The Economic Benefits of Reducing the Environmental Effects of Landfills: Heterogeneous Distance Decay Effects


This paper estimates the economic benefits of different alternative policies considered for reducing the external effects caused by the utilization of landfills in solid waste management. The preferences of the local population in the surrounding areas of a landfill are evaluated utilizing a discrete choice experiment in which subjects are presented with alternative policy decisions that involve reducing the material processed through the landfill. Various models were employed in order to capture heterogeneous preferences, resulting in a mixture of normals modelling approach (MN-MNL) outperforming other alternative models of heterogeneity. The results show that the policy of moving the landfill away from the population does not provide the most benefits when compared to a policy of increasing recycling in the household. The economic benefits of the waste management policies are heterogeneous across the population surrounding the landfill and so the distance decay functions. Thus, the economic benefits for most waste management policies can increase or decrease the further away the subject lives from the landfill, depending on the preferences of his/her segment and the type of policy employed.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6


  1. 1.

    There is a relatively large number of applications of CV focusing on the economic assessment of WM and recycling policies in developed countries (e.g. Tiller et al. 1997; Aadland and Caplan 1999, 2006; Lake et al. 1996; Jamelske and Kipperberg 2006; Gallagher et al. 2008; Sterner and Bartelings 1999; Huhtala 1999, 2010) and in developing countries (e.g. Altaf and DeShazo 1996; Bluffstone and DeShazo 2003; Huang and Ho 2005; Fonta et al. 2007; Afroz et al. 2009; Ichoku et al. 2009; Afroz and Masud 2011).

  2. 2.

    See for instance Caplan et al. (2002) in Odgen (US), Jin et al. (2006) in Macao, Karousakis and Birol (2008) in London, Pek and Jamal (2011) in Malaysia, Gillespie and Bennett (2012) in Brisbane (Australia), and Kipperberg and Larson (2012) in Seattle (US).

  3. 3.

    The question of the location of WM facilities has been also assessed in economic terms in the studies of Robert et al. (1991) and Groothuis and Miller (1994) utilizing the CV method.

  4. 4.

    There are several recent works aimed at reviewing alternative econometric specifications to account for heterogeneity in choice models (Keane and Wasi 2013; Balcombe et al. 2009 among others). Since the aim of the present work is not to compare all the specifications in detail and just test differences in practice, the framework for this section mainly comes from Keane and Wasi (2013), for all model details.

  5. 5.

    Represented by a strictly increasing, continuous, and strictly quasi-concave utility function.

  6. 6.

    Note that since \(U_{\textit{ijt}}\) is a indirect conditional utility function of all the determinants of individuals’ choices that do not differ among alternatives at each choice occasion (e.g. individuals’ socio-demographics) then it can not be included in (1) unless they are interacted with a constant term.

  7. 7.

    It is worth to note that this specification is equivalent to assuming that \(\eta _{i}\) follows a multivariate normal distribution with zero mean and a diagonal covariance matrix \(\underline{\Sigma _{\beta }}\).

  8. 8.

    Since this integral does not have a closed form solution, McFadden and Train (2000) suggested to estimate the model using simulated maximum likelihood methods or Markov Chain Monte Carlo simulation techniques.

  9. 9.

    Carson and Louviere (2011) argue that DCCV can be seen as a simple and reduced version of conventional DCE in which each individual receives only one choice task, that involves a choice between the status quo and a single alternative combination of attributes and levels that normally implies an improvement in some attribute (i.e. environmental quality) and a monetary payment in order to move from the status quo option. In terms of the model outlined here, a DCCV follows from DCE by considering \(\hbox {t}=1\) and \(\hbox {j}=2\).

  10. 10.

    By incorporating a multinomial probit model to define state probabilities, well known identification issues from this model also need to be addressed. Geweke and Keane (2001) show that the scale problem can be overcome by employing a fixed variance matrix, and that the identification question of the posterior distribution would be resolved by using any proper distribution for \(\Gamma \). See Geweke and Keane (2001) for specific details about the specification and the inference of the functions of interest, and Geweke (2007) for an in-depth discussion about identification and interpretation issues in the mixture of normals and SMR models, and how to overcome them. In a entirely frequentist setting, also Yao (2015) provides alternative solutions to deal with this issue.

  11. 11.

    Following Frühwirth-Schnatter (2001) and León and Araña (2012) recommendations, we examine marginal density plots and two dimensional scatter plots of posterior simulations of group specific parameters to choose restrictions that guarantee unique labelling, and test whether or not label switching may be an issue for our specific application. Also, different starting values were considered, and final results were insensitive to them. Based on this analysis and the inspection of graphs for all posterior values, the chain of parameter draws shows to hold good convergence properties and no label switching issues raised. Following a reviewer recommendation we include some of these graphs in “Appendix 2”.

  12. 12.

    Official population statistics can be found at Regional Statistical Office (ISTAC),

  13. 13.

    The utilization of quotas for gender and age was intended to guarantee a representative sample in these two parameters. The utilization of these parameters is a common procedure in pooling studies (Araña and León 2013).

  14. 14.

    The professional interviewers were qualified and trained holding academic degrees and under contract with the survey company hired for this study, with between 5 and 11 years of experience in field studies. They were a team of five interviewers and two supervisors, with ages ranging from 30 to 47 years old.

  15. 15.

    The bid vector design was determined following Cooper (1993) procedures for mean square error minimization. This vector was also checked for policy relevance by the panel of experts. The price vector was also explored in focus group discussions.

  16. 16.

    The questionnaire and all the materials utilized in survey research are available from the authors upon request.

  17. 17.

    A reviewer to this paper has raised the issue of the potential credibility of this policy. Based on careful pre-testing and work with focus groups it was found that subjects clearly understood the feasibility of this policy, since there are other areas in the island which are not located near populations and which can be planned for landfill development. The fact that this specific area was not specified was not an issue, since subjects were only interested in the policy of getting rid of the landfill near their homes. Thus, based on our focus group interviews we decided that there was no need to specify any alternative location in the island. On the other hand, policy makers had not already defined this specific area.

  18. 18.

    The remainder of personal income and potential substitutes follows recommendations by the NOAA panel protocol on the use of the stated preference methods such as contingent valuation and DCE for measuring environmental damages (Arrow et al. 1993).

  19. 19.

    A nice review of the advantages of this methodology in our context can be found in Balcombe et al. (2009) .

  20. 20.

    Several approaches have been proposed in the econometric literature to estimate the impact of covariates on segment probabilities based on GS draws. These approaches may require either the inclusion of an extra stage within the GS algorithm, or a post-simulation treatment of the GS draws (see Ghosh et al. 2011 for a recent review). Here we employed a post-simulation approach. We use the GS draws to estimate the posterior membership probabilities, and to assign individuals to the segment with the highest probability. Then, a standard Probit regression was employed to estimate the impact of socio-demographics on the probability of belonging to each segment.

  21. 21.

    Highest Posterior Density Intervals (HPD) in Table 5 were calculated by using the 2.5th and 97.5th percentile of the posterior distribution of the WTP. A reviewer pointed out that the Poe et al. (2005) approximations of using a convolution approach is useful when target variables are non-normal. However, since mixture of normals can approximate any functional form (Fergunson 1973) the use of HPD for the posterior distribution allows researchers to approximate a non-parametric comparison of the full WTP distribution.

    The methodology is based on intensively drawing from a multivariate normal distribution, using the estimated parameter coefficients as means and the estimated covariance matrix of the coefficients as covariance. As noticed by Efron and Tibshirani (1993), a non-parametric bootstrapping distribution of the WTP is obtained by using the simulated values of the parameters in each draw.


  1. Aadland DM, Caplan AJ (1999) Household valuation of curbside recycling. J Environ Plan Manag 42(6):781–799

    Article  Google Scholar 

  2. Aadland DM, Caplan AJ (2006) Curbside recycling: Waste resource or waste of resources? J Policy Anal Manag 25(4):855–874

    Article  Google Scholar 

  3. Adamowicz W, Louviere J, Williams M (1994) Combining stated and revealed preference methods for valuing environmental amenities. J Environ Econ Manag 26:217–292

    Article  Google Scholar 

  4. Afroz R, Hanaki K, Hasegawa-Kurisu K (2009) Willingness to pay for waste management improvement in Dhaka city, Bangladesh. J Environ Manag 90(1):492–503

    Article  Google Scholar 

  5. Afroz R, Masud MM (2011) Using a contingent valuation approach for improved solid waste management facility: evidence from Kuala Lumpur, Malaysia. Waste Manag 31(4):800–808

    Article  Google Scholar 

  6. Akaike H (1987) Factor analysis and AIC. Psychometrika 52:317–332

    Article  Google Scholar 

  7. Altaf MA, DeShazo JR (1996) Household demand for improved solid waste management: a case study of Gujranwala, Pakistan. World Dev 25(5):857–868

    Article  Google Scholar 

  8. Araña JE, León CJ (2005) Flexible mixture distribution modeling of dichotomous choice contingent valuation with heterogenity. J Environ Econ Manag 50(1):170–188

    Article  Google Scholar 

  9. Araña JE, León CJ (2006) Modelling unobserved heterogeneity in contingent valuation of health risks. Appl Econ 38(19):2315–2325

  10. Araña JE, León CJ (2013) Can defaults save the climate? Evidence from a field experiment on carbon offsetting programs. Environ Resour Econ 54(4):613–626

  11. Arrow K, Solow R, Portney PR, Leamer EE, Radner R, Schuman H (1993) Report of the NOAA panel on contingent valuation.

  12. Azapagic A (2007) Energy from municipal solid waste: large-scale incineration or small-scale pyrolysis. Environ Eng Manag J 7(5):337–346

    Google Scholar 

  13. Balcombe K, Chalak A, Fraser I (2009) Model selection for the mixed logit with Bayesian estimation. J Environ Econ Manag 57(2):226–237

  14. Bateman I, Langford I (1997) Non-users WTP for a national park: an application of the contingent valuation method. Reg Stud 31(6):571–582

    Article  Google Scholar 

  15. Bateman IJ, Day BH, Georgiou S, Lake I (2006) The aggregation of environmental benefit values: welfare measures, distance decay and total WTP. Ecol Econ 60(2):450–460

  16. Ben Akiva M, Swait J (1986) The Akaike likelihood index. Transp Sci 20:133–136

    Article  Google Scholar 

  17. Ben-Avika M, McFadden D, Abe M, Bockenholt U, Bolduc D, Gopinath D, Morikawa T, Ramaswamy V, Rao V, Revelt D, Steinberg D (1997) Modeling methods for discrete choice analysis. Mark Lett 8(3):273–286

    Article  Google Scholar 

  18. Bluffstone R, DeShazo JR (2003) Upgrading municipal environmental services to European Union levels: a case study of household willingness to pay in Lithuania. Environ Dev Econ 8:637–654

    Article  Google Scholar 

  19. Boxall PC, Adamowicz WL, Swait M, Williams M, Louviere J (1996) A comparison of stated preference methods for environmental valuation. Ecol Econ 18(3):243–253

    Article  Google Scholar 

  20. Bozdogan H (1987) Model selection and Akaike’s information criterion (AIC): the general theory and its analytical extensions. Psychometrika 52:345–370

    Article  Google Scholar 

  21. Bozdogan H (1994) Proceedings of the first US/Japan conference on the frontiers of Statistical modeling: an informational approach. Kluwer Academic Publishers, Dordrecht

    Book  Google Scholar 

  22. Brey R, Walker JL (2011) Latent temporal preferences: an application to airline travel. Transp Res Part A Policy Pract 45(9):880–895

  23. Brisson I, Pearce D (1998) Literature survey of hedonic property pricing studies of landfill disamenities cited in European Commission (2000). In: A study on the economic valuation of externalities from landfill disposal and incineration of waste management. Final Appendix Report conducted by COWI on behalf of the European Commission, October, Brussels

  24. Burda M, Harding M, Hausman J (2008) A Bayesian mixed logit-probit model for multinomial choice. J Econom 147(2):232–246

    Article  Google Scholar 

  25. Caplan T, Grijalva C, Jakus PM (2002) Waste not or want not? A contingent ranking analysis of curbside waste disposal options. Ecol Econ 43(2002):185–197

    Article  Google Scholar 

  26. Caplan A, Grijalva T, Jackson-Smith D (2007) Using choice question formats to determine compensable values: the case of a landfill-siting process. Ecol econ 60(4):834–846

  27. Carson RT, Louviere JJ (2011) A common nomenclature for stated preference elicitation approaches. Environ Resour Econ 49(4):539–559

    Article  Google Scholar 

  28. Concu GB (2007) Investigating distance effects on environmental values: a choice modelling approach\(^\ast \). Aust J Agric Resour Econ 51(2):175–194

  29. Cooper JC (1993) Optimal bid selection for dichotomous choice contingent valuation surveys. J Environ Econ Manag 24(1):25–40

    Article  Google Scholar 

  30. Czajkowski M, Kądziela T, Hanley N (2014) We want to sort! Assessing households’ preferences for sorting waste. Resour Energy Econ 36(1):290–306

    Article  Google Scholar 

  31. del Saz Salazar S, García Menéndez L (2007) Estimating the non-market benefits of an urban park: does proximity matter?. Land Use Policy 24(1):296–305

  32. Efron B, Tibshirani RJ (1993) An introduction to the bootstrap. Monographs on Statistics and Applied Probability (EUA), (57)

  33. Fiebig D, Keane M, Louviere J, Wasi N (2010) The generalized multinomial logit model: accounting for scale and coefficient heterogeneity. Mark Sci 29(3):393–421

    Article  Google Scholar 

  34. Fonta WM, Ichoku HE, Ogujiuba KK, Chukwu JO (2007) Using a contingent valuation approach for improved solid waste management facility: evidence from Enugu state. Nigeria J Afr Econ 17(2):277–304

    Article  Google Scholar 

  35. Frühwirth-Schnatter S (2001) Markov Chain Monte Carlo estimation of classical and dynamic switching and mixture models. J Am Stat Assoc 96:194–209

    Article  Google Scholar 

  36. Gallagher L, Ferreira S, Convery F (2008) Host community attitudes towards solid waste landfill infrastructure: comprehension before compensation. J Environ Plan Manag 51(2):233–257

    Article  Google Scholar 

  37. Garrod G, Willis K (1998) Estimating lost amenity due to landfill waste disposal. Resour Conserv Recycl 22(1–2):83–95

    Article  Google Scholar 

  38. Gelfand AE, Ghosh SK (1998) Model choice: a minimum posterior predictive loss approach. Biometrika 85(1):1–11

    Article  Google Scholar 

  39. Georgiou S, Bateman I, Cole M, Hadley D (2000) Contingent ranking and valuation of water quality improvements. CSERGE discussion paper 2000–18, University of East Anglia

  40. Geweke J (2007) Interpretation and inference in mixture models: simple MCMC works. Comput Stat Data Anal 51(7):3529–3550

    Article  Google Scholar 

  41. Geweke J, Keane M (2001) Computationally intensive methods for integration in econometrics. Handb Econom 5:3463–3568

    Article  Google Scholar 

  42. Gillespie R, Bennett J (2012) Willingness to pay for kerbside recycling in Brisbane, Australia. J Environ Plan Manag 0(0):1–16

    Google Scholar 

  43. Ghosh J, Herring AH, Siega-Riz AM (2011) Bayesian variable selection for latent class models. Biometrics 67(3):917–925

    Article  Google Scholar 

  44. Groothuis PA, Miller G (1994) Locating hazardous waste facilities: the influence of NIMBY beliefs. Am J Econ Sociol 53(3):335–346

    Article  Google Scholar 

  45. Guntermann KL (1995) Sanitary landfills, stigma and industrial land values. J Real Estate Res 10(5):531–542

    Google Scholar 

  46. Hanley N, Wright RE, Adamowicz V (1998) Using choice experiments to value the environment: design issues, current experience and future prospects. Environ Resour Econ 11:1–16

    Article  Google Scholar 

  47. Hanley N, Schläpfer F, Spurgeon J (2003) Aggregating the benefits of environmental improvements: distance-decay functions for use and non-use values. J Environ Manag 68:297–304

    Article  Google Scholar 

  48. Hirshfeld S, Vesilind PA, Pas EI (1992) Assessing the true costs of landfills. Waste Manag Res 10(6):471–484

    Article  Google Scholar 

  49. Huang C-J, Ho Y-H (2005) Willingness to pay for waste clearance and disposal: results of the Taichung city study. Bus Rev Cambr 4(2):136–141

    Google Scholar 

  50. Huber J, Zwerina K (1996) The importance of utility balance in efficient choice designs. J Mark Res 33:307–317

    Article  Google Scholar 

  51. Huhtala A (1999) How much do money, inconvenience and pollution matter? Analysing households’ demand for large-scale recycling and incineration. J Environ Manag 55(1999):27–38

    Article  Google Scholar 

  52. Huhtala A (2010) Income effects and the inconvenience of private provision of public goods for bads: the case of recycling in Finland. Ecol Econ 69:1675–1681

    Article  Google Scholar 

  53. Ichoku HE, Fonta WM, Kedir A (2009) Measuring individuals’ valuation distributions using a stochastic payment card approach: application to solid waste management in Nigeria. Environ Dev Econ 11(2009):509–521

    Google Scholar 

  54. IIIEE (2013) Catalysing change: five stories of advancing sustainable solutions. The International Institute for Industrial Environmental Economics (IIIEE), Lund University

  55. INE (2010) Estadísticas sobre medio ambiente. Estadísticas de residuos. Instituto Nacional de Estadística, Madrid, España

    Google Scholar 

  56. Jamelske E, Kipperberg G (2006) A contingent valuation study and benefit/cost analysis of the switch to automated collection of solid waste with single stream recycling in Madison, Wisconsin. Public Works Manag Policy 11(2):89–103

    Article  Google Scholar 

  57. Jin J, Wand Z, Ran S (2006) Comparison of contingent valuation and choice experiment in solid waste management programs in Macao. Ecol Econ 57:430–441

    Article  Google Scholar 

  58. Kamakura WA, Russell GJ (1989) A probabilistic choice model for market segmentation and elasticity structure. J Mark Res 26(4):379–390

    Article  Google Scholar 

  59. Karousakis K, Birol E (2008) Investigating household preferences for kerbside recycling services in London: a choice experiment approach. J Environ Manag 88:1099–1108

    Article  Google Scholar 

  60. Keane M, Wasi N (2013) Comparing alternative models of heterogeneity in consumer choice behavior. J Appl Econom 28(6):1018–1045

  61. Kipperberg G, Larson DM (2012) Heterogeneous preferences for community recycling programs. Environ Resour Econ 53(4):577–604

    Article  Google Scholar 

  62. Kohlhase JE (1991) The impact of toxic waste sites on housing values. J Urb Econ 30:1–26

    Article  Google Scholar 

  63. Lake I, Bateman I, Parfitt J (1996) Assessing a Kerbside recycling scheme: a quantitative and willingness to pay case study. J Environ Manag 46:239–254

    Article  Google Scholar 

  64. León CJ, Araña JE (2012) The dynamics of preference elicitation after an environmental disaster: stability and emotional load. Land Econ 88(2):362–381

  65. Luce RD, Tukey JW (1964) Simultaneous conjoint measurement: a new type of fundamental measurement. J Math Psychol 1:1–27

    Article  Google Scholar 

  66. Mazzanti M, Zoboli R (2009) Municipal waste Kuzunets curves: evidence on socio-economic drivers and policy effectiveness in the EU. Environ Resour Econ 44:203–230

    Article  Google Scholar 

  67. McFadden D (1974) Conditional logit analysis of qualitative choice behaviour. In: Zarembka P (ed) Frontiers in econometrics. Academic Press, New York

    Google Scholar 

  68. McFadden D, Train K (2000) Mixed MNL models for discrete responses. J Appl Econom 15(5):447–470

    Article  Google Scholar 

  69. Nelson AC, Genereux J, Genereux M (1992) Price effects of landfills on house values. Land Econ 68(4):359–365

    Article  Google Scholar 

  70. Pate J, Loomis J (1997) The effect of distance on willingness to pay values: a case study of wetlands and salmon in California. Ecol Econ 20(3):199–207

  71. Pek C-K, Jamal O (2011) A choice experiment analysis for solid waste disposal option: a case study in Malaysia. J Environ Manag 92:2993–3001

    Article  Google Scholar 

  72. Poe GL, Giraud KL, Loomis JB (2005) Computational methods for measuring the difference of empirical distributions. Am J Agric Econ 87(2):353–365

    Article  Google Scholar 

  73. Roberts RK, Douglas PV, Park WM (1991) Estimating external costs of municipal landfill siting through contingent valuation analysis: a case study. South J Agric Econ 23:155–165

    Google Scholar 

  74. Rolfe J, Windle J (2012) Distance decay functions for iconic assets: assessing national values to protect the health of the Great Barrier Reef in Australia. Environ Resour Econ 53(3):347–365

  75. Sasao T (2004) An estimation of the social costs of landfill siting using a choice experiment. Waste Manag 24(8):753–762

    Article  Google Scholar 

  76. Smith VK, Desvousges W (1986) The value of avoiding a LULU: Hazardous waste disposal sites. Rev Econ Stat 68:293–299

    Article  Google Scholar 

  77. Schwartz G (1978) Estimating the dimension of a model. Ann Stat 6:461–464

    Article  Google Scholar 

  78. Sterner T, Bartelings H (1999) Household waste management in a swedish municipality: determinants of waste disposal, recycling and composting. Environ Resour Econ 13(1999):473–491

    Google Scholar 

  79. Sutherland RJ, Walsh RG (1985) Effect of distance on the preservation value of water quality. Land Econ 281–291

  80. Tiller KJ, Jakus PM, Park WM (1997) Household willingness to pay for dropoff recycling. J Agric Resour Econ 22(2):310–320

    Google Scholar 

  81. Train K (2003) Discrete choice methods with simulation. Cambridge University Press, Cambridge

    Book  Google Scholar 

  82. Vuong QH (1989) Likelihood ratio tests for model selection and non-nested hypothesis. Econometrica 57(2):307–333

    Article  Google Scholar 

  83. Yao W (2015) Label switching and its simple solutions for frequentist mixture models. J Stat Comput Simul 85(5):1000–1012

Download references


Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The author(s) received financial support for the research from the Spanish Government (Ministerio de Economía y Competitividad) under the research program ECO2011-30365.

Author information



Corresponding author

Correspondence to Javier de León.


Appendix 1: Scenario Description and Choice Question

In the city of Las Palmas and in the island of Gran Canaria in general, there is a problem with the management of urban solid waste. Currently, the wastes of municipalities of the Central and Northern regions of the island of Gran Canaria are managed at the landfill named Salto del Negro, in the city of Las Palmas, as you can see in this map here (show and explain the map to the respondent).

All waste is accumulated and buried at this landfill. The problem is that the landfill system has a limited capacity, and when no more materials can be buried another place will need to be found on the island, which, at the same time has also a limitation of space. In addition, there can be other risks to health and to the quality of life of the persons living nearby the landfill, because of the gases, bad odor and the animals that can proliferate in these areas such as rodents or insects.

The current system of waste management in Las Palmas de Gran Canaria recycles about 10 % of all the waste generated in all households in the city, without the utilization of specialized machinery, facility or equipment in the landfill. The higher the amount of waste collected for recycling the lesser would be the amount of risks to health and the quality of life of the persons living nearby, and the higher would be the lifespan of the current landfill.

Please let me now explain to you some of the policy measures that could work to solve the problem described above:

  1. A.

    (Show card A). Firstly, it can be implemented a system of separation facilities based on machinery in the landfill which will be able to collect all materials that can be reused, such as organic material, plastic, glass and metals, as you can see illustrated in this picture.

  2. B.

    (Show card B). Secondly, you could also separate materials at your home before taking it to the trash. To this aim there will be implemented special containers close to your home where you could drop various types of materials, such as organic material, glass, paper, plastic and wood.

  3. C.

    (Show card C). Thirdly, all the waste management system with the landfill can be moved away to some other location far away from any population, and therefore there will be no potential health or quality of life effects for your household, your neighborhood or any population in the island.

Since these measures cost money, in this study we would like to know whether you would vote for the implementation some of these measures, or all of them, if you have to pay an annual tax that will be utilized only for the purposes of financing these measures.

Please, we would like you recall that you and your family have a limited amount of income and there might be some other public issues that you may also favor and for which you may be interested in their implementation, as well as other necessary expenses that you might have for your life.

The next card shows two alternatives for the implementation of the policy measures presented above (please, interviewer explain policy content of the alternatives), together with the current situation regarding these measures (please, interviewer explain that none of the polices are implemented under this alternative). The policy alternatives involve paying a tax for their implementation. Considering that you are presented with only these three alternatives, which one would you choose?


Appendix 2: Diagnostic Plots of MN-MNL

This section collects son diagnostic plots employed to estimate the necessary number of segments and to choose a suitable variable for ordering the segments.

Fig. 7

Diagnostic plots to identify efficient parameter ordering restrictions

Figure 7 presents three plots useful for implementing some pre-analysis diagnostics. The plot on the left (Plot A) represents simulations of the mean parameter (\(\alpha \)) versus the variance (\(\sigma ^{2}\)). It can be seen that identification via ordering the segments with respect to \(\alpha \) is a sensible constraint to endure identification, since three groups are clearly separate it in that domain. However, the alternative of using \(\sigma ^{2}\) as an ordering criteria does not seem as effective as using \(\alpha \), since there is not a clear grouping of responses in finite components.

Plot B is presented in the middle of Fig. 7 and reflects the same information but for the values of the means of each component of the mixture (\(\alpha _{k}\)) versus the mean of the other estimated components (\(\alpha _{l}\) for \(k \ne l\)). It shows that the components of the mixture are well separated when considering ordering \(\alpha \) in the restriction, as there are no points on the diagonal. This support the identification restrictions presented in the MN-MNL model section. Finally, it can be seen in the plot on the right (Plot C) that there are many points close to the diagonal discouraging the use of the variance as a ordering restriction.

Figures 8 and 9 show the MCMC draws and posterior density plots for the unrestricted MN-MNL model (Fig. 8) and a MN-MNL model with ordering constraints via the mean parameters (Fig. 9), that is, \({\upalpha } 1 < {\upalpha } 2 < {\upalpha } 3 < {\upalpha } 4\).

Fig. 8

MCMC traces and posterior densities for \(\alpha \) and \(\sigma ^{2}\) for the unrestricted version of the MN-MNL

It can be seen on Fig. 8 that there are clear suspects of label switching when no restrictions are included. However, Fig. 9 shows that when the proposed ordering restrictions are included, the posterior looks uni-modal and evidences of convergence and stability in the MCMC posterior simulations are found.

Fig. 9

MCMC draws obtained for \(\alpha \) from the sampler under the constraint \(\alpha _{1} < \alpha _{2} < \alpha _{3}<\alpha _{4}\)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

León, C.J., Araña, J.E., de León, J. et al. The Economic Benefits of Reducing the Environmental Effects of Landfills: Heterogeneous Distance Decay Effects. Environ Resource Econ 63, 193–218 (2016).

Download citation


  • Discrete choice
  • Distance decay effects
  • Heterogeneity
  • Landfills
  • Recycling
  • Stated preference
  • Waste management