not faced any explicit price for waste collection and processing. This can, therefore, be viewed and studied as a new service, in a manner not altogether different from that, which is applied to new products. However, in this case we deal with a non-market service, because there customarily do not exist markets for municipal waste collection services. Household’s demand schedule for these services must be inferred by means of a CV-type survey. Alternatively, we could have analyzed the process of source waste-separation in the framework of a household production function, combining time and labor inputs (as well as some capital costs in the form of additional containers and bags) to produce sorted waste. Here, the price of the service could be in the form, say, of a penalty if the waste were not thus separated. We opted for the first alternative. The data were obtained from a survey conducted at the end of a one year source separation pilot project. In a project lasting through 1994, 6000 families in the cities of Ra’anana and K’far Saba in central Israel were asked to separate the waste into wet and dry streams. To evaluate the project’s performance, a follow-up telephone survey of heads of households (husband or wife, randomly selected) of 600 of these families was carried out when the project terminated. Altogether, the questionnaire contained 28 questions, dealing with the household familiarity with and attitude towards waste separation and recycling, participation rates, demographics, and a contingent valuation (CV) scenario. Table 2 gives the socioeconomic and demographic characteristics of the 600 household sample. Table 2 In the CV part, respondents were queried on their willingness to pay (WTP) for municipal services which would enable source (i.e., at home) waste separation, in order to facilitate the recycling component in the various MSW options. The CV scenario is given in Appendix 2. 3. Results and Discussion 3.1. Comparing Costs of Alternative Waste Management Options The results of the cost analyses of the different options for a “representative” town are presented in Figure 1 and Table B (in Appendix 1). The least-cost solution involves the disposing of waste to a nearby landfill, if one is available. This, as elaborated above, is fast becoming an infeasible solution. The externality costs of landfilling due to leachates and landfill gas are in the range of 5.75 US$/ton (DoE report, 1993). When external costs attributable to transportation (congestion, road accidents, noise and pollution) are added, the total external costs could increase total disposal costs by more than 10 per cent (Table C in Appendix 1). According to an analysis conducted by the US EPA, (Glebs, 1989), the expected costs of a well operated sanitary landfill are about 20 US$ per ton. Therefore, if tipping fees are higher than 20 US$ per ton, external costs mainly road-related could raise total disposal costs by about 6 per cent, NIMBY “costs’ notwithstanding. Figure 1 As noted above, recycling of “conventional” recyclables (paper, plastics and glass) is often viewed by decision makers. There is no doubt that using recycled materials, compared with incineration or landfilling, conserves natural resources and sometimes energy (e.g. aluminum recycling). Therefore, recycling has a high rate of public acceptance (Chilton, 1993). However, in Israel, conventional recycling methods curbside collection of recyclables does not reduce the need for landfilling, since, due to its composition, this option would still leave 80 per cent of the waste to be landfilled. Regarding the other solutions, it can be seen that as long as tipping fees (TF) are relatively low, less than 20 NIS/ton (ca. 5 US$), there is a cost advantage to sorting mixed waste at a material recovery facility (MRF). This option includes collection of unsorted, wet, sticky and fermenting waste, the waste is being mechanically sieved for the organic fraction and manually sorted for recyclables. These characteristics of the waste impose several environmental and economic disadvantages: the compost might be contaminated with heavy metals and broken glass (which are sieved as well); the ability to sort the waste is limited (due to the wetness and stickiness of the waste); and the quality of recyclables is reduced (with a corresponding decline of marketability and price). When TF are above 55 NIS (ca. 15 US$), the cost-effective solution is the wet/dry source separation, which would also promise to keep the highest ratio of waste away from landfills. In this option, 40 to 50 per cent of the waste are directly composted and the clean compost is applied to the soil, increasing crop yields and improving soil structure. A further 10 to 20 per cent of the waste is recycled, and about 10 per cent may be used as Refuse Derived Fuel (RDF). Only 20 to 35 per cent of the waste would be delivered to a landfill. 3.2. Factors Affecting WTP for Waste Separation About 40 per cent of the surveyed sample population refused to pay any sum for the introduction of a recycling option, a slightly higher percentage than that usually encountered in collection is customarily viewed as a basic and essential public service, paid for by our municipal taxes. Three separate analyses of WTP responses were carried out: WTP=0 vs. WTP>0 responses, WTP>0 responses, and a combined analyses of these two response groups. First, the customary logistic regression analysis of zero vs. non-zero responses (Table 3) indicates that the probability of a zero response is higher if the person is a salaried employee, the household owns only 1 car (an indicator of relatively lower income level in the case of these two rather affluent suburban towns), is older, and resides in a single-unit home. Except for the latter, the explanatory variables may be interpreted (in the present framework) as proxies for less affluence. Not surprising, the higher the bid, the higher is the probability of getting a WTP of zero, noting that the addition of 20 NIS to the monthly municipal bill amounts to almost 50 per cent increase in the aforementioned flat charge for waste handling presently incorporated in it. Table 3 Next, the positive WTP responses were analyzed with a semi-log regression equation, which performs better when money is the dependent variable. The analysis indicates that WTP is higher if the respondent is younger, the spouse is self-employed, lives in a single-unit housing, resides in the town of K’far Saba, and randomly assigned a higher bid level (the familiar “anchoring” effect). Thus, again, the regression results suggest that a higher WTP to help introduce and operate the recycling option is associated with a higher level of affluence. Finally, given the high percentage of zero responses, we applied a mixture “spike” model (Reiser and Shechter, 1997), which has been designed to analyze the combined WTP response distributions. The mixture model assumes that the population of interest could be considered to be composed of two sub-populations. One sub-population is not willing to pay for the good or service in question, while the other sub-population is willing to pay and has a continuous WTP distribution. That is, the model does not compel the investigator to assume that both groups are drawn from the same population in terms of preferences and other relevant attributes, when combining the two groups, as is the case with standard analyses of CV data. In applications of this kind, one usually is made aware of the fact that due to the strong covariate effects, examining the overall mean (or median) of the sample may be misleading. Consequently, rather than the customary overall sample means, we present in Table 4 means for a number of particular sub-groups, representing different combinations of socioeconomic and demographic covariates. For example, given a bid of 10 NIS, the probability of WTP=0 (last column in Table 4) is higher if the household (in the case of a salaried employee) owns only 1 car instead of 2 or (for self-employed) if the head of the household is older. This of course would be reflected in the mean and median of the corresponding sub-groups. Thus, in analyzing the accurate representation of the relevant population, and, in our case, the range of the community’s responses to the introduction of an MSW incentive scheme. We believe that this kind of information is important in designing a workable incentive scheme, such as a per-bag fee, since it might indicate, for instance, neighborhoods where there might be tendencies towards less cooperation. In such cases, more events of illegal dumping or other forms of non-cooperation would be more likely to occur and, therefore, should be taken into consideration prior to the launch of the incentive scheme. Table 4 3.3 Economic Incentives as a Reflection of WTP To illustrate how one might go about converting expressed WTP into a (hopefully...) workable incentive scheme, we computed mean WTP of each community using the mixed model (i.e., averaging for each place of residence over all responses) and used it as a basis for calculating a per-bag fee the economic incentive proposed to promote separation at source. That fee is intended to “internalize” disposal externalities, moving the system to operate closer to the optimal disposal scheme. In the following table we present mean monthly WTP with and without the zero’s for the two towns: K’far Saba Ra’anana WTP>0 14.65 NIS (176 NIS)* 13.62 NIS (163 NIS) WTP≥0 7.17 NIS (86 NIS) 7.69 NIS (92 NIS) *(annual payment is given in parenthesis). Given that an average family generates approximately 2 tons of solid waste per year, the lowest WTP value of 7 NIS per month implies that each family would be willing to pay an extra fee (i.e., marginal WTP) of about 40 NIS per ton of MSW. The relevance of this specific value in the present case will be shown momentarily. Figure 2 shows the three ‘runner-up’ options: “Close landfill”, “Sorting mixed waste”, and “Wet-dry source separation”, but with a “cosmetic uplifting” of the first alternative via a vertical upwards shift of its cost schedule by 40 NIS. This shift causes the second alternative, namely, “Sorting mixed waste”, to become the cost-effective solution at a TF of 7 to 8 NIS (the point where the two cost schedules intersect). One could interpret this shift as representing the shadow price, or the implicit marginal supply cost of in-situ waste disposal services, following the aforementioned national policy of closing “close” municipal landfills. Figure 2 However, we argue that the next best option, the separation at the source to wet and dry fractions, merits a closer examination. It has already been noted that if the TF were to rise to about 55 NIS per ton, it will become the cost-effective solution. Now, in Appendix 1 it is argued that, compared with the “wet-dry” solution, the “sorting mixed waste” option produces an inferior quality compost and recyclables. But as no price tag was attached to either the production of compost (of any quality) or to the quality component of the recyclables stream, one may surmise that if these aspects were taken into account (causing a corresponding downwards vertical shift of the relevant curve in Figure 2), the “Wet-dry” option could turn out to be the most cost-effective solution, intersecting the “Close landfill” schedule at a point near that where the “Sorting mixed waste” curve now intersects it. Consider now the marginal WTP of 40 NIS per ton calculated above: Given the closure of “close” landfills, this WTP turns out to equal the marginal supply cost of the most costeffective option(s). If municipalities were willing to pass over this added supply cost to the citizenry, through some form of a decentralized incentive payment, the system could reach the desired optimal solution in a decentralized fashion. In that case, the numbers above would translate into a fee tag of 0.6 NIS per bag in a home-sorting waste scheme, accompanied, of course, by an appropriate monitoring and enforcement scheme. [If the authorities decide that bag price should also reflect current collection and disposal costs, its price would rise from 3.90 NIS to 4.50 NIS] This approach could substitute for (or at least supplement) Command and Control-type regulations, or subsidies to municipalities (to cover the added cost of centralized sorting facilities) financed by the general tax-paying public. 4. Conclusions MSW policy should be adjusted to the unique characteristics of any country. MSW in Israel contains more organic kitchen waste and less paper and cardboard than the European and US waste; The populace, in general, is less motivated to voluntary “environment-friendly” behavior; The typical residential pattern of more than 70 per cent of the dwelling units is one of small apartments (including small storage space in kitchens) in multi-apartment buildings. 9 40 NIS divided by 70, the number of waste bags needed for 1 ton of waste. 10 Average collection and disposal fees are about 270 NIS per ton; dividing by 70 (the number of waste bags per ton) = 3.90 NIS/bag. Our results show that sorting mixed waste and the wet/dry options, involving recycling of both the dry and organic fractions, are the most cost-effective as well as social welfare optimizing solutions. Our findings call for the establishment of a ‘pay-per-bag’ incentive scheme, namely, a tax of 0.6 NIS per bag, which will induce households to internalize the optimal solution. This price would not constitute too heavy a burden on households and will not induce non-cooperation or illegal dumping. Furthermore, in arriving at a suitable solution to MSW policy, we should avoid adopting “politically correct” solutions. In both, popular and political circles, recycling rather than the total management of the solid waste stream is often considered a “politically-correct” national priority. Green lobbies have often argued that recycling is the solution to the MSW problem, promoting “nobler” goals such as “saving raw materials, energy, and land resources”, often sadly regardless of the associated cost tab or other indirect environmental costs, such as the pollution generated by additional collection vehicles, or water and energy needed to wash reusable glass bottles. Countries like Israel should learn from other nations’ mistakes, and try to avoid repeating them. This is particularly true in regard to the hasty imposition of CAC rules and regulations (e.g.. the packaging ordinance in Germany) prohibiting landfilling, which have proved to be economically inefficient (Anonymous1, 1993, Anonymous 2, 1993). 11 To quote from a recent newsletter of “Man, Nature, and Law”, the Israeli Union for Environmental Defense: “Recycling should be viewed as a national objective (underlined in the original text) without constraining its execution by economic considerations, or at least by narrow economic tests. Recycling is an expression of important values, such as preventing squandering of resources, saving of natural and energy resources, and treating