Vous êtes sur la page 1sur 14

A Cost/Benefit Analysis of the University of Wisconsin - Marshfield/Wood County Campus Recycling Program

Daniel J. Burkhart, Todd Witt, Tyler Gruen, and Malcolm Gold

University of Wisconsin Marshfield / Wood County Spring 2011

Abstract: The University of Wisconsin Marshfield/Wood County recycling program conforms to state and city mandates and currently uses receptacles that vary by type, color, and location throughout the campus buildings. Survey responses by faculty, potential, and current students reveal the intrinsic valuation and preference toward homogenous receptacles outweighs the cost of switching some of the campus containers. Survey results indicated that a small reassessment of the type of the recyclable receptacles would be beneficial received by the campus.

Corresponding Author. We greatly appreciate the ability to interview Brian Panzer with the city of Marshfield and Tom Waalkens and Dan McCollum of UW-M/WC, and thank them for their time. This report was generated as a result of the Spring 2011 Eco 342 Environmental Economics class project.

Background
The Waste and Recycling Program on the University of Wisconsin Marshfield/Wood County (UW-M/WC) Campus divides waste, paper, and commingled recyclables into three respective containers within the building and also to their own separate dumpsters outside. UW-M/WC recycling program coordinates statutory regulations and needs from the institution (UW Colleges), state, and the city and county (UW-M/WC Commission). The campus waste and recycling program is a portion of the Veolia contract covered by the City of Marshfield. This program complies with the priorities described by Wisconsin State Statute 287.05 (12) and initiative 287.05 (4m) That the initiatives of current recyclers which facilitate reuse and recycling through separation, collection and processing of substantial volumes of scrap and waste material, reducing the amount of mixed solid waste that is disposed of in landfills or burned without energy recovery in incinerators, should be encouraged. In addition the City of Marshfield Municipal Code Chapter 12 Section 12-61 requires the separation of paper from commingled recyclables, which includes aluminum, glass, plastic, steel and other metal. The municipal code provides the more stringent legislature by which UWM/WC operates its recycling and waste disposal.

Research Question
Given the background and statutory obligations mentioned above, we conduct a costbenefit analysis (CBA) where we focus on the improvement of campus-recycling program. A CBA compares the benefits to costs associated with a specific action or decision, and is a common decision-making tool required for policy recommendations encompassing health, public

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

2|Page

transportation, and many political actions.1 A CBA is one an economic analysis utilized by private businesses and nonprofits as well. (Young and Steinberg, 1995) The basic concept of a CBA is if the benefits outweigh the costs, the CBA supports the decision. If the opposite is true the CBA indicates the decision should be avoided. We focus our CBA on different campus recycling program attributes since the existence of the campus-recycling program is a state and city mandate. Based on conversations with students, staff, and faculty and our independent research of the current recycling options on campus we chose to examine the following attributes: the number of recycling containers on campus (Count), consistency of style and color among recycling containers (Consistent), and whether recycling containers are in each classroom (Classroom). To estimate the benefits of each of these attributes, we utilize a survey which targets preferences toward each of these options. A statistical analysis of the survey responses allows us to determine the dollar valuation associated with each attribute, as well as whether the recycling program option actually determines individual preferences.

Survey Design
We use a dual-response choice-based conjoint analysis survey to elicit individual preferences. This survey design replicates the typical decision making process of consumers at a store, where an individual is faced with a limited set of selections and chooses one or none. The main characteristic distinguishing choice-based conjoint analysis from other types is that the respondent expresses preferences by choosing concepts from sets of concepts, rather than by rating or ranking them. (Sawtooth 2011) The dual-response nature of the survey allows the

Examples of a CBA include US Department of Health And Human Services (1993), US Federal Highway Administration (2003), and Sunstein (2002). A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 3|Page

respondents to first choose the preferred option from the limited choices, and then whether none is better than their preferred option. By offering this option, the researcher can identify the decrease in demand from the preferred option if it became undesirable by the respondent. This survey design allows researchers to interpret data across all respondents into a quantifiable outcome. The choice-based conjoint analysis has many advantages that allow the researcher to obtain a quantifiable result; however there are disadvantages with this design. If the number of attributes involved is large, the respondent has a lot of information to take in to accurately answer each option and this analysis is not recommended. (Sawtooth 2011) The survey asks respondents to pick between a set of recycling program alternatives, which vary in their attributes and price. The number of recycling containers on campus, Count, varied across three values (10, 20, or 30 containers)2, Consistent and Classroom were binary (Yes or No), and additional price for the option ranged from -$2 to $10 per person. To keep the survey brief, we implement a fractional factorial design which asks individuals to choose the best of three options only a handful of times. The fractional factorial design requires an orthogonal condition in order for the statistical estimates to be identified. The three attributes Count, Consistent, and Classroom are perfectly orthogonal in the survey design. After the decision is made from the three available options, we implement a dual-choice design which asks a followup question asking whether the preferred option is desired compared to changing nothing. The first survey question is below and is an example of the conjoint choice based survey design.

The UW-M/WC campus has 18 comingled recycling containers and 20 paper recycling containers on campus, thus we used 20 as a reference point and allowed the containers to vary by 50% either direction. A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 4|Page

A second follow-up question uses the option selected above and asks whether the individual would prefer this option compared to doing nothing. The status quo on campus is the reference point implied within the question.3 Each survey respondent faced a similar question, where the option varies according to the previous answer. Eight set of dual-choice problems were utilized in the survey.

Survey responses were consistent between on campus individuals and potential students, indicating the reference point and current campus program did not influence survey responses. A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 5|Page

Results
The survey was deployed electronically and 149 respondents completed 20 questions during a five day period. The recycling program and any changes to it affect not only current students, but also faculty and staff as well as future students. Each vested group of campus affiliates are represented in the survey. The table below describes the demographic categories of survey respondents. These response rates are consistent with overall campus demographics. Table 1: Survey Summary Statistics Demographic Category N Age 21 and under 95 22 and over 54 Affiliation Current Student 126 Employee 13 Potential Student 10 Total 149

Percent 63.76 36.24 84.56 8.72 6.71

Survey responses indicate the relative importance, out of 100, of the various attributes as Count 4.6, Consistent 35.6, Classroom 23.4, and price at 36.4.4 The overall relative importance is calculated as the frequency of each attribute being present in the preferred option. This analysis examines the correlation between the preferred option and various attribute presence, and does not account for the simultaneity of the various attributes within the decision process. This basic analysis reveals that consistent containers, price, and classroom containers are important campus recycling program attributes but the number of containers is not important.

Price varies in a second follow-up set according to an individuals responses in the first dual-choice set, and is not perfectly orthogonal to the other attributes. As such the interpretation of relative importance is best made between Count, Consistent, and Classroom. 6|Page

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

Chart 1 below reveals relative importance by current students compared to the other two groups of respondents. Pair wise t-tests indicate with statistical significance that demographic groups value the program attributes differently.

We utilize a conditional logit statistical model which can simultaneously capture colinearity of price and other attributes, the difference between demographic groups, and test for statistical significance. The conditional logit model uses maximum likelihood to estimate the survey probabilities associated with making a decision according to the following model:

P(Yij j )

exp( X ij )

J j 1

exp( X ij )

where Yij are the different alternatives each individual faces, and Xij are the set of explanatory variables. To capture the differentiated valuation of each attribute across demographic groups we create additional variables to capture the attribute importance for students. Thus we interact each attribute with Student, which is a binary variable set to 1 if the survey respondent is a current

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

7|Page

student. The full set of explanatory variables include recycling program attributes (Count, Consistent, Classroom, and Price) interacted with Student that vary with each observation i and across alternatives j.5 Below are the estimates from the conditional logit model.

Table 2: Conditional Logit estimation of UW-M/WC recycling attributes Variable Estimate Odds-Ratio Count -0.121 0.886 Student x Count Additional Effect 0.181 1.198 Consistent 1.636 5.136 Student x Consistent Additional Effect -0.667 0.513 Classroom 0.962 2.616 Student x Classroom Additional Effect -0.253 0.776 Price -0.127 0.881 Student x Price Additional Effect 0.260 1.296 Pseudo R2 0.204 Number of Observations 5940

p-value 0.255 0.120 0.000 0.009 0.000 0.215 0.030 0.000

Note: The statistical model has the binary decision variable (1 for yes) as the dependent variable with the eight independent explanatory variables above. Grouping occurs at the 1192 individual choice-based decisions.

The first column indicates the maximum likelihood estimate from the model, with the second column being the marginal odds-ratio interpretation, and the third column the p-values indicating statistical significance. The Count estimates are not statistically significant; indicating the number of containers on campus is not a deciding factor when choosing amongst alternatives. The estimates for Consistent indicate on average non-student survey respondents will choose an option with consistent recycling containers 84% of the time.6 Students do not value Consistent containers as much as other survey respondents, as the additional effect are negative and statistically significant from zero. The overall estimate for the student population is 0.969 with

This model estimates the overall average effect of each attribute within the survey. Fixed effect estimates pick up the difference between current students and other survey respondents, even though the student interaction do not vary within the an observation. 6 The Odds-Ratio represents the P (Attribute = 1) / P (Attribute = 0), which can be rearranged to find the probability that the attribute will choose that option, with all other attributes held constant. A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 8|Page

an odds-ratio of 2.63, indicating controlling for other attributes students chose options with consistent containers 72.5% of the time. Survey estimates indicate having recycling containers in the classroom is also a statistical significant effect for students and all survey respondents, with the likelihood of choosing an option with Classroom controlling for all else at 67% for students, or approximately 2 out of every 3 choices. Lastly, price was a statistically significant determinant in choosing options, with current students choosing more expensive options and other survey respondents preferring cheaper options. Overall an option that is an additional dollar is 53% likely to be chosen by students, which is statistically significant from 50% but a small overall impact once other attributes are controlled for.

Policy Considerations
Based on the estimates from the above model, we focus the CBA on recommendations associated with statistically significant estimates. As such we examine the potential of changing to consistent recycling containers on the UW-M/WC campus and the possibility of classroom containers.

Consistent Campus Recycling Containers


We first examine the policy recommendation of changing the campus recycling containers such that paper recycling containers are consistent throughout campus and comingled recyclable containers are consistent throughout campus. As such an individual would be able to identify the difference between recycling and waste based on the type and style of container in addition to any signage. There is nothing within our survey that indicates whether the two types of recycling containers should be the same as each other, or different. Based on the logit model estimates a student (entire sample) is indifferent between a $7.31 ($10.84) increase in cost and
A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 9|Page

consistent containers on campus, which indicates the per person intrinsic valuation of having consistency amongst campus recycling containers. Table 3 below includes the estimated benefits per person and campus wide with an implementation of consistent campus recycling containers.

Table 3: Estimated benefits of consistent recycling containers. Per-Person Benefit Campus-Wide Benefit Students $7.31 $4,571 Entire Sample $10.84 $7,931
Note: 625 Students and 80 Faculty and Staff are used as an estimate of the campus. (UW-M/WC 2011)

In order to obtain consistent recycling containers on campus, different options would need to be explored. One such possibility is to replace a non-blue plastic campus recycling containers at $75 each, which would require the purchase of 26 additional containers. (McCollum, 2011) This would require an outlay of $1,950 one time, but would return over $4,500 each year in student valuation. The longevity of the benefits as well as the one-time snapshot estimates all indicated that the benefits of such a change outweigh the costs. The recommended change is only one such possibility and it appears through the estimate of the benefits that most changes directed at consistency would be beneficial to campus even at a greater cost.7

Classroom Recycling Containers


We now consider a policy recommendation of including campus recycling containers in each classroom. Survey estimates indicate a student is indifferent between an additional $5.33 cost per person and having classroom containers. Overall this translates into an additional $3,331

If the return on investment were to be required to be recouped in five years with a discount rate of 5%, expenses up to $20,000 would be supported by this analysis. A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 10 | P a g e

of student satisfaction by having recycling containers in the classroom. The overall campus estimated benefits are contained in Table 4 below.

Table 4: Estimated benefits of classroom recycling containers. Per-Person Benefit Campus-Wide Benefit Students $5.33 $3,331 Entire Sample $6.96 $4,908
Note: 625 Students and 80 Faculty and Staff are used as an estimate of the campus. (UW-M/WC 2011)

In order to implement classroom recycling containers, additional manpower would be utilized in addition to the purchase of extra containers. If were to only take one additional minute to empty the two recycling containers in a classroom, this would result in 20 minutes per night for the campus classrooms. This result is an estimated annual cost of $2,600 in labor alone based on a conservative estimate of the time required. Purchase of 40 14-quart classroom recycling containers at 10.50 each, would require a one-time outlay of $420. Additional containers in the classroom also impose a space constraint and logistical constraint on maintenance, which was not considered in this analysis. Without this extra space consideration, the policy recommendation requires additional analysis but at best will result in a small net benefit with the annual benefits being offset with the annual labor costs.

Consistent Classroom Recycling Containers


The above policy consideration relies on the statistical analysis which only examines the main effects of changing one attribute and not any interaction effect between two attributes. An alternative specification of the logit model in which Classroom and Consistent are interacted is estimated to determine whether there would be any positive effects of changing to consistent

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

11 | P a g e

recycling containers on campus with recycling containers in the classroom as well. Estimates from this model specification are contained in Table 5.

Table 5: Conditional Logit estimation of UW-M/WC recycling attributes Variable Estimate Odds-Ratio Count 0.033 1.033 Student x Count Additional Effect 0.156 1.168 Consistent 2.493 12.097 Student x Consistent Additional Effect -0.648 0.523 Classroom 2.050 7.764 Student x Classroom Additional Effect -0.400 0.671 Classroom x Consistent -2.179 0.113 Student x Classroom x Consistent 0.257 1.293 Additional Effect Price -0.095 0.909 Student x Price Additional Effect 0.258 1.295 2 Pseudo R 0.204 Number of Observations 5940

p-value 0.768 0.200 0.000 0.046 0.000 0.191 0.000 0.515 0.104 0.000

Note: The statistical model has the binary decision variable (1 for yes) as the dependent variable with the eight independent explanatory variables above. Grouping occurs at the 1192 individual choice-based decisions.

Survey respondents were dichotomous in their preferences, preferring either classroom containers or consistent containers on campus. The interaction between Classroom and Consistent is statistically significant and negative, offsetting the positive benefits from one of the policy changes recommended above. The benefits from changing to consistent containers or classroom containers are nearly identical, but are not augmented by changing both simultaneously.

Conclusion
Recycling at UW-M/WC is currently mandatory according to state and city laws. We examine efficiency within the program and use a cost-benefit analysis to determine what recycling program attributes should be modified. We collect data through a conjoint-based dual
A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program 12 | P a g e

choice survey and find the campus would benefit from changing to consistent recycling containers as the intrinsic benefits outweigh the costs. This recycling program on campus impacts the city, county, state, and campus individuals. Recycling reduces the landfill and pollution effects on the area and state. Current political discourse is examining whether recycling will be included in the Wisconsin State Statutes. (Bergquist and Sandler, 2011) Campus and local recycling currently does not capture 100% of all recyclable materials. (Panzer, 2011) The city waste and recycling contract determines prices for dumpsters, but the size and frequency of pickup are flexible. All of these possibilities were considerations in the Eco 342 class project, but we focused on our project due to the state and local mandates and through the timeliness of getting survey information. Each of the above considerations is important, but not included in our analysis.

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

13 | P a g e

Bibliography
Bergquist, Lee, and Larry Sandler, (2011) "Recycling Mandate Eliminated in Budget, Milwaukee Journal Sentinel.1 Mar. 2011. McCollum, Dan. (2011) UW-M/WC Assistant Campus Dean class interview, 30 Mar. 2011. Panzer, Brian. (2011) Marshfield City Garage class interview, 30 Mar. 2011. Sawtooth Software Inc. (2008) "The CBC System for Choice-Based Conjoint Analysis." Sawtoothsoftware.com. Web May 11, 2011. Sunstein, Cass R. (2002) Risk and Reason: Safety, Law and the Environment. Cambridge, United Kingdom: Cambridge University Press. US Department of Health and Human Services. (1993) Feasibility, Alternatives, And Cost/Benefit Analysis Guide Web May 11, 2011. US Federal Highway Administration. (2003) US Federal Highway Administration: Economic Analysis Primer: Benefit-Cost Analysis Web May 11, 2011. UW-M/WC. (2011) Enrollment up at UW/Marshfield-Wood County Web April 27, 2011. Young, Dennis and Richard Steinberg. (1995) Economics for Nonprofit Managers Foundation Center.

A Cost/Benefit Analysis of the UW-M/WC Campus Recycling Program

14 | P a g e

Vous aimerez peut-être aussi