Summarizing the impacts of policies that limit the use of single-use plastic items: a systematic literature review

1.1 Background

Plastic provides a wide array of benefits to modern human life. Its lightweight, durable, and cost-effective properties bestow users with convenience, versatility, and economic attributes (Barbir et al. 2021). The world’s growing demand for plastic catalyzed its mass production, and more is made every year (Tan et al. 2021). However, plastic pollution is one of the major contributors to global environmental pollution. The World Bank estimates that 242 million tons of plastic waste are produced each year, and up to 12.7 million tons end up in the oceans (Kaza et al. 2018). Plastic can also degrade into microplastics, embed in the environment thereby polluting it, and biomagnify (Yates et al. 2021). Additionally, single-use plastic (SUP) utensils may release chemicals directly into food (Schweighuber et al. 2019). These phenomena can directly lead to negative consequences for human health, including gastrointestinal toxicity, liver toxicity, neurotoxicity, and reproductive toxicity (Chang et al. 2020).

Indirect health consequences may also arise from exposure to microplastics (Chang et al. 2020), or to pollution from other stages of the plastic’s life cycle, including greenhouse gas (GHG) emissions associated with its production (Major 2021; Tan et al. 2021), and air pollution associated with its incineration (UNEP 2019). The burden of arthropod-borne infectious diseases associated with plastic pollution has been identified in mostly developing countries, such as Malaysia, Indonesia and Thailand (Maquart et al. 2022). Notably, the extent of potential health impacts from plastic remains unclear (Yates et al. 2021). Therefore, researchers recommend the monitoring of plastic waste, reduction of plastic consumption (Palmas et al. 2022), and development of a circular economy (Agamuthu et al. 2019).

An estimated 33–50 % of plastic consumed is intended for single use (Excell et al. 2018; Tan et al. 2021). Several SUP items are termed problematic because they cannot be recycled, and are therefore more likely to exist as environmental waste (UNEP 2018a). Globally, up to five trillion SUP bags are consumed annually (UNEP n.d.). The European Union alone consumes two billion plastic takeaway containers and 36 billion straws per year (The Lancet Planetary Health 2018), while the USA consumes up to 500 million straws daily (Jonsson et al. 2021), and China about 46 billion straws per year (Qiu et al. 2021). Remarkably, G20 countries produce two-thirds of the world’s plastic waste (Fadeeva and van Berkel 2021). However, only about 9 % of all plastic waste gets recycled (Geyer et al. 2017).

Despite there not yet being a globally binding agreement on reducing plastic (Ortiz et al. 2020), governments around the world are proposing and implementing policies to mitigate the negative impacts of SUPs. Germany and Denmark were the first countries to implement a SUP reduction policy in 1991 and 1994, respectively, via a plastic bag tax (Xanthos and Walker 2017). By 2019, there were 739 national or subnational policies globally aimed to address plastic pollution, including to reduce initial plastic consumption (Karasik et al. 2020).

Despite the existance of many SUP policies, and the ongoing development of new ones, the level to which these policies are implemented, whether they achieve the desired outcome, and whether there are unintended consequences on a global level, are concepts not well outlined (Yates et al. 2021). These are important questions that researchers attempt to answer for other climate change mitigation policies as well, like the introduction of carbon taxes (Ambasta and Buonocore 2018).

A handful of reviews summarize SUP bag policies and their effectiveness in defined regions like three countries in Africa (Behuria 2021) or Europe (Kasidoni et al. 2015). However, many focus on qualitative impacts or the respondents’ perspectives on the policy, rather than environmental impact. Some also include voluntary policies, or those targeting other stages of the life cycle of plastic instead of end-user consumption. Two studies have conducted a systematic review on the effectiveness of policies limiting the use of SUP bags globally, however one focused only on bans (Muposhi et al. 2022), and the other only reviewed national-level policies (Adeyanju et al. 2021). Finally, one systematic review gathered information for several plastic items from several sources until 2018, aiming to provide an inventory of policies in all stages of the plastic’s life cycle. It did not explore global patterns of policy effectiveness specifically for policies preventing the use of SUPs, and did not discuss co-benefits (Diana et al. 2022; Karasik et al. 2020).

Many sustainable actions have the potential to provide a secondary gain, or co-benefit, to individuals or society. Some co-benefits have been identified for tax-based SUP bag policies, including revenue that is generated from sales, or an increased societal awareness of plastic pollution (Cornago et al. 2021). However, the level to which policies restricting SUP use result in co-benefits, thus potentially optimizing their uptake, needs to be better described. In contrast, there may also be trade-offs, or unintended consequences, through negative impacts on equity or consumption patterns.

This paper focuses on all preventive policies aimed at reducing consumption of SUPs, broadening the scope globally and aiming to gather evidence on co-benefits or unintended consequences associated with these policies.

2.1 Searching the databases

A literature search strategy was applied to four peer-reviewed literature databases and one grey literature database (Google) (Table 1).

The peer-reviewed search strategy was kept broad with only two key terms, plastic and policy. This was done in an effort to capture all possible relevant papers regardless of outcome.

In Medline, the following strategy was applied:

1. Ovid MEDLINE(R): (8,089 articles, July 11, 2022)

2. ALL <1980 to 2023>

1. ((plastic* or disposable or single-use) adj5 (single-use or straw* or bag or bags or stick* or ware)).mp.

2. exp Plastics/

3. 1 or 2

4. (policy or policies or regulation* or guideline* or legislation* or legal* or “ban” or “bans” or banning or banned or strategy).mp.

5. exp Policy/

6. 4 or 5

7. 3 and 6

The same search strategy was applied to the three other databases by adapting the format to their own algorithm (Appendix A, Supplementary Material).

The search strategy for grey literature used three key terms, including plastic, policy, and impact, and the first 10 pages were reviewed. Subsequent searches were done using four key terms: the three above plus each of the major problematic SUPs as defined by Canada’s new policy (Appendix B, Supplementary Material). The first five pages were reviewed for each problematic SUP. Terms were chosen and refined after undergoing the peer-reviewed search, identifying the most common and correct terms used.

Finally, reference searching was applied to review papers and other articles found relevant to the topic.

2.2 Selection criteria

Inclusion criteria:

• Papers presenting quantitative results from implementation of a SUP policy

• Papers including at least one environmental outcome

• Journal articles

• Policy reports

• All years of publication

Overall, a broad inclusion criterion was applied such that peer-reviewed journal articles, as well as grey literature articles and reports were included. As well, all years of publication were accepted, although a natural time limitation existed for when the first plastic policy was implemented. Finally, papers discussing an implemented SUP policy needed to include an outcome measure that demonstrates environmental impact.

• Exclusion criteria:

• Review papers or future modelling studies

• Editorials, opinion pieces, news reports

• Policy does not aim to prevent or reduce consumption of problematic SUPs

• Policy described is not governmental and enforceable

• Environmental impact is not measured as a change in consumption or a reasonable proxy

• Non-English texts

Review papers and future modelling studies were excluded as they do not provide primary, measured data. Similarly, report styles such as editorials, opinion pieces, and news reports present uncertainty regarding bias and the source of their data, and were therefore also excluded. Further, papers were excluded if the discussed policy did not aim to reduce consumption of problematic SUPs, such as recycling policies, or if the outcome did not measure plastic reduction through a reasonable proxy. To reduce confounders and ensure comparability between policies, policies that were not governmental and enforceable were excluded. Finally, this review did not include non-English texts due to resource constraints.

The search strategy and selection criteria were cross-referenced with Adeyanju et al. 2021, but were expanded and adapted to this review’s objectives.

2.3 Screening

Peer-reviewed database results were exported to Covidence (2022) by downloading ris files from each database. Duplicates were automatically deleted on Covidence, or manually marked as duplicates. Screening was done using three stages: title, then abstract, and finally full text when applicable.

EndNote (2013) was used to acquire full texts when possible, and missing texts were searched through subscribed databases at McMaster University Health Sciences Library, University of London Online Library and LSHTM Library. A message was sent to one author through ResearchGate to access a full text, with no response.

For papers included in the review using the selection criteria, data were then extracted using Covidence and a self-created list of headings. These included the type of policy, type of plastic it targets, where and when the policy was implemented, the change in consumption of the SUP, and other impacts (Appendix C, Supplementary Material).

Grey literature searches in Google were done manually and results were cross-referenced with Covidence for duplicates. Screening was done by title, followed by full webpages. Data were then extracted into a Microsoft Excel table using the same self-created headings in Covidence. This table was also used for all reference search papers. Finally, all results were combined into one table (Appendix C, Supplementary Material). Referencing was managed using Mendeley (2022) throughout this report.

2.4 Harmonization of data

Reasonable data manipulation was done to make results more comparable. The primary outcome was a percent reduction in the consumption of SUP items at the consumer end, either by item count or by weight. When not available, reasonable proxies were deemed as follows, in order of priority: percent reduction in retailers’ acquisition of items, percent reduction in manufacturing, percent reduction in household use, and percent reduction in litter/waste collected. Lastly, percent reduction as reported by surveyed individuals was included only if the question explicitly asked if the individual reduced consumption to zero, thereby using percent reduction of individuals using SUPs as proxy. Percent reduction was calculated in relative terms by comparing the current level of use with the level of use before the policy implementation. In scenarios where several datapoints were provided for level of use over time, additional datapoints were created under the same paper for different timeframes, resulting in a larger number (56) than the total number of studies (51). Of note, one policy in Nepal, representing a partial ban on specific thin plastic bags, showed an 80 % increase in plastic use by weight because individuals used thicker, non-banned plastic bag alternatives. This value is distinctly different than all other data points and may not truly represent an increase in SUP if the thicker plastic could be re-used. Therefore, it was removed as an outlier.

2.5 Data analysis

Descriptive analysis of the entire dataset was done using Microsoft Excel and Datawrapper (Datawrapper 2025).

Statistical analysis of comparable data (aggregated SUP bag policies) was done in Stata (https://www.stata.com/). Given the data’s skewed distribution, a test for comparison of medians was done using a Mann-Whitney test.

2.6 Critical appraisal

The peer-reviewed search identified cohort, mixed methods, and cross-sectional studies. Validated tools were used to appraise the quality of each study type, including the CASP Cohort Study Checklist (Critical Appraisal Skills Programme 2022) (Appendix D, Supplementary Material), Mixed Methods Appraisal Tool (Nha HONG et al. 2018) (Appendix E, Supplementary Material), and Appraisal tool for Cross-Sectional Studies (AXIS) (Downes et al. 2016) (Appendix F, Supplementary Material). Each question was regarded as equal in value and a point was given for each question the study scored a “yes” on (or the equivalent for low risk of bias). Questions seven and eight were omitted from the CASP Cohort Study Checklist as they were not easily applicable to the data and scoring them for risk of bias was not possible.

Grey literature was appraised using the Public Health Ontario guide to appraising grey literature (Public Health Ontario 2015) using the same scoring method (Appendix G, Supplementary Material). Data from the year 2000 or later were considered recent data for the relevant question. One peer-reviewed article was found through this search and was consequently appraised with AXIS. Reference search results were appraised accordingly, where one required a peer-reviewed appraisal tool, and the remainder qualified for the grey literature tool.

Studies that received a “yes” in more than three-quarters of the questions were identified as studies with low risk of bias. Those that received a “yes” in more than half and up to three-quarters of the questions were labelled as having a medium risk of bias. Those that only received a “yes” in half or less of the questions were labelled as having a high risk of bias. Studies of all quality were included in the review, to avoid potentially systematically excluding a certain pool of evidence.

2.7 Ethics

This project’s proposal was approved by the London School of Hygiene & Tropical Medicine Ethics Committee.

3.1 Overview of the available data

After accounting for duplicates, 10,650 titles were found through our search: 10,314 from peer-reviewed and 336 from grey literature searches. 362 articles were selected for full-text screening, 63 of which came from the grey literature. During the full-text screening, 334 papers were excluded, most commonly because the paper did not provide a measured change in SUP use or a reasonable proxy, or the policy studied was not specifically targeting the reduction in use of SUPs. Ultimately, 21 studies were selected from the peer-reviewed search, and seven from the grey literature search, in addition to 23 articles from reference searching – 22 of which were from the grey literature. Therefore, a total of 51 papers were identified for this review (Figure 1).

Figure 1:

PRISMA (preferred reporting items for systematic reviews and meta-analyses) search strategy flow diagram, combining peer-reviewed, grey literature, and reference searches. Adapted from Covidence (2022).

3.2 Descriptive analysis

The included papers analyzed policies from 21 different countries, and these policies were implemented between the year 1994 and 2020. One paper covered a policy addressing several SUP items, another one studied single-use straws, and the remaining 49 addressed single-use bags (Table 2). Of the policies introduced, 28 were implemented nationally, and 23 were sub-national – including at the level of the city, county, region, province, or state (Figure 2). 28 of the policies taxed a SUP item, 19 policies banned the item, three presented a combination of a tax and ban, and one introduced a default-choice modification scheme (Figure 3). European countries tended to provide evidence for the effectiveness of national-level, tax-based policies, while other regions were more diverse. However, most of Africa, Asia, and South America lacked evidence on the effectiveness of any policies (Figure 2, Figure 3).

Figure 2:

Geographical distribution of the available evidence on the effectiveness of policies that aim to reduce the consumption of SUPs by the level of policy. Countries shaded in blue introduced a national-level policy, countries shaded in yellow introduced a sub-national-level policy, and countries shaded in green have introduced both national and sub-national level policies. Created with Datawrapper 2025 . (Map lines delineate study areas and do not necessarily depict accepted national boundaries).

Figure 3:

Geographical distribution of the available evidence on the effectiveness of policies that aim to reduce the consumption of SUPs by the type of policy. Countries shaded in blue implemented a tax-based policy, countries shaded in red imposed a ban, and those shaded in purple have implemented various ban and tax-based policies. A striped pattern indicates a default choice modification policy, which was unique to the United States of America. Created with Datawrapper 2025 . (Map lines delineate study areas and do not necessarily depict accepted national boundaries).

Of the total papers, 23 were peer-reviewed and 28 were grey literature (Appendix C, Supplementary Material). Most papers had a low (n=26) or medium risk (n=21) of bias, with only four assessed as having high risk of bias (Table 2). When grouped by country income level, G20 countries had a higher proportion of evidence with low risk of bias when compared with non-G20 countries (65 % and 43 %, respectively). Of note, most of the evidence came from G20 countries (Figure 4). The largest amount of evidence covered SUP bag taxes, then SUP bag bans. There is a clear gap in evidence for most other SUPs, and more so in non-G20 countries (Figure 4).

Figure 4:

Heatmap displaying the frequency of evidence for the effectiveness of SUP reduction policies, by each SUP item, each type of policy, and whether it came from a G20 country or not (n=51).

The average reduction in use of all SUPs in response to all policies was 62 %, with a median of 66 % and a very wide range encompassing a minimum of 3 % and a maximum of 100 % (Figure 5).

Figure 5:

Percent reduction in SUP consumption due to all policies by location (n=56). * (orange) refers to straws, while the others refer to bags. Dashed line represents the overall mean reduction (62 %). One outlier removed.

3.3 Single-use plastic bags

The average reduction in consumption of SUP bags in response to a policy was 63 %. There was no apparent difference between the median effectiveness of bans (63 %) compared with taxes (64 %), nor was there a change in effectiveness with length of time passed since the policy’s implementation (Appendix H, Supplementary Material). Compared to policies implemented at a national level, those implemented sub-nationally had a 9 % higher median in reduction of use, although the difference did not reach statistical significance, and the data has a notably wide range (p=0.31, Figure 6).

Figure 6:

Box plots showing the reduction in consumption of SUP bags by the level of the policy (p=0.31). Blue (left) represents ban and tax-based policies at a national level, while red (right) represents those at a sub-national level (n=55). The horizonal line crossing each box represents the median.

Further, the median reduction in use of SUP bags from policies in G20 countries was 19 % higher than non-G20 countries (p=0.4, Figure 7). Similarly, the difference did not reach statistical significance and the dataset has a notably wide range.

Figure 7:

Box plots showing the reduction in consumption of SUP bags from policies implemented in G20 countries (purple, left) compared with non-G20 countries (orange, right). The horizonal line crossing each box represents the medians (p=0.4, n=55).

3.4 Other single-use plastics

It was not possible to conduct any quantitative analysis for plastics other than bags due to a lack of data. However, Belgium’s picnic tax included SUP utensils/table ware, and although not specifically quantified, consumption of utensils and table ware was reported to show no change as a result of the tax (Table 2). Tax revenue has increased since the policy’s implementation, but the items’ purchase price has increased concurrently, so consumption may have remained stable.

Additionally, San Luis Obispo, California, implemented a default-choice modification policy for SUP straws, which only provided straws to customers who requested them. Consequently, straw consumption declined by 32 % (Table 2).

4.1 Summary of results

Of the 51 papers identified, the overwhelming majority discussed SUP bag policies. Tax-based policies were most common, followed by ban-based policies, and most were effective in reducing consumption, with an average reduction of 62 %. There was no obvious difference in effect between bans and taxes, or with duration of time since policy implementation. Slightly more than half of the polices were implemented nationally, which appeared to be less effective than the remaining policies that were implemented sub-nationally (medians of 66 % and 75 %, respectively). Policies appeared to be more effective in G20 countries (75 %), compared with non-G20 countries (56 %). Although neither association reached statistical significance, the limitations of the available data – including small sample size and challenges with harmonization – suggest further investigations are needed to better quantify the association.

Finally, many co-benefits and unintended consequences were identified as a result of these policies. Co-benefits could indirectly improve human health; For example, reduced visible litter improves walkability of communities, while revenue collected from tax was sometimes donated to charity that supported environmental or human health causes. Other climate-friendly outcomes, like the estimated CO₂ emission reductions, could help mitigate climate change. However, 11 papers noted an increase in consumption of other disposable items like garbage bags. Additional negative impacts manifested on the plastic industry, equity, and even one report of possible increased intestinal infections attributable to unsanitary reusable bag use.

4.2 Implications of this work

The review showed that both tax- and ban-based policies were effective at reducing SUP bag consumption, a default-choice modification scheme reduced SUP straw consumption somewhat, and a tax did not seem to impact SUP utensil consumption. These findings may inform future policies, although the wide range in policies’ percent reductions highlights the need to consider other factors that influence the policies’ impact.

Policy implementation can be contrasted between top-down and bottom-up approaches. Top-down implementation perceives the policy process as a linear and rational approach, where policy is developed at higher levels of a hierarchy and communicated to the frontline for implementation (Buse et al. 2012). This creates a theoretical division between formulation and implementation and fails to consider the impact those implementing the policy have on the outcome. In contrast, bottom-up implementation recognizes that those in subordinate levels have a meaningful influence on shaping policy and its outcome from what was envisaged higher up (Buse et al. 2012). Naustdalslid (2014) explains that top-down processes were used to implement policies supporting a circular economy for plastic, which are based on scientific knowledge alone, lack public support and indicators for national monitoring of impact. Additionally, Muposhi et al. (2021) conclude that critical success factors for tax-based policy effectiveness include performing a baseline assessment of the plastic problem, stakeholder involvement and acceptance of the tax, establishing proper tax governance and continuous monitoring, and enforcing the tax. Therefore, local support and consumer behaviour are key factors for effective policies that aim to reduce consumption and may be better achieved through subnational policies. A report from Germany found that sub-national authorities implement policies better than national policies due to less vertical governance logistics and administrative burdens, better engagement with stakeholders at the implementation level, and better ex ante assessment of the situation (OECD 2010). Of note, this administrative burden is consistent with the unintended consequence found by Convery et al. regarding administrative costs in Ireland’s national policy (Table 2).

4.3 Strengths and limitations

This review followed a systematic approach and strategy, with multiple databases used, and produced a comprehensive output of relevant peer-reviewed articles. Although several review papers were identified for reference searching, only one new peer-reviewed paper was consequently added, and our Google search identified one additional peer-reviewed article. Further, this review employs an explicit methodology, and the search strategy is easily reproducible.

However, the grey literature search could be improved. Limited by time and resources, the search was made more specific to identify reports in the first 35 pages of Google hits. Although this search provided proportionally more articles that were ultimately included from the total hits when compared with the peer-reviewed search, the reference search identified 22 grey literature reports not otherwise found. Many of these reports would have possibly been found if the Google search was not limited to the first 35 pages. Further, in general, search engines are biased towards North America. To mitigate geographic bias, Google.com was used.

Another drawback is the language bias, as no resources were available for translation of non-English texts. The English search potentially neglected existing data in other languages, and the exclusion of non-English documents during screening removed several relevant papers, including some covering less frequently researched plastics like straws and food ware. This limitation was also noted by Diana et al. (2022).

Although a tool-based approach was used to assess the studies’ quality and risk of bias, the resulting categorization is only valid in context of the specific tool used and the cut-off scores decided.

Further, limitations in economics literacy led to the generalization of all tax policies under one category, although some taxes were applied at different points of the supply chain and may have impacted cost, price, supply, and demand differently.

Finally, the published evidence presented some strengths and limitations of our study design. First, reduction in plastic consumption is a more easily quantifiable outcome compared with other environmental policies that mitigate different outcomes, such as air pollution or GHG emissions, as it is less limited by technical knowledge or capacity. This allows a broader comparison for a better understanding of policy impacts across many countries. However, lack of data and heterogeneity of study designs, as well as the skew in literature towards plastic bag regulations, prevented further quantitative analysis, such as a meta-analysis. Additionally, it may not be a good indicator of the extent of the plastic industry’s influence on policy success. In fact, only 20 of the 56 data points in this review came from non-G20 countries despite them outnumbering G20 countries almost 10 to one.

4.4 Comparison with previous studies

Diana et al. (2022) similarly reviewed SUP policies globally. They found 40 policies, the majority of which similarly discussed SUP bags. However, their review was conducted in March 2018, and included voluntary and other policies. They graphed tax- and ban-based policies on SUP bags by percent reduction in consumption, but only for national-level policies with select subnational ones (total of 27 regions) (Diana et al. 2022). Their overall average reduction in SUP bag consumption was 66 %, which is similar to this review’s 63 % when removing the straw datapoint. They included fewer papers in their summary compared with this review, and they included three papers that did not meet inclusion criteria for this paper. They reached the same conclusion regarding the policies’ effectiveness as well as the indifference between tax- and ban-based ones (Diana et al. 2022). Additionally, they published a secondary report that mentions some of the unintended consequences identified in this review including the increased garbage bag consumption and pushback from the plastic industry (Karasik et al. 2020).

An OECD review similarly discussed SUP policies and their impacts but focused on OECD countries. It discovered more grey literature reviews by including non-English literature, and included policies targeting other stages in the plastic’s life cycle. It did not cumulatively quantify reduction of all policies, but it similarly concluded that market-based and regulatory (ban) policies, among others not relevant to enduser consumption, are effective in supporting the elimination of SUPs (Cornago et al. 2021). They discussed the need to incorporate other approaches and included qualitative context to explain policy effectiveness. For example, market-based policies’ effectiveness depended on consumers’ price elasticity of demand, while regulatory policies’ effectiveness depended on the alternatives available (Cornago et al. 2021). The report also described many co-benefits that are like the ones discovered in this review, including CO₂ emission reductions and charitable donations, and unintended consequences including an increase in garbage bag consumption, impacts on the plastic industry, and administrative costs (Cornago et al. 2021).

Muposhi et al. (2022) qualitatively summarized secondary impacts of SUP bag bans. Although they discovered similar litter reductions, they also found one paper that reported no significant global marine litter reduction. There is little global evidence quantifying the impact of these policies on plastic litter and it is therefore challenging to conclude definitive impacts of these policies on a macro scale. Further research should therefore aim to assess this trend globally.

4.5 Further research recommendations

The results highlight quite well the gaps in knowledge and need for further research. As evident by the heatmap (Figure 4), data are missing on all types of policies addressing other problematic SUPs including stir sticks, carrier rings, straws, and utensils/food ware. One report from the UK noted a 46 % reduction in overall problematic plastic consumption including the above items, however, it does not specify which policies were attributed to this change and some policies, like recycling, do not aim to prevent consumption (WRAP 2021). Additionally, as previously mentioned, there was more evidence from G20 countries, suggesting more evidence, or knowledge translation, is needed from non-G20 countries, especially in Africa, South America, and Asia. Ultimately, it is difficult to make global conclusions based on evidence from 21 countries alone.

Further research is also needed to better quantify the net plastic reduction considering the possible subsequent increase in other plastic use, its impact on plastic pollution, air pollution, and climate change, and assess how those impact human health. Research is consequently needed to quantify the net impact on human health accounting for co-benefits and unintended consequences summarized here. It is important that each region produces evidence that considers its local context. One approach may include performing more informed and comprehensive life cycle assessments prior to implementing the policy in each location. Of course, timeliness is critical with climate change and there is an argument to implore the precautionary principle. The precautionary principle is the idea that policy makers should make prudent policy decisions if there is no evidence available, in order to solve a problem that is potentially more harmful than the possible consequences of their decision (Ball 2006). Collaboration with policy makers, researchers, and industry will therefore be needed.