Reducing air pollution through behavioral change of wood-stove users: Evidence from an RCT in Valdivia, Chile

Ruiz-Tagle, J. Cristóbal; Schueftan, Alejandra Working Paper Reducing air pollution through behavioral change of wood-stove users: Evidence from an RCT in Valdivia, Chile IDB Working Paper Series, No. IDB-WP-959 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Ruiz-Tagle, J. Cristóbal; Schueftan, Alejandra (2019) : Reducing air pollution through behavioral change of wood-stove users: Evidence from an RCT in Valdivia, Chile, IDB Working Paper Series, No. IDB-WP-959, Inter-American Development Bank (IDB), Washington, DC, https://doi.org/10.18235/0001894 This Version is available at: https://hdl.handle.net/10419/208154 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/igo/legalcode Reducing Air Pollution through Behavioral Change of Wood-Stove Users: Evidence from an RCT in Valdivia, Chile J. Cristóbal Ruiz-Tagle A lejandra Schueftan IDB WORKING PAPER SERIES Nº IDB-WP-959 September 2019 Department of Research and Chief Economist Inter-American Development Bank September 2019 Reducing Air Pollution through Behavioral Change of Wood-Stove Users: Evidence from an RCT in Valdivia, Chile J. Cristóbal Ruiz-Tagle* A lejandra Schueftan** * Environmental Defense Fund ** Instituto Nacional Forestal, Chile Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Ruiz-Tagle, J. Cristobal. Reducing air pollution through behavioral change of wood-stove users: evidence from an RCT in Valdivia, Chile / J. Cristobal Ruiz-Tagle, Alejandra Schueftan. p. cm. — (IDB Working Paper Series ; 959) Includes bibliographic references. 1. Stoves, Wood-Chile. 2. Indoor air pollution-Chile-Prevention. I. Schueftan, A lejandra. II. Inter-American Development Bank. Department of Research and Chief Economist. III. Title. VI. Series. IDB-WP-959 Copyright © Inter-American Development Bank. This work is licensed under a Creative Commons IGO 3.0 AttributionNonCommercial-NoDerivatives (CC-IGO BY-NC-ND 3.0 IGO) license (http://creativecommons.org/licenses/by-nc-nd/3.0/igo/ legalcode) and may be reproduced with attribution to the IDB and for any non-commercial purpose, as provided below. No derivative work is allowed. Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the UNCITRAL rules. The use of the IDB's name for any purpose other than for attribution, and the use of IDB's logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this CC-IGO license. Following a peer review process, and with previous written consent by the Inter-American Development Bank (IDB), a revised version of this work may also be reproduced in any academic journal, including those indexed by the American Economic Association's EconLit, provided that the IDB is credited and that the author(s) receive no income from the publication. Therefore, the restriction to receive income from such publication shall only extend to the publication's author(s). With regard to such restriction, in case of any inconsistency between the Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives license and these statements, the latter shall prevail. Note that link provided above includes additional terms and conditions of the license. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. http://www.iadb.org 2019 1 Abstract* Ambient air pollution is a serious problem in cities in south-central Chile because of massive combustion of wood as fuel for residential heating. To reduce air pollution emissions, Chile’s environmental authority has implemented a largescale program consisting of replacing old, highly polluting residential wood stoves with new, less polluting ones. However, to extend burning time and save on wood fuel expenditures, users tend to dramatically constrain the air flow in these new wood stoves, which creates a highly polluting combustion process. To address this issue a behavioral intervention was designed to provide users with feedback on their wood stoves’ air pollution emissions. The intervention consists of a metallic sign that aligns with the wood stoves damper lever and that informs users of the level of pollution emissions according to the chosen setting of the wood stove’s damper. To assess the effectiveness of this information sign, a randomized controlled trial (RCT) is conducted in selected households in the city of Valdivia, Chile. Results from this intervention show that the information sign induced a behavioral change in wood stove users that translates into a 17.3 percent reduction in residential pollution emissions. JEL classifications: C93, O13, Q53, Q56. Keywords: Field experiment, Environment and development, Air pollution, Wood stoves * We appreciate the financial support of the Inter-American Development Bank (IDB). We additionally benefited from discussions with Bridget Hofmann and Sebastián Miller (both at IDB) that greatly improved the experimental design. We are also grateful to Rene Reyes (at Instituto Forestal, Sede Los Ríos) and Oscar Pilichi (at Universidad Austral de Chile) for valuable support in conducting the field work. All errors and omissions are our own. Corresponding author email: cristobal.ruiz[email protected]. 2 1. Introduction Ambient air pollution caused by burning wood fuel for heating and cooking is a serious concern in many urban areas of the developing world (Chávez, Stranlund and Gómez, 2011). This problem is particularly serious during the cold months in the south-central region of Chile, where ambient air pollution concentrations, largely from wood stoves, have reached extremely high levels, causing serious consequences for the health and wellbeing of the population. According to a 2018 ranking of the world’s most polluted cities for fine particulate matter (PM2.5), eight of the top 10 most polluted cities in South America are in Chile’s south-central region. Valdivia, the city where we conducted the field experiment discussed in this paper, is ranked fourth. In addition, seven more cities in this region appear among the top 20.1 To address this issue, Chile’s environmental authority has pushed in recent years for a large-scale program to replace high-polluting residential wood stoves with new and less polluting models.2 These heavily subsidized woodstoves, as well as all those that have been available in the market in the last decade, feature state-of-the-art combustion technologies. This results in minimal indoor pollution emissions and low outdoor pollution emissions, but only under “optimal” operating conditions and not necessarily under “real world” operating conditions. This is because wood stove outdoor pollution emissions vary largely according to the setting of the stove’s damper, which regulates airflow inside the combustion chamber, and these new wood stoves allow users to freely choose the damper’s setting. Moreover, to extend wood-burning time and reduce expenditure on wood fuel, users usually set the damper so as to fully choke airflow. Wood stove outdoor pollution emissions present a highly non-linear relationship with respect to the damper setting, and those emissions when fully choked are many times larger than those in any other setting (even compared to emissions when only partially choking them). As a 1 The cities in Chile’s south-central region listed in this ranking are: Padre Las Casas (Rank #1), Osorno (#2), Coyhaique (#3), Valdivia (#4), Temuco (#5), Linares (#8), Rancagua (#9) and Puerto Montt (#10). Santiago, Chile’s capital but not in the south-central region, is ranked #6. Source: www.airvisual.com/world-most-polluted-cities 2 Chile’s environmental authority has designed and implemented Air Pollution Control and Prevention Plans (PPDA, according to the Spanish acronym) in most of the cities in this region. There are currently 10 PPDAs at different stages of implementation. The PPDAs consider four lines of action focused on improving the energy efficiency of homes, the quality of heating systems, the quality of fuels for energy, and education programs. However, these policies have not yielded the expected results and, despite their implementation, ambient air pollution concentrations have continued to increase in all major cities of the region (Schueftan and González, 2015). 3 consequence, the choking of wood-stoves can result in outdoor pollution emissions as much as five times greater than under optimal wood fuel burning conditions (Jordan and Seen, 2005).3 Moreover, because most dwellings in cities of south-central Chile suffer from high rates of infiltration (through windows, doors, walls, roofs and ceilings that are not properly sealed) ambient air pollution from wood fuel burning leaks to the interior of dwellings. This generates a problem of local air pollution where the most affected individuals are members of households using wood stoves and their close neighbors. As a consequence, when aggregated at the neighborhood and city level in these cities in south-central Chile, individual household behavior creates large systemic effects on overall concentrations of ambient air pollution.4 As per the individuals that operate their woodstoves, it is certainly more convenient to choke them, as doing so extends wood fuel burning time of the wood-fuel. Thus, in addition to the convenience of not constantly having to load their woodstoves, choking wood stoves allows users to save on wood fuel expenditures. Therefore, users’ decision on whether or not to choke their own wood stoves can be characterized as a voluntary contribution to a local public good in which wood stove users face individual (or household-level) incentives to choke their woodstove at the expense of (marginally) contributing to higher local ambient air pollution concentrations. Furthermore, whereas the benefits of using their wood stoves in such a way that saves on wood fuel are immediately perceptible (via a comfortable warm dwelling and savings in fuel costs), the costs of air pollution are shared among the neighboring population (regardless of who emits air pollutants). While those costs do not necessarily produce immediate consequences, they are likely to be cumulative over time.5 In this paper we examine whether users can be nudged into low outdoor air pollution emissions by providing information feedback on their wood stove emissions. As wood stove pollution emissions vary according to damper setting, which in turn, adjusts the airflow in the 3 The choking of these wood stoves creates smoldering and low-temperature combustion, thus creating a highly polluting combustion process with no vivid flame that efficiently burns wood-fuel. Moreover, when choking these wood stoves the smoke from the burning of wood-fuels exits the chimney at a low temperature and so that reaching a low elevation upon exiting. Thus, as opposed to being taken away by stronger winds at higher elevations, this causes the smoke from burning wood-fuel to remain at ground-level, which further exacerbates the problem of high air pollution concentrations. 4 This problem is even more serious in low-income neighborhoods where there is a higher density of dwellings (with low quality construction materials and more infiltrations), which creates higher concentrations of air pollution with more emitters and less distance between them. 5 Miller and Ruiz-Tagle (2018) provide causal estimates of the effect of exposure to air pollution on infant mortality for Santiago, Chile. The authors find that the cumulative effects (over a six-month period of exposure) of air pollution on infant mortality could be twice as large as the acute (same-week) effects. 4 combustion chamber—the information sign informs users of current air pollution emissions of their own respective wood stoves according to the damper setting selected. Our hypothesis is that users will respond to this visual aid by decreasing the frequency of the choking of their wood stoves. Therefore, in this paper we evaluate whether providing an information sign that delivers real-time feedback on wood stove pollution emissions can effectively reduce outdoor air pollution emissions. We assess this by conducting a randomized controlled trial (RCT) involving a framed field experiment in which we provided a random group of households (treatment group) with an information sign. Users subject to this intervention were free to set their woodstove’s damper in any way they want—choking it to save on wood fuel expenditure, opening the airflow in the combustion chamber to decrease outdoor pollution emissions, or any setting in between. However, by means of the information sign users in this group knew that any setting of the damper they choose is associated with a certain level of their wood-stoves’ actual outdoor pollution emissions. On the other hand, users in the control group were not provided with this information sign. Moreover, the actual setting of wood stoves dampers was monitored and recorded for both treatment and control groups. Our results show that conveying feedback on wood stove emissions induces a significant behavioral change in wood-stove users by reducing the frequency with which they choose the fully choked damper setting. Moreover, this induced behavioral change has also important effects on reducing wood-stoves’ outdoor air pollution emissions. Our results show that providing users with this information significantly reduces wood stove outdoor emissions of fine particulate matter (PM2.5). The rest of the paper is organized as follows. The next section reviews the related literature, and Section 3 describes in greater detail the problem of air pollution in these cities as well as policies targeted to address this problem. Section 4 explains the experimental design, and Section 5 presents the main results of the behavioral intervention. Finally, Section 6 presents concluding remarks. 2. Literature Review 2.1 Literature on Air Pollution from Burning Wood Fuel in Chile Chile’s environmental authority (Ministry of Environment) has designed a battery of policies to address the problem of ambient air pollution under the Air Pollution Control and Prevention 5 Plans (PPDA). Chile’s Ministry of Environment has produced official reports that assess ex ante the expected effect of different policies on mitigating the problem of ambient air pollution (MMA, 2014a). These reports predict scenarios for the different programs that are being implemented under the PPDA as well as the expected effects that those programs could have in the next 15 years. Using official data from laboratory tests of the performance and emissions produced by different types of wood stoves, these reports assess the expected effects of the wood-stove replacement programs. Users’ behavior regarding actual operation of these woodstoves, however, has not yet been assessed, and that behavior can have a large effect on pollution emissions. On the other hand, recent academic papers assess the potential effect of alternative policies to reduce air pollution. For instance, Chávez et al. (2009) examine the effect of price incentives such as demand and supply subsidies as well as strengthening command-and-control policies of standards on wood fuel consumption.6 Additionally, UFRO-CONAMA (2009), Chávez, Gómez and Salgado (2011); and Gómez et al. (2017) have focused on evaluating the wood-stove replacement program, particularly for the city of the Temuco—which was the first city to have a PPDA and to extensively apply these pollution reduction policies.7 Furthermore, more recently Gómez et al. (2014); Gómez et al. (2013); and Jaime, Chávez and Gómez (2017) have focused on assessing the design of subsidies and the effect of prices and income in order to increase adoption of cleaner wood fuels and improved wood stove devices. However, all these papers acknowledge that the real effect on emissions reduction is uncertain due to the impact of users’ behavior when operating their wood stoves. Moreover, a recent paper by Schueftan et al. (2016) compared the environmental, economic and social effect of different policies focused on improving heating systems, quality of fuels and energy efficiency of dwellings. Schueftan et al. (2016) argue that the most effective strategy is to focus on reducing energy demand by improving the energy efficiency of 6 Chávez et al. (2009) examined the use of certified wood fuel with appropriate moisture content. The authors conclude that command-and-control policies are more cost effective and therefore be the preferred environmental policy for tackling air pollution caused by burning wood fuel in cities in the south-central region of Chile. Therefore, the use of certified-dry wood fuel with an appropriate moisture content should be enforced, as only 1 percent of dwellings use certified-dry wood fuel due to its higher price. However, the effects of command-and-control policies on reducing air pollution are expected to be small. 7 These studies presented different technology options to induce the adoption of cleaner and more efficient wood burning technologies. Jaime, Chávez and Gómez (2017) considered a pilot program and studied the factors that affected the decision to participate in these programs. They found that the availability of a subsidy and the possibility of having a credit were both very important, especially in low-income households. 12 Figure 4. Information Sign Attached to the Top of the Wood Stove Note: The information sign aligns with the damper’s lever and informs wood-stove users on emissions produced by each damper setting. To register the actual damper setting at each point in time this device was attached to the wood stove of each participating household for the entire duration of the experiment. Damper adjusts the air inflow inside the woodstove’s combustion chamber Damper setting monitoring device 13 4.3 Field Experiment We recruited 80 households to participate in a field experiment. All participating households had a wood stove that they used as the only source of heating in their dwelling. A household was eligible to participate as long as the damper setting monitoring device could be installed in the household’s wood stove. The damper setting monitoring device was designed so to fit the most popular brand and model of wood stoves in Valdivia (Bosca Limit 360 or Bosca Limit 380), but there were a few additional cases in which the monitoring device was also installed in wood stoves of a different brand and model. The experiment was conducted in two phases of 40 households each, and each phase lasted for one month. The two phases of the experiment were implemented in the city of Valdivia, Chile during the months of August and September of 2017. For each phase, a subset of participating households was randomly assigned to a treatment group (comprised of 26 households in each phase) and a control group (comprised of 14 households in each phase), so that adding up both phases there were 52 households randomly assigned to the treatment group and 28 household randomly assigned to the control group.19 Household members were not aware of whether they were assigned to a treatment or a control group and did not know to which group other participating households were assigned. Furthermore, we believe that there was no communication between participating households. All participating households signed an informed consent form before starting the experiment which informed them of the different phases of the experiment.20 As an incentive for participation, each participating household received one cubic meter (1 m3) of certified-dry wood fuel. This also allowed us to make sure that the quality of wood-fuel being used throughout the experiment does not affect users’ behavior regarding the setting of the wood-stove’s damper. For all participating households (both in the treatment as well as in the control group) the wood stove’s damper setting was recorded for a full month, using the damper setting monitoring device as illustrated in Figures 4 and 5. Furthermore, after two weeks, the information sign was installed only in the wood stoves of those households in the treatment group (see Figure 4). 19 Random assignment to treatment and control groups assures that the two groups are virtually identical, at least from a statistical point of view. 20 The consent form as well as the full design of the experiment was approved by an external ethics review board from the Centre for Experimental Social Sciences of Nuffield College at the University of Oxford (CESSNuffield). The application record is ETH-170526299-3 and is titled “Informational Interventions to Reduce Air Pollution in Chile’s Southern Cities.” 14 Moreover, the installation of this information sign was supplemented with an information flyer in the form of a refrigerator magnet. This flyer, in addition to clearly explaining the information conveyed by the information sign, contains more detailed information on wood stoves’ effect on the city’s ambient air pollution (see Figure 5). During the installation visit, the technician explained in detail to the household head the meanings of both the information sign and the flyer and answered any questions household members may have had. We believe that the flyer strengthened the link between the chosen damper setting and their associated effects in term of wood pollution emissions (Fischer, 2008). Figure 5. Magnetic Flyer Explaining Meaning of the Information Sign Note: This flyer was provided to all households in the treatment group at the time that the information sign was installed in their wood stoves. The flyer was a magnet that can be posted on their refrigerator, which makes it always visible. Original Flyer Translation Damper use of our wood stoves What is its effect on the air we breathe? Valdivia [city] suffers from high air pollution due to inefficient use of wood stoves. Damper use drives wood stoves’ emissions of air pollutants. How? A choked wood stove emits much more air pollutants than a wood stove with an open damper. This air pollution also filters inside the dwelling through doors, windows and drafts. Information sign The installed signage represents your wood stove’s air pollution emissions for each setting of the damper. Very High High Mid-Level Good Ignition 15 Furthermore, for each participating household, in both the treatment and control groups, an indoor temperature monitoring device was installed approximately 2 meters from the wood stove (at about 2 meters from the woodstove, which is usually located in the living room). The purpose of this temperature monitoring device is to record indoor temperature so that, by contrasting it with outdoor temperature, we are able to determine when the wood stove is actually in use. In addition, an enumerator applied a survey to each participating household to gather information on socio-economic characteristics of the households, experience of any healthrelated problems, characteristics of dwellings, quality of wood fuel used and means for acquiring it. The survey also asked questions regarding the household head’s opinion on air pollution as well as frequency of use of the wood stove.21 Moreover, a team of enumerators and technicians paid periodic visits to participating households to ensure that the monitoring devices were readily recording information. A follow-up survey was additionally conducted at the end of the participation period to complement information gathered by the initial survey. After a month, and once the intervention was finished, all monitoring devices were collected, the data from them were downloaded and the survey data were digitized. 5. Results 5.1 Descriptive Statistics for Treatment and Control Groups The randomization of participating households into treatment and control groups guarantees that there are no systematic characteristics of those participating households that may affect the outcome of the experiment. Table 1 provides descriptive statistics of a selected group of household characteristics from the household survey for those households in the treatment and control group. Table 1 presents mean and standard deviation for the following variables: i) number of household members and distribution of their age groups; ii) whether there is any household member that suffers from respiratory or cardiovascular disease; iii) the average number of hours that the wood stove is in use, both during weekdays and on weekends; iv) a self-reported score of the indoor temperature; v) monthly household income; vi) dwelling characteristics regarding ownership, surface area, number of floors and construction year; and vii) survey respondent’s characteristics, such as gender, age, marital status and educational 21 In the next section we present descriptive statistics of these questions from the household survey. 16 attainment. We conducted t-tests and Kolmogorov-Smirnov tests on the means of each of those variables and found no statistically significant difference between households in the treatment and control groups. Consequently, the only observed feature that set these households apart is whether they were assigned to the treatment or control group, which occurred via the random process described in the previous section. Table 1. Descriptive Statistics of Participating Households (treatment and control groups) 5.2 Effect on Frequency of Wood Stove’s Damper Setting In this section we present results from the analysis of the data on wood stove use obtained with the damper setting monitoring device. We restrict our analysis to only those hours of the day Variables Mean S.D. Mean S.D. Household members 3.15 1.03 3.37 1.46 Less than 4 years old 0.19 0.48 0.19 0.44 Between 5 and 14 0.30 0.54 0.50 0.75 Between 15 and 65 2.44 1.15 2.33 1.32 65 and older 0.22 0.58 0.35 0.56 Num. hours woodstove is in use Weekdays 13.4 6.3 12.3 5.0 Weekends 14.9 6.5 15.5 4.8 Indoor temp. score (self reported) 0.88 0.12 0.88 0.11 Monthly HH income (perc.) Less than USD 800 0.41 0.50 0.42 0.50 Between USD 800 - 1,700 0.33 0.48 0.38 0.49 More than USD 1,700 0.19 0.40 0.19 0.40 Dwelling's Ownership = own (perc.) 0.70 0.47 0.71 0.46 Surface area (sq. meter) 70.6 39.2 78.6 39.6 Floors (perc. 1 floor) 0.37 0.49 0.38 0.49 Construction before year 2000 0.56 0.51 0.60 0.50 Const. between 2000 and 2007 0.19 0.40 0.21 0.41 Const. after 2007 0.11 0.32 0.06 0.24 Const. year N/A 0.15 0.36 0.12 0.32 Respondent's Gender (1=male) 0.41 0.50 0.40 0.50 Age 43.0 13.8 47.3 14.2 Marital status = single (perc.) 0.33 0.48 0.31 0.47 Marital status = married (perc.) 0.44 0.51 0.58 0.50 Marital status = divorced/widowed 0.22 0.42 0.12 0.32 Educ. attainment = primary (perc.) 0.15 0.36 0.19 0.40 Educ. attainment = secondary (perc.) 0.41 0.50 0.37 0.49 Educ. attainment = Terc. (technical) 0.19 0.40 0.19 0.40 Educ. attainment = Terc. (university) 0.26 0.45 0.25 0.44 Treatment Group Control Group HH member suffer from resp. or cardio. disease 0.26 0.45 0.25 0.44 17 when the wood stove was in use. According to our survey data, participating households report that their wood stoves are in use an average of 13.32 hours a day (that is, about 55.5 percent of the time). To establish whether a wood stove is in use during any given period of time we contrast the outdoor city-wide temperature with the indoor temperature in the room where the wood-stove is located (as recorded by a temperature monitoring device). We determine that a woodstove is in use when the difference between outdoor and indoor temperature is greater or equal to 11.31 degrees Celsius, which corresponds to the 55.5 percentile of the temperature difference.22 Figure 6. Indoor and Outdoor Temperature Profile for Participating Households in Valdivia, Chile: Average for August and September 2017 Note: The black line denotes average hourly indoor temperature (in Celsius degrees) for participating households and the light blue line denotes average hourly outdoor temperature (in Celsius degrees) for the city of Valdivia. Whereas the outdoor temperature is obtained from Chile’s meteorological service, the indoor temperature is obtained from the Speck sensors, which were located in each of the dwelling’s living room at a distance of about two meters from the household’s wood stove. 22 Figure 6 shows average indoor and outdoor temperature for those participating households. The indoor temperature corresponds to the recorded temperature of the Speck sensor, which was located in the dwelling’s living room at a distance of about two meters from the wood-stove. This means that when the wood-stove is in use, the indoor temperature as recorded by this device is considerable higher than that of the rest of the dwelling. Figure 7 shows that the average indoor temperature peaks soon after midnight and drops monotonically until about 13:00 hrs. (when it reaches its lowest temperature). This temperature profile reflects the most common pattern of household wood stove in Valdivia. Most wood stoves are lit in the late afternoon or evening, and households usually fully load them with wood fuel before going to bed (between 22:00 and 24:00 hrs.), so that the wood stove reaches its peak temperature about an hour later. Then, the wood stove slowly cools down until loaded and started again in the morning, or in the afternoon. On the other hand, outdoor temperature drops overnight slowly until dawn (around 8:00 hrs. during winter), to slowly increase and peak at around 15:00 hrs. 18 Results of the behavioral intervention are presented in Figures 7-9 below.23 Figures 7 and 8 show the frequency of use of the damper setting for the treatment group (Figure 7) and control group (Figure 8), and Figure 9 summarizes the main results by showing the magnitude of the effect of the information sign. Figure 8 shows that, for wood stoves in the treatment group, the most frequent setting of the wood-stoves’ damper before the sign was installed (blue bars) was “Choked” (37.8 percent of the time). On the other hand, the least frequent setting of the wood-stoves’ damper before the sign was installed was “Mostly Open” (6.3 percent of the time). Figure 8 shows an almost identical pattern for wood stoves in the control group. The most frequent damper setting prior to installation of the information sign (dark green bars) was “Choked” (37 percent of the time), and the least frequent setting “Mostly Open” (5.4 percent of the time). Furthermore, by conducting a Kolmogorov-Smirnov test on the distribution of the damper setting before the information sign was installed, we cannot reject the null hypothesis of equal distributions of both control and treatment groups. In other words, according to the labelling and color marks in the information sign, our data shows that most of the time wood stove users set the damper to “Very High” pollution emissions (red mark—see Figure 3). Figure 7 shows that, after information signs were installed on wood stoves in the treatment group, there was a statistically significant reduction in the frequency the damper was set at “Choked” (“Very High” pollution emissions). The figure also shows that, after information signs were installed, there was a statistically significant increase in the frequency at which the damper was set at “Mostly Open” (“Low” pollution emissions). More precisely, the frequency of the “Choked” setting decreased 4.8 percentage points (from 37.8 to 33 percent), and the frequency of the “Mostly Open” setting increased by 4.2 percentage points (from 6.3 to 10.5 percent). Conversely, for all other damper settings no statistically significant difference occurred in the treatment after information signs were installed. 23 Whereas when presenting results we combined results from the two phases, an analysis for each phase (not shown here) yields the same result. Results by phase are available upon request. 19 Figure 7. Frequency of Damper Setting for Treatment Group, Percentage of Time the Wood Stove Is On Note: Blue bars denote the distribution of damper settings of households in the treatment group before the information sign was installed. Red bars denote the distribution of damper settings of the same households after the information sign was installed. The settings are grouped into five categories (Choked, Mostly Choked, Mid-Level, Mostly Open, and Fully Open), and these distributions are calculated based on the percentage of time the wood stoves are actually on. In addition, capped lines denote 95% confidence intervals (C.I.) of these distributions for each damper setting. On the other hand, in order to assess what would have happened without information signs, we examine the control group during the same period. Figure 8 shows that the frequency of the “Choked” setting (“Very High” pollution emissions) experienced a statistically significant increase for wood stoves in the control group. That is, the frequency of the “Choked” setting increased by 7.9 percentage points (from 37.0 to 44.9) on the days that followed the installation of the sign on wood stoves in the treatment group. Our data also show an increase in the frequency of the “Open” setting during that same period of time. Conversely, during that period, 20 there was a decrease in the frequency of all intermediate settings (“Mostly Choked,” “MidLevel” and “Mostly Open”).24 Figure 8. Frequency of Damper Setting for Control Group, Percentage of Time the Wood Stove Is On Note: Green bars denote the distribution of damper settings of households in the control group before the information signage was installed. Turquoise bars denote the distribution of damper settings of the same households after the information sign was installed. The wood stove’s damper settings are grouped into five categories (Choked, Mostly Choked, Mid-Level, Mostly Open, and Fully Open), and these distributions are calculated based on the percentage of time the wood stoves are actually on. In addition, capped lines denote 95% confidence intervals (C.I.) of these distributions for each damper setting. To assess the overall effect of the information sign on wood stove users’ behavior we contrast the difference in frequencies in each setting before and after installing the information 24 These results from the control group could be interpreted as evidence of an experimenter effect or Hawthorne effect, which vanishes over time. Whereas households in both treatment and control groups received long visits from our team of technicians and enumerators at the onset of the experiment (and thereafter only brief check-in visits), only those households in the treatment group received a second long visit two weeks into the experiment (to install the information sign, hand out the flyer and provide instructions on how to interpret it). That is, whereas all households may have perceived that their behavior regarding the use of their wood-stoves was being observed during the entire length of the experiment, this perception may have vanished over time for those households in the control group. If this vanishing experimenter effect holds true, then, as time went by, households in the control group should have gradually returned to their normal wood stove use. 21 sign and the difference between treatment and the control group. This strategy is commonly known as a difference-in-difference approach (Diff-in-Diff). Figure 9 shows the results of the Diff-in-Diff approach in terms of change in percentage points for each damper setting. Figure 9 shows a reduction of 12.7 percentage points in the frequency of the “Choked” setting, as well as a reduction of 3.4 percentage points in the frequency of the “Open” setting.25 On the other hand, Figure 9 shows increases of 4, 6.3 and 5.8 percentage points, respectively, in the frequency of the “Mostly Choked,” Mid-Level” and “Mostly Open” damper settings. Therefore, our results suggest that the information sign effectively nudged wood-stove users to choosing less-polluting damper settings. That is, whereas we observe a large drop in the frequency of the setting that emits “Very High” pollution (“Choked” setting), we also observe an important increase in the frequency of damper settings for “Moderate” and “Low” pollution (“Mid-level” and “Mostly Open” damper settings). Figure 9. Change in Frequency of Damper Setting for Induced by Information Sign: Percentage Points for Each Damper Setting Note: Maroon bars denote the difference-in-difference distribution of the damper setting. That is, the bars show the distribution of the difference in frequencies in each setting before and after installing the information sign and the difference between treatment and the control group (shown in Figures 8 and 9). As noted above, the settings are grouped into five categories (Choked, Mostly Choked, Mid-Level, Mostly Open, and Fully Open), and this distribution is calculated based on the percentage of time the wood stoves are actually on. 25 Recall that the information sign notes that the “Open” damper setting should be used only during ignition, and not at other times. Only those households in the treatment group received this information, which is made explicit on the sign as well as in the magnetic flyer. Therefore, the observed reduction in the frequency of use of the “Open” damper setting attributed to the information sign may likely reflect the fact that wood stove users effectively learned not to use this setting at times other than during ignition. 28 In Table 6 we summarize the information presented in Figures 8 through 10 (or equivalently, that of Panel B of Table 3) and compute the effect of the information sign on reductions in wood stove emissions of fine particulate matter (PM2.5). The first row of Table 6 presents the frequency of damper setting, as shown in Figures 8 and 9, where we have regrouped the categories into “Choked,” “Mid-level” and “Mostly Open” to match those used in the lab studies conducted by Díaz-Robles (2014). Likewise, we also re-grouped the categories of damper setting in the second row of Table 6, which presents the change in the frequency of use of the damper setting induced by the information sign (as shown in Figure 9). Table 6. Damper Setting Usage, PM2.5 Emissions by Wood Fuel Type and Effect of Information Sign on Reducing PM2.5 Emissions Choked Mid-Level Mostly Open Frequency of damper setting usage (%) 1 37.4 39.7 22.9 Change of damper setting usage due to sign (percentage points) 2 -12.7 10.3 2.4 PM 2.5 emissions before sign (g/h) 3 9.8 6.4 2.4 1.0 Reductions in PM 2.5 emissions due to sign (g/h) 4 1.4 2.2 -0.6 -0.1 PM 2.5 emissions before sign (g/h) 3 21.0 15.1 3.9 2.0 Reductions in PM 2.5 emissions due to sign (g/h) 4 3.9 5.1 -1.0 -0.2 (2): Settings 'mostly choked' and 'choked' were agregated into column 'Mid-Level', and setting 'Mos tl y Open' was agregated with 'Open'. See Figure 10 or Panel B Table 3. (3): Avera ge PM 2.5 emissions in gra ms per hour (g/h). Emissions for each damper setting is aweighted average calculated by multiplying frequency of damper setting (first row) by emission factors from Table 4. (4): Reductions in PM 2.5 emissions in gra ms per hour (g/h). Emissions for each damper setting is aweighted average calculated by multiplying change of damper setting due to signage (second row) by emission factors from Table 4. Dry Wood-Fuel (certified) High Moisture Wood-Fuel Damper Setting Total (1): Avera ge damper setting a cros s treatment and control groups before signage. Settings 'mostly choked' and 'Midlevel' were agregated into column 'Mid-Level', and setting 'Mostly Open' was agregated with 'Open'. See Figure 8and Figure 9, or Panel B Table 3. 29 The results on wood-stove emissions of PM2.5 are presented in the third, fourth, fifth and sixth rows of Table 6. Whereas the third and fourth rows presents results for the case of burning (certified) dry wood-fuel, the fifth and sixth rows present results for burning wood fuel with high moisture content. For the case of certified dry wood fuel, and given the actual setting of the wood-stove’s damper, the third row of Table 6 shows that average PM2.5 emissions for a wood stove like those used in Valdivia is 9.8 g/h. Furthermore, our results show a reduction in PM2.5 emissions drop of 1.4 g/h due to the information sign (fourth row total). This represents a 14.7 percent decrease in PM2.5 pollution emissions with respect to the baseline (absence of information sign). On the other hand, for the case of wood fuel with high moisture content, the fifth row of Table 6 shows that wood stove’s PM2.5 emissions amount to 21.0 g/h, and that the information sign induces a reduction in PM2.5 emissions of 3.9 g/h (sixth row total). This represents an 18.7 percent reduction with respect to the baseline for wood fuel with high moisture content. To obtain an overall effect for the effect of the information sign on PM2.5 emission reductions we weight these estimates by the percentage of consumption of certified dry wood fuel and wood fuel with high moisture content. According to survey data for households in Valdivia, about one third of households use certified dry wood fuel for heating, whereas about two-thirds use wood fuel with high moisture content (INFOR, 2015). Using these proportions to obtain a weighted estimate for the average household in Valdivia, we estimate that the information sign can reduce wood stove PM2.5 pollution emissions by 17.3 percent on average. 5.6 Effect on Indoor Concentrations of Fine Particulate Matter (PM2.5) In the previous section we considered the effect of the information sign on wood-stove outdoor air pollution emissions and assessed the change in these emissions that was induced by change in the damper setting. Once pollution from a given wood stove is emitted outdoors, those pollutants combine with those emitted by neighboring dwellings, making it difficult to establish a direct link between the emissions of an individual wood stove and ambient air pollution concentrations. As stated earlier, however, much of the pollution emissions that are emitted outdoors filter into dwellings through drafts entering doors, windows and ceilings. As individuals spend a large amount of their time indoors (particularly during the cold winter months), we are also interested in learning whether there is an effect of the information sign intervention on indoor 30 pollution concentrations. To do so, in this experiment we additionally measured indoor concentrations of particle pollution with the Speck sensor (the same that measures indoor temperature in the living room of each of the participating dwellings). Figure 10 below shows average indoor particle concentrations for each damper setting. The figure shows that households those that largely use the damper in the “Mostly Open” setting also experience lower indoor particle pollution. Figure 10. Average Indoor PM Pollution by Wood Stove Damper Setting Note: Blue bars denote average indoor pollution (fine particulate matter concentrations) for each damper setting. As noted above, settings are grouped into five categories (Choked, Mostly Choked, Mid-Level, Mostly Open, and Fully Open), and average indoor pollution is computed only for when the wood stoves are actually on. In addition, capped lines denote 95% confidence intervals (C.I.) for each damper setting. To examine whether the information sign effectively induced a change in indoor particle pollution concentrations, we run an ordinary least square regression for the following regression equation 𝐼𝐼𝑆𝑆𝑜𝑜𝐺𝐺𝐺𝐺𝐷𝐷𝐼𝐼𝑀𝑀 =𝛼𝛼0+𝛼𝛼1𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆+𝛼𝛼2𝑇𝑇𝐷𝐷𝐷𝐷𝐷𝐷𝑆𝑆𝐺𝐺𝐷𝐷 +𝛼𝛼3𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆_𝑇𝑇𝐷𝐷𝐺𝐺𝐷𝐷 +𝛿𝛿𝑋𝑋 +𝜀𝜀 (3) where 𝐼𝐼𝑆𝑆𝑜𝑜𝐺𝐺𝐺𝐺𝐷𝐷𝐼𝐼𝑀𝑀 refers to the particle count of indoor air pollution and the rest of the variables are the same as in equations (1) and (2) above, with 𝜀𝜀 denoting the error term. 31 The average indoor pollution for each of the groups is given by: (a) Treatment group, before sign = 𝛼𝛼0+𝛼𝛼2 ; (b) Treatment group, after sign = 𝛼𝛼0+𝛼𝛼1+𝛼𝛼2+𝛼𝛼3 ; (c) Control group, before sign = 𝛼𝛼0 ; and (d) Control Group, after sign = 𝛼𝛼0+𝛼𝛼1. Therefore, the Diff-in-Diff estimate [(b-a)-(d-c)] is captured by the parameter 𝛼𝛼3. Table 7 below presents parameter estimates for the 𝛼𝛼’s in equation (3) above, both without further controls for household characteristics (column 1) and adding those controls (column 2). Results from Table 7 show that the information sign had no significant effect on indoor particulate pollution of participating households. This is result is not surprising. As stated above, outdoor pollution from surrounding emitting sources mixes outside with that of other emitters (neighbors’ wood stoves) and then enters indoors through drafts. In this work we did not attempt to model and directly account for the mixing of ambient air pollutants. In addition, as stated earlier, the wood stoves of those participating households have a sealed combustion chamber that is designed to leak little air pollution inside the dwelling. Because of this, we did not expect to find significant effects of the information sign on indoor particle pollution. Table 7. Parameter Estimates for OLS Regression Dependent Variable: Particle Pollution Sign On is a dummy variable that takes on value equal to 1 when the information sign is on, and 0 otherwise. Similarly, Treatment Group takes on value equal to 1 when the household belongs to the treatment group, and 0 otherwise. Furthermore, Sign On & Treatment Gr. takes on value 1 when the household is in the treatment group and the information sign is on, and 0 otherwise. Column (1) presents results without controlling for observable household characteristics (such as those from Table 1), whereas column (2) explicitly accounts for household characteristics. VARIABLES (1) (2) Sign On -4.117 -3.754 (2.654) (2.500) Treatment Group -2.838 -1.451 (4.966) (5.791) Sign On & Treatment Gr. 0.274 0.372 (3.056) (2.818) Constant 33.09*** 56.20** (4.019) (21.42) Observations 98,880 98,880 R-squared 0.004 0.016 Standard errors clustered at the household level. *** p<0.01, ** p<0.05, * p<0.1 32 6. Concluding Remarks and Policy Recommendations To address the problem of high concentrations of ambient air pollution in cities in south-central Chile, the environmental authority has pushed for a large-scale wood stove replacement program. Under this program, eligible households receive a free wood stove with state-of-the-art combustion technology so long as they effectively dispose of their old (dirty) wood stoves. These new wood stoves have a market value starting at about USD 200. In addition, the only new wood stoves now available are wood stoves with the same combustion technologies as those under the government replacement program. However, because of the way consumers use these new wood stoves, reductions in pollution emissions from upgrading to these new wood-stoves have not been realized. Unless users change the way they operate their woodstoves, upgrading to new wood stoves alone cannot effectively address the problem of high ambient air pollution in this region. In this work we assessed an information sign that aims to induce a behavioral change in the way users operate their wood stoves, so that to reduce air pollution emissions. We found that reductions in wood stoves’ pollution emissions vary from 14.7 to 18.7 percent, depending on the type of wood fuel that is burned (either certified-dry or high-moisture wood fuel), and a weighted average yields reduction of 17.3 percent. Current government policies have focused on inducing users to choose cleaner wood fuels, such as certified-dry wood fuel and pellets. However, as mentioned earlier, these policies have not proven very successful at tackling the problem of air pollution in cities in this region. We believe that a policy measure that attaches an information sign to existing wood stoves can be complementary to current efforts by Chile’s environmental authorities. We see a big potential in reduction of wood-stove pollution emissions from a large-scale program that nudges wood stove users into changing the way they operate their devices. For example, at the end of the intervention we asked in our survey whether users would be willing to choke their wood stoves less frequently in order to reduce their wood stoves’ pollution emissions. Only 7.4 percent of those in the control group and 26 percent of those in the treatment group said they would. However, when asked whether they would choke their wood stoves less frequently if there were an educational campaign to educate and motivate the community, 88 percent of respondents in both groups said they would. Therefore, the data and analysis conducted in this work allow for a better understanding of users’ behavior regarding actual 33 operation of their wood stoves. This can be used for predicting realistic scenarios of the impact of a similar large-scale intervention that provides wood stove users with an information sign similar to the one assessed in this work. Moreover, the information sign evaluated in this work costs only a small fraction of the cost of these new wood stoves, at about USD 3 (plus costs of installation and information flyer). We believe that a large-scale intervention has real potential to tackle a substantial portion of the problem of air pollution in cities of south-central Chile. It is simple and low-cost so that it can be scaled up in a relatively short period of time and at little additional cost. Moreover, it can be complementary to other existing policies and programs. Significant gains in reducing air pollution emissions could additionally be achieved by a large-scale communications campaign that teaches users that fully choking their wood stoves is highly polluting and informs them that they can considerably reduce their pollution emissions by simply adjusting the damper setting away from fully closed. Chile’s Ministry of Energy is currently undertaking communications campaigns in all cities suffering from wood stove air pollution, and adding this layer should be quite affordable. Indeed, this sort of communications campaign is currently for a small subgroup of users, low-income elderly residents, in a nearby city (Temuco). Furthermore, as the government (Ministry of Environment) dictates the technology and pollution emission standards of the wood stoves currently available in the market, it could simply mandate that all new wood stoves come with a built-in information sign similar to the one assessed in this study. This should only slightly increase manufacturers’ costs and could prove to be a cost-effective policy measure. 34 References Bernstein, J.A. et al. 2008. “The Health Effects of Nonindustrial Indoor Air Pollution.” Journal of Allergy and Clinical Immunology 121(3): 585-591. Bustamante, W. et al. 2009. “Eficiencia Energética en la Vivienda Social, un Desafío Posible.” In: Camino al Bicentenario: Propuestas para Chile. Santiago, Chile: Gobierno de Chile. Castillo, C. 2001. Estadística Climatología. Tomo II. Santiago, Chile: Dirección Meteorológica de Chile, Climatología y Meteorología Aplicada. Chávez, C., W. Gómez and S. Briceño. 2009. “Costo-efectividad de Instrumentos Económicos para el Control de la Contaminación: El Caso del Uso de Leña.” Cuadernos de Economía 46(134): 197-224. Chávez, C. et al. 2011. Diseño, Implementación y Evaluación deun Programa Piloto de Recambio de Actuales Tecnologías Residenciales de Combustión a Leña por Tecnologías Mejoradas, en las Comunas de Temuco y Padre Las Casas.”” Informe Final. Chávez, C., J. Stranlund and W. Gómez. 2011. “Controlling Urban Air Pollution Caused by Households: Uncertainty, Prices, and Income.” Journal of Environmental Management 92(10): 2746-2753. Chow, J.C., and J.G. Watson. 1998. “Guideline on Speciated Particulate Monitoring.” Report prepared for US Environmental Protection Agency, Research Triangle Park, NC, by Desert Research Institute, Reno, NV (1998). Available at: https://www3.epa.gov/ttn/amtic/files/ambient/pm25/spec/drispec.pdf CDT (Corporación de Desarrollo Tecnológico de la Cámara Chilena de la Construcción). 2010. Estudio de Usos Finales y Curva de la Oferta de Conservación de la Energía en el Sector Residencial de Chile. Santiago, Chile: CDT. CONAMA (Corporación Nacional de Medio Ambiente). 2007. “Analisis Técnico-Económico de la Aplicación de una Norma de Emisión para Artefactos de Uso Residencial Que Combustionan con Leña y Otros Combustibles de Biomasa.” Santiago, Chile: Ambiente Consultores. Delmas, M.A., M. Fischlein and O.I. Asensio. 2013. “Information Strategies and Energy Conservation Behavior: A Meta-Analysis of Experimental Studies from 1975 to 2012.” Energy Policy 61:729-739. 35 Díaz-Robles, L.A. 2014. “Investigación y Generación de Factores de Emisión de Contaminantes Atmosféricos para Artefactos Residenciales que Combustionan Biomasa de Relevancia Nacional.” Temuco, Chile: Universidad Católica de Temuco (UCT). Egan, D. 1975. Concepts in Thermal Comfort. Upper Saddle River, United States: Prentice Hall. EPA (Environmental Protection Agency of the United States). 2016. National Ambient Air Quality Standards (NAAQS). Available at: https://www.epa.gov/laws-regulations/summary-clean-air-act Faruqui, A., S. Sergici and A. Sharif. 2010. “The Impact of Informational Feedback on Energy Consumption—A Survey of the Experimental Evidence.” Energy 35(4):1598-1608. Fischer, C. 2008. “Feedback on Household Electricity Consumption: A Tool for Saving Energy?” Energy Efficiency 1(1): 79-104. Fuenzalida Díaz, M., M. Miranda Ferrada and V. Cobs Muñoz. 2013. “Análisis Exploratorio de Datos Espaciales Aplicado a MP10 y Admisión Hospitalaria: Evidencia para Areas Urbanas Chilenas Contaminadas por Humo de Leña.” Geografía y Sistemas de Información Geográfica 5(5, Sección I): 109-128. Gómez, C., S. Yep and C. Chávez, 2013. “Subsidios a Hogares para Inducir Adopción de Tecnologías de Combustión de Leña Más eficiente y Menos Contaminantes: Simulación para el Caso de Temuco y Padre Las Casas.” Estudios de Economía 40(1): 21-52. Gómez, W. et al. 2014. “Using Stated Preference Methods to Design Cost-Effective Subsidy Programs to Induce Technology Adoption: An Application to a Stove Program in Southern Chile.” Journal of Environmental Management 132: 346-357. Gómez, W. et al. 2017. “Lessons from a Pilot Program to Induce Stove Replacements in Chile: Design, Implementation and Evaluation.” Environmental Research Letters (12): 11. Gómez-Lobo, A. et al. 2006. “Diagnóstico del Mercado de la Leña en Chile. Informe Final Preparado para la Comisión Nacional de Energía de Chile.” Santiago, Chiles: Centro Micro Datos, Universidad de Chile, Departamento de Economía, Centro Micro Datos. Available at: http://www.sinia.cl/1292/articles-50791_informe_final.pdf. INFOR (Instituto Forestal). 2015. “Encuesta Residencial Urbana sobre Consumo de Energía, Uso de Combustibles Derivados de la Madera, Estado Higrotérmico de las Viviendas y calefacción en las ciudades de Valdivia, La Unión y Panguipulli.” Santiago, Chile, 36 Instituto Forestal, Observatorio de los Combustibles Derivados de la Madera. Unpublished database. Jaime, M., C. Chávez and W. Gómez. 2017. “Fuel Choices and Fuelwood Use for Residential Heating and Cooking in Urban Areas of South-Central Chile: The Role of Prices, Income, and the Availability of Energy Sources and Technology.” Working Paper 201706. Caracas, Venezuela: Corporación Andina de Fomento, Development Bank of Latin America. Jordan, T.B., and A.J. Seen. 2005. “Effect of Airflow Setting on the Organic Composition of Wood Heater Emissions.” Environmental Science and Technology 10: 3601-3610. Available at http://www.ncbi.nlm.nih.gov/pubmed/15952364 Miller, S., and J.C. Ruiz-Tagle. 2018a. “Air Pollution and Premature Mortality in Emerging Economies: Evidence from Santiago, Chile.” College Park, Maryland, United States: University of Maryland. Manuscript. Available at: https://sites.google.com/prod/view/cristobalruiztagle/research Miller, S., and J.C. Ruiz-Tagle. 2018b. “Adverse Effects of Air Pollution on Probability of Stillbirth: Evidence from South-Central Chile,” College Park, Maryland, United States: University of Maryland. Manuscript. Available at: https://sites.google.com/prod/view/cristobalruiztagle/research. MMA (Ministerio de Medio Ambiente). 2011. “Establece Norma Primaria de Calidad Ambiental para Material Particulado Fino Respirable MP2,5.” Santiago, Chile: MMA. Available at: http://www.mma.gob.cl/transparencia/mma/doc/D12.pdf MMA (Ministerio del Medio Ambiente). 2014a. “Planes Descontaminación Atmosférica. Estrategia Nacional 2014 – 2018.” Santiago, Chile: MMA. MMA (Ministerio del Medio Ambiente). 2014b. “Revisa Norma de Emisión de Material Particulado para los Artefactos que Combustionen o Puedan Combustionar Leña y Derivados de la Madera, Contenida en el Decreto N° 39 de 2011. Santiago, Chile: MMA. Price, M.K. 2014. “Using Field Experiments to Address Environmental Externalities and Resource Scarcity: Major Lessons Learned and New Directions for Future Research.” Oxford Review of Economic Policy 30(4):621-638. 37 Reyes, R. 2017. “Consumo de Combustibles Derivados de la Madera y Transición Energética en la Región de Los Ríos, Periodo 1991-2014.” Informes Técnicos BES 3(6). Santiago, Chile: Instituto Forestal. Ruiz-Tagle, J.C. 2018. “Air Pollution and Urgent Care Visits: Estimation of a Causal Relationship Using Exogenous Variation in Concentrations of Fine Particulate Matter (PM2.5) in Santiago, Chile.” College Park, Maryland, United States: University of Maryland. Manuscript. Available at https://sites.google.com/prod/view/cristobalruiztagle/research. Schueftan, A., and A. González. 2013. “Reduction of Firewood Consumption by Households in South-Central Chile Associated with Energy Efficiency Programs.” Energy Policy 63(1): 823–832. Schueftan, A., and A. González. 2015. “Proposals to Enhance Thermal Efficiency Programs and Air Pollution Control in South-Central Chile.” Energy Policy 79(1): 48-57. Schueftan, A., J. Sommerhoff and A. González. 2016. “Firewood Demand and Energy Policy in South-Central Chile.” Energy for Sustainable Development 33: 26-15. SINCA (Sistema de Información Nacional de Calidad del Aire). 2017. Santiago, Chile: Ministerio del Medio Ambiente de Chile. http://sinca.mma.gob.cl Tiefenbeck, V. et al. 2016. “Overcoming Salience Bias: How Real-Time Feedback Fosters Resource Conservation.” Management Science 64 3):1458-1476. UFRO-CONAMA. 2009. “Diseño de un Programa de Recambio de Artefactos que Combustionan Leña por Tecnología Menos Contaminantes, en las Comunas de Temuco y Padre Las Casas.” Temuco, Chile: Universidad de la Frontera and Comisión Nacional del Medio Ambiente. WHO (World Health Organization). 1987. “Health Impact of Low Indoor Temperatures. Report on a WHO Meeting: Copenhagen, 11-14 November 1985.” Environmental Health Series 16. Copenhagen, Denmark: World Health Organization, Regional Office for Europe. WHO (World Health Organization). 2005. WHO Air Quality Guidelines for Particulate Matter, Ozone, Nitrogen Dioxide and Sulfur Dioxide. Global Update 2005. Available at: http://apps.who.int/iris/bitstream/10665/69477/1/WHO_SDE_PHE_OEH_06.02_eng.pdf