Journal of Sustainable Marketing

Waste Management: Social Marketing Perspective

Prior insights on the effectiveness of waste management interventions are, at best, equivocal, although, in general, they show a positive correlation with recycling and sorting behaviors (Miafodzyeva & Brandt, 2013). Some recent studies (Tian et al., 2022; Ng, 2020) show that implementing a waste management policy promotes higher waste-sorting outcomes by enhancing users’ knowledge. While some studies find a positive association between informational and educational interventions and waste-sorting outcomes (Page & Sharp, 2012; Xia et al., 2023), some authors directly question the effectiveness of such interventions (Varotto & Spagnolli, 2017). For instance, in her meta-analysis, Bernstad (2014) found that informational interventions modestly increased waste sorting. In their meta-analysis, Osbaldiston & Schott (2012) show that convenience and financial incentives are the most effective interventions for increasing recycling and sorting. In contrast, informational and educational interventions have a smaller effect. Matiiuk & Liobikienė (2021) find that informational interventions did not significantly affect the level of waste sorting. Henceforth, based on the initial literature review, the findings regarding how informational and educational interventions affect waste sorting remain mixed.

Waste prevention, sorting, and recycling are “wicked problems” that often prompt decision-makers to use social marketing (Kim et al., 2019). Social marketing is a strategic method for fostering behavioral change and advancing social good by applying marketing principles beyond commercial settings (Issock Issock et al., 2021). It focuses on improving societal well-being through behavior change (Andreasen, 1994; Thaler & Helmig, 2013). Consequently, social marketing literature has examined how to promote positive waste-management behaviors, particularly through informational and educational interventions (Kim et al., 2019; Donovan, 2011).

However, many interventions fail due to a lack of rigor in meeting social marketing benchmark criteria (behavior, customer orientation, theory, insight, exchange, competition, segmentation, and methods mix), which directly affects their effectiveness (Kim et al., 2019; Kim et al., 2020; Sewak et al., 2021). Some social marketing literature argues that success isn’t solely about fulfilling social marketing benchmark criteria but also about producing social value (e.g., societal well-being via durable infrastructure-behavior fit, reduced environmental burden, reduced administrative frictions) through relationship-building and exchange (French & Russell-Bennett, 2015). Both the waste management and social marketing literature emphasize the need to focus on the broader context and system in which interventions operate (Gordon et al., 2016). This has led to calls for greater use of macro-social marketing, which aims for systemic, societal change rather than individual-level change (Truong, 2017; Kennedy, 2016). Macro-social marketing, often synonymous with systems social marketing, involves co-designing interventions with multiple stakeholders to address “wicked problems” like improper waste management (Flaherty et al., 2020). Macro-social marketing focuses on streamlining institutional settings and societal contexts to maximize the effectiveness of interventions (Kennedy & Parsons, 2012). It requires alignment between upstream, midstream, and downstream levels for greater effectiveness (Cherrier & Gurrieri, 2014; French & Russell-Bennett, 2015). Upstream and downstream actors in the social marketing system are interdependent, with actions at the downstream level influencing behavior at the upstream level (Newton et al., 2016).

Beyond macro-social marketing, literature indicates that other sustainable behavior frameworks, such as SHIFT emphasize levers at the individual and social interface (e.g., social proof, habit disruption, affect, tangibility; White et al. (2019)). These are necessary but not sufficient levers when upstream design imposes significant constraints on midstream and downstream actors. For instance, Andrade & Vieites (2025) synthesize cross-level obstacles and opportunities, complementing macro-social marketing system orientation.

However, centralized decision-makers often develop generalized policies without field-level insights, resulting in ineffective interventions (Schmidtke et al., 2021). Hierarchical structures within the social marketing system need further scrutiny to identify weaknesses and improve outcomes (Gordon et al., 2016). Behavioral change can be hindered by institutional and regulatory systems, inadequate infrastructure, and user behavior (Kemper & Ballantine, 2017). Policymakers use interventions to institutionalize behavioral change at the societal level, ensuring downstream actors have the resources and infrastructure needed for success (Brennan et al., 2014). Expanding interventions to consider social and institutional factors is recommended (Mathew et al., 2023). One solution is upstream social marketing, which influences behavior change at the highest decision-making levels by interacting with policymakers (Gordon, 2013). Upstream social marketing aims to shape policy through proactive engagement, informing policymakers of what works (French & Russell-Bennett, 2015). This approach is especially relevant to waste management systems (Kennedy & Smith, 2022; Al-mosa et al., 2017). The institutional setting plays a key role in waste prevention and sorting, with scholars calling for more innovative thinking (Trischler et al., 2019; Reschovsky & Stone, 2017). “Wicked problems” require proactive institutional involvement, and social marketing may help resolve these challenges. However, societal and political needs often conflict, leading to institutional resistance to change (Acemoglu & Robinson, 2008). Despite this, little research critically evaluates the institutional settings guiding social marketing interventions in waste management behaviors. This paper explores waste management interventions through the lens of institutional theory to address this gap.

An Institutional Perspective on Waste Management

Institutions are both formal and informal rules that organize social, political, and economic relations, shaping behavior through norms, regulations, and structures (Hodgson, 2006). Formal institutions are codified laws, contracts, and government structures, while informal institutions include customs, values, and beliefs (North, 1990). The regulatory pillar of institutions comprises laws and rules that constrain behavior, with this paper focusing on formal institutions. Formal institutions reassure societal members about the “rules of the game” and simplify behavior (North, 1990), but their persistence can be detrimental if they fail to adapt to societal changes (Thelen, 2009). Institutional rigidity refers to a stable but inflexible institutional setting that cannot meet evolving societal needs (Hodgson, 1989). This rigidity often stems from excessive formalization, overregulation, and bureaucratization (Andina-Díaz et al., 2021), leading to inefficiencies and stifling innovation (Gratton et al., 2021). Institutional rigidity can result from bureaucratic elites fearing the loss of power if change occurs, even though change might benefit society (Acemoglu & Robinson, 2006; Robinson & Acemoglu, 2006). Rigid institutions serve the interests of elites, creating a ”local maximum” that prevents optimal societal outcomes (Lustick, 2011).

In waste management, expert-driven policies often lack a user-centric approach, hindering innovation and positive behavior (Sewak et al., 2021). Similarly, rigidity can result from upstream actors’ fear of policy change and high perceived transaction costs (Khayame & Abdeljawad, 2020; Jagadale & Kemper, 2022). This study examines the EU's formal waste management framework, which sets measurable goals for waste prevention, sorting, recycling, and re-use by 2020, 2025, 2030, and 2035 (Parliament & Union}, 2008). Member States create plans, policies, and incentives to meet these goals within the EU framework.

Building on a macro-social marketing approach, we frame institutional rigidity as an upstream mechanism that reduces the efficiency with which upstream policy designs translate into downstream behavioral outcomes. We conceptualize a pathway in which institutional rigidity increases upstream-downstream misalignment, constrains co-design potential, segmentation, and adaptation to local infrastructure, thereby attenuating both the level and persistence of municipal waste prevention and sorting outcomes at the downstream level. Our mixed-methods evidence provides an empirical probe of this mechanism in a waste management intervention scaled at the national level. In Table 1, we showcase how macro-social marketing and its sub-dimensions operate across levels in the waste management context, mapping each to concrete manifestations in our study.

Research Context

Our study focuses on waste management in Croatia, an EU member state since 2013. Between 2018 and 2019, the Ministry of Economy and Sustainable Development (MESD) launched a public call for grants under the title “Implementing informational and educational programs for sustainable waste management.” The EU’s Cohesion Fund allocated €6.26 million for this initiative. Eligible applicants were local self-government units (cities or municipalities) with populations over 10,000, aiming to implement educational interventions for sustainable waste management. Both individual and consortium applications were accepted, and funding awards were determined based on compliance with the rules, the proposal’s quality, and available funds.

The funding call included mandatory and recommended interventions. Mandatory activities were: 1) leaflets, 2) brochures, 3) posters, 4) specialized radio and TV shows, 5) website creation, and 6) public roundtables. Recommended activities included: 1) workshops and events for children (school and kindergarten), 2) environmental celebrations, 3) media ads, 4) mobile apps, 5) internet banners, 6) paid ads, 7) contests, and 8) leaflets for tourists. Applicants had to choose a set number of mandatory and recommended activities based on population size. For example, municipalities with populations between 10,000 and 40,000 were required to select at least three mandatory and two recommended activities. These interventions were outlined in the Programme of Informational and Educational Activities for Sustainable Waste Management published by MESD.

Methodological Strategy

We applied a mixed-methods approach to gain a more comprehensive understanding of the effectiveness of informational and educational interventions for waste sorting and prevention (Tashakkori & Teddlie, 2010). The mixed-methods approach combines quantitative and qualitative research strategies to provide more detailed insights into the problem under investigation within a single study setting (Johnson & Onwuegbuzie, 2004). Compared to single-use approaches, the mixed-methods perspective is pragmatic and draws on both interpretivist and positivist approaches to research (Morgan, 2007). It enables higher data accuracy and allows for data triangulation. Consequently, researchers can gain a deeper understanding of the phenomena under investigation, thus increasing the validity of inference and making it more relevant for practitioners (Creswell & Plano Clark, 2007). We follow a procedure in which we conduct quantitative analysis (study 1) and then proceed to qualitative analysis (study 2) to gain a deeper understanding of the quantitative findings.

Data and Measures

Surveys have been commonly used to explore waste management behaviors, but are often criticized for lacking objectivity. Previous studies emphasize the importance of measurable behavioral outcomes in waste prevention (Kim et al., 2019; Rundle-Thiele et al., 2019). For this reason, we used secondary data on waste sorting and prevention outcomes. The MESD provides comprehensive waste-sorting statistics at the city and municipal levels, including total waste and sorted waste percentages. Data on the interventions were gathered through the Waste Prevention Portal (http://sprjecavanjeotpada.azo.hr/index.htm), which lists funded projects.

Outcome data are annual administrative statistics curated by MESD’s Directorate, standardized across years following national waste management reporting protocols. This allowed us to match project characteristics with waste sorting outcomes (prevention and sorting) from 2018-2021, addressing calls for a long-term perspective in evaluating pro-environmental interventions (Allcott & Rogers, 2014). Projects funded through earlier calls (e.g., 2016-2018) were excluded due to the limited number of funded projects, heterogeneity in calls, and incomplete documentation.

Table 2 provides critical information on these projects. Waste sorting data was sourced from the MESD’s Directorate for Environmental and Natural Protection (https://www.haop.hr/hr/tematska-podrucja/otpad-registri-oneciscavanja-i-ostali-sektorski-pritisci/gospodarenje-otpadom). Published yearly, this data includes municipal waste production and sorted waste percentages for each city and municipality in Croatia. Outcomes are annual administrative measures. Sorted waste percentage denotes the share of separately collected streams of waste in total municipal waste as reported by municipal systems (e.g., sorted plastics). We define short-term change as 2019–2018 and long-term change as 2021–2018.

Study 1: Evaluating the Impact of Interventions Through Propensity Score Matching

We used the PSM technique to evaluate the impact of interventions. The role of PSM is to determine the efficacy of the applied treatment and to help remove selection and endogeneity biases. PSM is an extensively used method to estimate treatment effects in observational settings by improving covariate balance between treated and control units (Rosenbaum & Rubin, 2022). Using PSM, researchers compare outcomes between a treatment and a control group by ensuring that the two groups have similar observable characteristics. In our case, using PSM, we can evaluate the effects of informational and educational interventions by treating the cities and municipalities that received funding as the treated group and those that did not as the control group.

4.1.1 Findings

In our dataset, 289 cities and municipalities implemented interventions, while 267 did not. Due to a limited pool for the control group (after matching and enforcing PSM using caliper and observables), we focused on 100 grant-leading cities/municipalities funded in 2018-2019 as the treatment group. Grant-leading municipalities directly select, budget, and implement activities and are the accountable unit for outputs, whereas consortia members may share exposure without clear attribution.

Preselected covariates for matching included municipal waste volume, the percentage of waste sorted in the base year, and region (tourist or non-tourist). We used a Python-based PSM extension for SPSS that operates in two stages: calculating propensity scores for each unit and comparing outcomes (waste sorting increases) between matched cities/municipalities in the treatment and control groups.

The PSM procedure generated propensity scores via logistic regression, with a caliper width set to 0.1. Of the 100 cases, 85 were successfully matched to control-group counterparts, whereas 15 were not matched, primarily because they were located in large cities and therefore excluded. The logistic regression model achieved a mean accuracy of 77.1% in predicting group membership. The baseline characteristics of cities and municipalities across key covariates (size, waste-sorting status, and regional affiliation) before and after PSM are presented in Table 3. To validate, we box-plotted propensity score distributions before and after matching (see Figure 1).

Table 3. Baseline Characteristics for Cities and Municipalities Before and After PSM Application

Display Table

	Before PSM				After PSM
	Treatment (N = 100)	Control (N = 267)	t-test	p-value	Treatment (N = 85)	Control (N = 85)	t-test	p-value
Amount of municipal waste (base year = 2018)	6832,96	1332,98	$-3.88$	0.000	5018,75	2341,53	$-1.04$	0.298
Status of waste sorting (base year = 2018)	12,83%	5,83%	$-6.62$	0.000	12,09%	10,78%	$-0.77$	0.437
Regional affiliation	0.25	0.27	$-0.29$	0.766	0,42	0,38	$-0.94$	0.346

After applying PSM, differences between treatment and control groups on covariates were insignificant, suggesting satisfactory matching. To test differences in outcomes (short- and long-term waste sorting and prevention), we used the t-test for independent samples (see Table 4), comparing results before and after PSM to neutralize self-selection bias.

Table 4. Impact of Interventions on Waste Sorting Outcomes Before and After PSM Application

Display Table

	Before PSM				After PSM
	Treatment (N = 100)	Control (N = 267)	t-test	p-value	Treatment (N = 85)	Control (N = 85)	t-test	p-value
Short-term change in total waste (2019-2018)	$-145.31$	$-26.81$	2.87	0.004**	$-133.64$	$-33.95$	1.76	0.079
Long-term change in total waste (2021-2018)	$-1063.91$	$-245.46$	2.89	0.004**	$-858.16$	$-465.96$	0.74	0.456
Short-term change in waste sorting (2019-2018)	2.76%	1.06%	$-3.16$	0.002**	2.91%	1.08%	$-2.09$	0.038*
Long-term change in waste sorting (2021-2018)	8.83%	6.11%	2.54	0.011*	8.95%	6.04%	$-1.88$	0.061

Table 4 shows that before PSM, the treatment group performed better in both short- and long-term waste sorting and prevention. However, after PSM, the treatment group showed superior results only in short-term waste sorting. The impact on short-term waste prevention and long-term sorting was marginally significant, while long-term waste prevention was insignificant.

Post-hoc analysis

Beyond average effects, we also examined whether the composition of interventions within treated municipalities relates to persistence in waste-sorting outcomes. Specifically, the count (and share) of recommended activities (e.g., workshops in schools, digital media/banners/ads, contests) positively predicts long-term change in sorted waste percentage (2021–2018), while mandatory activities e.g., (e.g., leaflets, brochures, posters, project website, broadcast shows, public roundtables) shows negative associations with sorted waste percentage. The coefficient remains statistically significant with baseline covariates (the percentage of people with higher education degrees living in the city/municipality, the amount of budget spent on environmental management (county-level data), the level of economic development (GDP pc, county-level), the number of funds spent in the project and the project duration (see Table 5).

This pattern is consistent with the macro-social marketing mechanism: more flexible, insight-driven recommended activities enable local tailoring and upstream-downstream alignment, supporting persistence, whereas rigid mandatory items constrain the realization of social marketing benchmark criteria and attenuate long-run effects. We therefore turn to Study 2 to unpack how flexibility and local fit shape alignment and persistence.

Study 2: Reflection of Cities/Municipalities on Implemented Interventions and the Project Framework

We conducted follow-up interviews to explore reasons for the significant differences in effectiveness between mandatory and recommended interventions. We emailed 20 purposively selected municipalities (diverse baseline waste, sorting status, and region), and 11 project managers (one per municipality) eventually participated (55% response rate). They provided informed consent, and their responses were kept confidential. The interviews were held via videoconferencing and lasted about 30 minutes. All interviews were recorded, transcribed, and analyzed. Table 6 shows the informants' profiles, which reflect high diversity, ensuring theoretical representativeness. The semi-structured interview guide was derived from Study 1's patterns (attenuation in waste outcomes over time; heterogeneity in treatment groups based on the effectiveness of mandatory vs. recommended interventions). Two coders independently performed open coding, reconciled a shared codebook, and applied axial coding to aggregate themes; disagreements were resolved through adjudication. Coding was managed manually.

The informants were asked to assess the effectiveness of mandatory and recommended interventions in their areas and to share their experiences with the project's institutional setup, including the clarity of funding calls, project reporting, and flexibility. The data collection focused on two main questions, simplifying the analysis. We used thematic analysis (Clarke & Braun, 2016) to analyze the transcripts. The process began with initial coding, followed by reviewing and grouping the codes into broader themes. Two main themes emerged: 1) the effectiveness of interventions and 2) the role of the project funding framework, each with two sub-themes.

Theoretical and Policy Implications

Our study contributes to macro-social marketing literature by exploring the interactions between upstream, midstream, and downstream actors, emphasizing the importance of stakeholders in addressing “wicked problems” (Flaherty & Domegan, 2023; Dibb, 2014; Kemper & Ballantine, 2017; Kennedy et al., 2018; McHugh et al., 2018; Wymer, 2021). It extends prior macro-social marketing research in areas like health, well-being, education, and violence prevention (Bastos et al., 2021; Hamby et al., 2017; Huff et al., 2017; Truong, 2017). While previous studies focused on the role of upstream stakeholders in implementing programs, our research highlights how resource allocation at the upstream level of waste management policy can significantly affect intervention effectiveness at the downstream level. Our study bridges social marketing and waste management perspectives, revealing a lack of coherence in current waste management policies due to upstream actors' limited knowledge of social marketing benchmark criteria, limited downstream involvement in framework design, and rigidity in upstream policy frameworks.

Our findings provide key guidance for policymakers designing waste management interventions. First, they emphasize the need for co-designed interventions with users to ensure they meet the target population’s needs. Policymakers should involve users in the design process, fostering participatory approaches that enable users to help create interventions that promote behavioral change in their communities (Kim et al., 2019; Kubacki & Rundle-Thiele, 2016; Sewak et al., 2021). Formative research on users' attitudes, beliefs, and behaviors should be conducted before implementation to better inform intervention tailoring (Truong, 2017). Second, our study suggests that the problem may not lie in the interventions themselves but in the upstream framework that dictates which interventions should be implemented. Even well-designed social marketing interventions can fail if there’s misalignment between upstream and downstream elements of the macro-social marketing system. A potential recommendation for policymakers is to introduce more flexibility into the institutional framework governing waste management. This could involve evaluating interventions’ effectiveness at the institutional level, shifting interventions between mandatory and recommended categories, removing underperforming ones, and adding new suggestions. Our findings also question the value of the current institutional structure for waste management funding, proposing innovative solutions such as autonomous governance that allow users to design context-specific interventions (Mathew et al., 2023).

Limitations and Suggestions for Future Research

While our study provides valuable insights, it also has several limitations. First, our research focuses on a single case study from Croatia, meaning the results should be interpreted with contextual caution. Additionally, while PSM improves comparability between treated and control subjects in our study, the analysis relies on observable covariates available at the municipal and city levels. Data constraints limited the inclusion of additional characteristics such as income distribution at municipality/city level beyond waste volume and regional affiliation. As with all observational designs, unobserved heterogeneity cannot be fully ruled out, and estimated effects should be interpreted as associations under observed balance rather than definitive causal impacts. Second, the literature suggests that more holistic waste management programs should combine informational and educational interventions with other strategies (e.g., convenience and economic incentives) to achieve greater effectiveness (Kim et al., 2019). While our study supports the need for complementary interventions, further research is needed to explore this relationship in depth. Third, we only touched on potential imbalances between upstream social marketing and downstream social marketing. Future studies should collect data from downstream actors, such as waste management staff, community leaders, and residents, to conduct a comparative analysis of perspectives. This would help identify more nuanced barriers to effective intervention implementation and improve framework design. Finally, while we emphasize the importance of upstream social marketing for more effective waste management interventions, future research could assess upstream actors' knowledge and awareness of waste management strategies. Addressing the gap between their understanding and downstream realities could lead to targeted training or knowledge-sharing programs, ultimately bridging the gap between upstream goals and field-level outcomes.