Chater & Loewenstein (C&L) make the provocative claim that “The impact of…i-frame interventions has been disappointing and can reduce support for much-needed systemic reforms” (target article, short abstract). However, there is no solid evidence for the claim that i-frame and s-frame interventions generally compete with each other. A recent Bayesian meta-analysis (Geiger, Brick, Nalborczyk, Bosshard, & Jostmann, Reference Geiger, Brick, Nalborczyk, Bosshard and Jostmann2021) concludes that there is no consistent overall spillover effect across behaviors or intentions for proenvironmental behaviors. In fact, countering the evidence for negative spillovers between proenvironmental behaviors selectively cited by the authors, positive spillover effects do also occur (e.g., Henn, Otto, & Kaiser, Reference Henn, Otto and Kaiser2020; Kumar, Caggiano, Cuite, Felder, & Shwom, Reference Kumar, Caggiano, Cuite, Felder and Shwom2023; Sparkman, Attari, & Weber, Reference Sparkman, Attari and Weber2021; Xu, Zhang, & Ling, Reference Xu, Zhang and Ling2018).
In this commentary, we offer two rebuttals: (1) i-Frame interventions can and do increase support for systemic reforms (Sunstein, Reference Sunstein2022); and (2) i-frame and s-frame interventions should be pursued in parallel to address key societal issues. Without accompanying i-frame interventions, many s-frame interventions are likely to fail because of low compliance, lack of enforcement, or reactance (Antinyan & Asatryan, Reference Antinyan and Asatryan2019; Carlsson, Gravert, Johansson-Stenman, & Kurz, Reference Carlsson, Gravert, Johansson-Stenman and Kurz2021; Nwafor et al., Reference Nwafor, Singh, Collier, DeLeon, Osborne and DeYoung2021; Proudfoot & Kay, Reference Proudfoot and Kay2014).
i-Frame interventions can enhance s-frame interventions in at least two ways. First, i-frame interventions can be productively employed to instigate s-frame changes. For example, reflecting on how one's proenvironmental behaviors are connected to one's values or identity increased support for a carbon tax policy (Sparkman et al., Reference Sparkman, Attari and Weber2021). Drawing attention to rising global temperature increased support for climate policy for liberal individuals (Luo & Zhao, Reference Luo and Zhao2019). i-Frame interventions can also trigger positive policy spillovers. For example, a small fee discouraged the use of single-use plastic bags and also increased public support for other environmental policies in the United Kingdom, such as adding changes for plastic bottles, excessive packaging, and fuel consumption (Thomas, Sautkina, Poortinga, Wolstenholme, & Whitmarsh, Reference Thomas, Sautkina, Poortinga, Wolstenholme and Whitmarsh2019). Three recent meta-analyses suggest that positive spillovers tend to occur more often than negative spillovers (Kumar et al., Reference Kumar, Caggiano, Cuite, Felder and Shwom2023), that positive spillovers tend to occur when i-frame interventions target intrinsic motivation or when the behaviors are similar to each other (Maki et al., Reference Maki, Carrico, Raimi, Truelove, Araujo and Yeung2019), or when i-frame interventions support personal autonomy, involve an explicit rationale explaining why the behavior is important, and address normative goals (environmental protection) or personal gain goals (financial savings; Geiger et al., Reference Geiger, Brick, Nalborczyk, Bosshard and Jostmann2021).
Second, i-frame interventions are uniquely well-suited to complement s-frame changes and may sometimes be necessary to ensure the success of systemic reforms. Many existing policies are ineffective because individuals fail to comply, or adequate enforcement is not feasible. Nudges and other i-frame interventions have been shown to increase policy efficacy in numerous domains, including tax compliance (Holz, List, Zentner, Cardoza, & Zentner, Reference Holz, List, Zentner, Cardoza and Zentner2020), public health (Krawiec, Piaskowska, Piesiewicz, & Białaszek, Reference Krawiec, Piaskowska, Piesiewicz and Białaszek2021), and environmental policy (Carlsson et al., Reference Carlsson, Gravert, Johansson-Stenman and Kurz2021). i-Frame interventions have been particularly effective in solving the so-called “last-mile” problem in public policy to overcome the intention–action gap, improve compliance, and reduce reactance to achieve policy targets (Soman, Reference Soman2015). For example, the City of Vancouver passed a bylaw in 2015 to ban food waste in garbage bins and provided residents with compost bins, but a significant amount of food waste still remains in garbage bins years later (Metro Vancouver, 2020). To address this last-mile problem, i-frame interventions such as making composting easier by moving the bins closer to people's doors (DiGiacomo et al., Reference DiGiacomo, Wu, Lenkic, Fraser, Zhao and Kingstone2018) or making the signage easier to read (Wu et al., Reference Wu, Lenkic, DiGiacomo, Cech, Zhao and Kingstone2018) can substantially increase composting rates. i-Frame interventions such as personalized information, messaging, and reminders have increased participation rates of low-income individuals in social policy programs, helping these policies realize their intended benefits (Despard, Roll, Grinstein-Weiss, Hardy, & Oliphant, Reference Despard, Roll, Grinstein-Weiss, Hardy and Oliphant2022; Hotard, Lawrence, Laitin, & Hainmueller, Reference Hotard, Lawrence, Laitin and Hainmueller2019; Manoli & Turner, Reference Manoli and Turner2016; Page, Castleman, & Meyer, Reference Page, Castleman and Meyer2020; Umaña, Olaniyan, Magnelia, & Coca, Reference Umaña, Olaniyan, Magnelia and Coca2022). i-Frame interventions can even mitigate reactance from people who are reluctant to comply with s-frame changes, using ideologically consistent frames (Bain, Hornsey, Bongiorno, & Jeffries, Reference Bain, Hornsey, Bongiorno and Jeffries2012) or messaging from in-group authority figures (Goldberg, Gustafson, Rosenthal, & Leiserowitz, Reference Goldberg, Gustafson, Rosenthal and Leiserowitz2021).
C&L raise potential pitfalls of i-framed interventions but neglect the broader picture of how i-frame and s-frame interventions work together. The field of behavioral science needs a better framework to outline the conditions under which negative and positive spillovers are likely to occur, as a recipe to design effective, complementary, and mutually reinforcing i-frame and s-frame interventions. As a start, we have proposed a unifying framework to account for positive and negative spillovers from an operant conditioning perspective (Zhao, Radke, Chen, Sachdeva, & Luo, Reference Zhao, Radke, Chen, Sachdeva and Luo2023). Specifically, we argue that positive spillovers occur because the previous behavior has been positively reinforced and generalized (e.g., by social or symbolic rewards, or identity reinforcers like warm glow), and negative spillovers occur because the previous behavior has not been positively reinforced. Negative spillover is especially likely if the previous behavior involves personal sacrifice (e.g., costs, efforts), which functions as a form of punishment that can lead to the extinction of the behavior and other similar behaviors. When the i-frame intervention leads to a behavior that feels rewarding (e.g., enhancing identity or values), it will likely lead to positive spillovers. When the i-frame intervention leads to a behavior that feels punishing (e.g., paying more for renewable energy) without positive reinforcement, it will lead to negative spillovers (e.g., less support for a carbon tax policy). This framework bridges a critical gap in the literature by highlighting the importance of reinforcement in generating spillovers. Indeed, some of the most promising i-frame interventions (e.g., social recognition) reinforce desirable behaviors via operant learning principles (Schneider & Sanguinetti, Reference Schneider and Sanguinetti2021). Effective i-frame interventions to create positive spillovers should introduce positive reinforcement to sustain a given behavior and to trigger other similar behaviors that are likely to be reinforced.
We urge behavioral scientists to continue developing and refining i-frame interventions that enhance, rather than compete with, s-frame interventions. Focusing solely on negative spillovers is counterproductive. We also encourage behavioral scientists to be more thoughtful in developing i-frame interventions by capturing spillover effects and unintended consequences. Only then can we gather a more comprehensive picture of human behavior change.
Chater & Loewenstein (C&L) make the provocative claim that “The impact of…i-frame interventions has been disappointing and can reduce support for much-needed systemic reforms” (target article, short abstract). However, there is no solid evidence for the claim that i-frame and s-frame interventions generally compete with each other. A recent Bayesian meta-analysis (Geiger, Brick, Nalborczyk, Bosshard, & Jostmann, Reference Geiger, Brick, Nalborczyk, Bosshard and Jostmann2021) concludes that there is no consistent overall spillover effect across behaviors or intentions for proenvironmental behaviors. In fact, countering the evidence for negative spillovers between proenvironmental behaviors selectively cited by the authors, positive spillover effects do also occur (e.g., Henn, Otto, & Kaiser, Reference Henn, Otto and Kaiser2020; Kumar, Caggiano, Cuite, Felder, & Shwom, Reference Kumar, Caggiano, Cuite, Felder and Shwom2023; Sparkman, Attari, & Weber, Reference Sparkman, Attari and Weber2021; Xu, Zhang, & Ling, Reference Xu, Zhang and Ling2018).
In this commentary, we offer two rebuttals: (1) i-Frame interventions can and do increase support for systemic reforms (Sunstein, Reference Sunstein2022); and (2) i-frame and s-frame interventions should be pursued in parallel to address key societal issues. Without accompanying i-frame interventions, many s-frame interventions are likely to fail because of low compliance, lack of enforcement, or reactance (Antinyan & Asatryan, Reference Antinyan and Asatryan2019; Carlsson, Gravert, Johansson-Stenman, & Kurz, Reference Carlsson, Gravert, Johansson-Stenman and Kurz2021; Nwafor et al., Reference Nwafor, Singh, Collier, DeLeon, Osborne and DeYoung2021; Proudfoot & Kay, Reference Proudfoot and Kay2014).
i-Frame interventions can enhance s-frame interventions in at least two ways. First, i-frame interventions can be productively employed to instigate s-frame changes. For example, reflecting on how one's proenvironmental behaviors are connected to one's values or identity increased support for a carbon tax policy (Sparkman et al., Reference Sparkman, Attari and Weber2021). Drawing attention to rising global temperature increased support for climate policy for liberal individuals (Luo & Zhao, Reference Luo and Zhao2019). i-Frame interventions can also trigger positive policy spillovers. For example, a small fee discouraged the use of single-use plastic bags and also increased public support for other environmental policies in the United Kingdom, such as adding changes for plastic bottles, excessive packaging, and fuel consumption (Thomas, Sautkina, Poortinga, Wolstenholme, & Whitmarsh, Reference Thomas, Sautkina, Poortinga, Wolstenholme and Whitmarsh2019). Three recent meta-analyses suggest that positive spillovers tend to occur more often than negative spillovers (Kumar et al., Reference Kumar, Caggiano, Cuite, Felder and Shwom2023), that positive spillovers tend to occur when i-frame interventions target intrinsic motivation or when the behaviors are similar to each other (Maki et al., Reference Maki, Carrico, Raimi, Truelove, Araujo and Yeung2019), or when i-frame interventions support personal autonomy, involve an explicit rationale explaining why the behavior is important, and address normative goals (environmental protection) or personal gain goals (financial savings; Geiger et al., Reference Geiger, Brick, Nalborczyk, Bosshard and Jostmann2021).
Second, i-frame interventions are uniquely well-suited to complement s-frame changes and may sometimes be necessary to ensure the success of systemic reforms. Many existing policies are ineffective because individuals fail to comply, or adequate enforcement is not feasible. Nudges and other i-frame interventions have been shown to increase policy efficacy in numerous domains, including tax compliance (Holz, List, Zentner, Cardoza, & Zentner, Reference Holz, List, Zentner, Cardoza and Zentner2020), public health (Krawiec, Piaskowska, Piesiewicz, & Białaszek, Reference Krawiec, Piaskowska, Piesiewicz and Białaszek2021), and environmental policy (Carlsson et al., Reference Carlsson, Gravert, Johansson-Stenman and Kurz2021). i-Frame interventions have been particularly effective in solving the so-called “last-mile” problem in public policy to overcome the intention–action gap, improve compliance, and reduce reactance to achieve policy targets (Soman, Reference Soman2015). For example, the City of Vancouver passed a bylaw in 2015 to ban food waste in garbage bins and provided residents with compost bins, but a significant amount of food waste still remains in garbage bins years later (Metro Vancouver, 2020). To address this last-mile problem, i-frame interventions such as making composting easier by moving the bins closer to people's doors (DiGiacomo et al., Reference DiGiacomo, Wu, Lenkic, Fraser, Zhao and Kingstone2018) or making the signage easier to read (Wu et al., Reference Wu, Lenkic, DiGiacomo, Cech, Zhao and Kingstone2018) can substantially increase composting rates. i-Frame interventions such as personalized information, messaging, and reminders have increased participation rates of low-income individuals in social policy programs, helping these policies realize their intended benefits (Despard, Roll, Grinstein-Weiss, Hardy, & Oliphant, Reference Despard, Roll, Grinstein-Weiss, Hardy and Oliphant2022; Hotard, Lawrence, Laitin, & Hainmueller, Reference Hotard, Lawrence, Laitin and Hainmueller2019; Manoli & Turner, Reference Manoli and Turner2016; Page, Castleman, & Meyer, Reference Page, Castleman and Meyer2020; Umaña, Olaniyan, Magnelia, & Coca, Reference Umaña, Olaniyan, Magnelia and Coca2022). i-Frame interventions can even mitigate reactance from people who are reluctant to comply with s-frame changes, using ideologically consistent frames (Bain, Hornsey, Bongiorno, & Jeffries, Reference Bain, Hornsey, Bongiorno and Jeffries2012) or messaging from in-group authority figures (Goldberg, Gustafson, Rosenthal, & Leiserowitz, Reference Goldberg, Gustafson, Rosenthal and Leiserowitz2021).
C&L raise potential pitfalls of i-framed interventions but neglect the broader picture of how i-frame and s-frame interventions work together. The field of behavioral science needs a better framework to outline the conditions under which negative and positive spillovers are likely to occur, as a recipe to design effective, complementary, and mutually reinforcing i-frame and s-frame interventions. As a start, we have proposed a unifying framework to account for positive and negative spillovers from an operant conditioning perspective (Zhao, Radke, Chen, Sachdeva, & Luo, Reference Zhao, Radke, Chen, Sachdeva and Luo2023). Specifically, we argue that positive spillovers occur because the previous behavior has been positively reinforced and generalized (e.g., by social or symbolic rewards, or identity reinforcers like warm glow), and negative spillovers occur because the previous behavior has not been positively reinforced. Negative spillover is especially likely if the previous behavior involves personal sacrifice (e.g., costs, efforts), which functions as a form of punishment that can lead to the extinction of the behavior and other similar behaviors. When the i-frame intervention leads to a behavior that feels rewarding (e.g., enhancing identity or values), it will likely lead to positive spillovers. When the i-frame intervention leads to a behavior that feels punishing (e.g., paying more for renewable energy) without positive reinforcement, it will lead to negative spillovers (e.g., less support for a carbon tax policy). This framework bridges a critical gap in the literature by highlighting the importance of reinforcement in generating spillovers. Indeed, some of the most promising i-frame interventions (e.g., social recognition) reinforce desirable behaviors via operant learning principles (Schneider & Sanguinetti, Reference Schneider and Sanguinetti2021). Effective i-frame interventions to create positive spillovers should introduce positive reinforcement to sustain a given behavior and to trigger other similar behaviors that are likely to be reinforced.
We urge behavioral scientists to continue developing and refining i-frame interventions that enhance, rather than compete with, s-frame interventions. Focusing solely on negative spillovers is counterproductive. We also encourage behavioral scientists to be more thoughtful in developing i-frame interventions by capturing spillover effects and unintended consequences. Only then can we gather a more comprehensive picture of human behavior change.
Financial support
Jiaying Zhao is supported by the Canada Research Chair program.
Competing interest
None.