Widespread Misstatements of Emissions by Major U.S. Corporations

Widespread Misstatements of Emissions by Major U.S. Corporations | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Social Sciences - Article Widespread Misstatements of Emissions by Major U.S. Corporations Lauren Cohen, Ethan Rouen, Kunal Sachdeva This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5640071/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In the United States, corporations were directly responsible for more than 30 percent of the country’s 2022 total emissions, or approximately 2 billion metric tons of CO2 equivalents (1). Worried in part about the potential regulatory threats that companies face, investors and asset managers have called for greater transparency of publicly traded firms’ climate impacts. These calls have sparked a debate about whether government agencies should require such disclosures, or whether market pressures are sufficient to drive disclosure necessary for investors to make informed decisions (2). We show that these disclosures are, in fact, unreliable. Examining 900 emissions disclosures from 276 of the largest U.S. firms with available data, we find that 60% of these disclosures are eventually restated by the firm, a rate of restatement that has been consistent during the sample period. Furthermore, firms are more likely to understate than overstate, and the value of the understated emissions, on average, is more than twice the value of overstated emissions, suggesting that at least some of the misstatements are not errors. During the decade that we examine, the total gross understated public-firm emissions is 265 million tons, equivalent to the total emissions of Taiwan in 2020. Our results demonstrate the failure of market forces to produce reliable corporate environmental data. Earth and environmental sciences/Environmental social sciences/Energy and society/Energy economics Earth and environmental sciences/Environmental social sciences/Energy and society/Political economy of energy Scientific community and society/Social sciences/Carbon and energy/Energy modelling Scientific community and society/Social sciences/Climate change/Projection and prediction Figures Figure 1 Figure 2 Introduction Corporations are among the largest contributors to global greenhouse gas emissions. In the United States, the industrial sector (defined as the producer of goods and raw materials) was responsible for 30 percent of the country’s 2022 total emissions, or approximately 2 billion metric tons of CO2 equivalents ( 1) . Worried in part about the potential regulatory threats that companies face, investors and asset managers have called for greater transparency of publicly traded firms’ climate impacts. These calls have sparked a debate about whether government agencies should require such disclosures, or whether market pressures are sufficient to drive disclosure necessary for investors to make informed decisions (2) . To address investor demand, standard setting bodies such as the Securities and Exchange Commission in the United States and the International Accounting Standards Board are actively contemplating the mandating of climate disclosures such as firms’ greenhouse gas (GHG) emissions since policies to reduce emissions cannot be developed without credible data (3) . Still, some argue that market forces will be sufficient to pressure firms to act on environmental issues - including reporting the relevant information - obviating the need for regulation (4) . It is possible that market pressure will indeed lead firms to disclose all important information, as the top-performing companies will do so, prompting others to follow to avoid signaling themselves as lower quality—a process known as unraveling ( 5-6) . We conduct the first large-scale examination of reported scope 1 emissions for a sizeable set of the largest and most sophisticated corporations in the world. This analysis paints a clear — but bleak — picture of the voluntary disclosure mechanism in this landscape. We examine the CSR reports of 610 of the largest public firms and find that 276 report historical emissions data at least once from 2010 to 2024, a total of 1,800 unique disclosure years. We find that, across more than a decade of data, the majority (60%) of originally reported scope 1 emissions are changed in the future. In addition, the evidence suggests that at least some of these misstatements are intentional, with understatements (i.e., initially reporting a lower emissions number and increasing it) greater than overstatements in both frequency and magnitude. That this inaccurate information is so widespread and that the accuracy of the reported information does not appear to be meaningfully improving over time suggests that the current disclosure regime is insufficient, and, more importantly, that corporate contributions to global GHG emissions are distorted when incorporating this sizeable — yet misreported — piece. Data Our analysis is based on a hand collected sample of all Corporate Social Responsibility (CSR) Reports for 610 firms included in the Standard & Poor’s 500 index at least once from 2010 to September 2020. We require that the report contain a total scope 1 emissions number for the year of the report, and we require that this emissions number is reported at least once in the future in order to identify misstatements. These reports are voluntary, although there has been a rapid increase in their publication among this sample, with more than 85% of firms publishing one by 2020 (7) . This sample is also one that is ideal to study when examining whether market forces have led to effective environmental disclosures. These companies are among the largest in the world with sophisticated shareholders and reporting mechanisms to communicate financially material information to their stakeholders. While almost all of these firms voluntarily disclosed a CSR report in recent years, only 20% of all other publicly traded companies did so (8) . This group also is among the largest emitters of GHGs. Results The focus of our analysis is firms’ scope 1 emissions reporting behavior. These emissions are among the easiest to measure with established best practices. In addition, scope 1 emissions is a common data point, and it is unlikely that firms would report other climate-related data without reporting scope 1 emissions. In the United States and much of the world, the information in CSR reports remains unaudited with no enforcement body tasked with ensuring the accuracy of the information. Our analysis shows that these unaudited reports are rife with inaccuracies. From 2010 to 2020, 60% of all scope 1 emissions were restated by the company at some future date. For example, American Tower Corporation, a developer and operator of wireless and broadband infrastructure, in 2020 reported total scope 1 emissions of 596,969 metric tons of CO2e (Fig. S1). It restated its 2020 emissions in both 2021 and 2022, increasing it to 635,108 in 2021 and then 662,428 in 2022, while reporting the information in near-identical tabular format with no explanation for the restatement. While that represents an 11% increase in emissions, the absolute amount of underreported emissions, 65,459 tons, is equivalent to the annual emissions of 8,100 U.S. homes. The percentage of restatements has also been relatively stable over time, ranging between roughly 50-75% - and never being below 50% - suggesting that firms are not becoming more accurate in their measures over time (Fig. 1). To put these emissions restatements into perspective, during our sample period, restatements of financial performance metrics like revenue or net income—the most common indicators—occurred annually in less than 2% of public firms’ reports to the SEC. The magnitudes and direction of these restatements are also consequential and provide suggestive evidence that, to some extent, the misstatements are intentional. While it is impossible to identify the reasons for the changes, the accounting literature has examined the distribution of data and their deviation from a normal distribution as a way to highlight the possibility of intentional manipulation (9) . If the misstatements were made in error or changed solely for methodological reasons, it is likely that the distribution of the percent difference between the original and restated numbers would be symmetric around 0. This is not the case. The distribution of the logged difference between the last restated and original scope 1 number for all restatements is skewed to the right of 0, with the median slightly above 0 (Fig. S2). While the distribution shows that a majority of restatements are small, it also reveals that firms are more likely to understate their scope 1 emissions – 60% of all restatements are understatements. In addition, when firms understate, the understatements tend to be larger than overstatements. Of great concern is the distortion these restatements create for stakeholders seeking to understand corporations’ contributions to GHG emissions. The net aggregate amount of unreported emissions (i.e., the sum of the difference between restated and originally reported scope 1 numbers) from 2010 to 2020 is more than 150 million tons, but that disguises the magnitude of unreported emissions from those that underreport (Fig. 2). During the sample period, the total amount of scope 1 emissions that went underreported was 265 million tons, approximately equal to the total emissions of Taiwan in 2020. The total amount of emissions reductions due to initial overstatements was 115 tons, or less than half the value of understated emissions. It is important to reiterate that scope 1 emissions are among the easiest emissions-related measures that firms can quantify. Therefore, it is likely that these numbers represent a lower bound since there are no mechanisms requiring firms to systematically restate these numbers (i.e., there are likely to be unobservable misstatements), and they do not take into account harder-to-measure scope 2 and 3 emissions. In addition, our stringent data requirements exclude all firms that report scope 1 emissions in a given year but do not report that year’s emissions in future reports. Policy Implications & Conclusion This analysis reveals significant flaws in the current information environment related to firms’ environmental disclosures. It suggests that the information available to stakeholders and decision makers is inaccurate, and, despite the rapid increase in disclosure, market forces have failed to produce reliable information on corporate environmental impact. The persistence of these inaccuracies is particularly concerning given the rapid increase in the number of large companies that have set “net-zero” targets, which will be necessary to build lower-carbon economies, and yet depend on misstated emissions (10) . This concern is exacerbated by the fact that contracting on environmental performance is becoming an increasingly common practice. In addition to making public emissions-reduction commitments based on these data, firms are relying on environmental performance to write contracts with shareholders by tying executive compensation to emissions reduction targets and with lenders through green bonds that connect borrowing costs to environmental impact. The current system essentially asks firms to grade themselves on an honor system, with no explicit or perceived consequences for dishonesty within that reporting. Our analysis suggests a potential policy solution. Restatements of accounting information such as earnings and revenue are exceedingly rare, even though these measures are complex and require numerous estimates to calculate. The most glaring difference between these disclosures and scope 1 emissions is that financial reporting is regulated, and the consequences for misreporting are explicit and enforced. The long-term stakes of accurately reporting environmental impact are, arguably, far greater than that of precise quarterly or annual reporting of financial performance. Given the growing standardization of emissions measurement, it is feasible that firms’ environmental disclosures can achieve such levels of accuracy, too. The path forward, though, requires a system similar to that in place for financial reporting, with regulated, standardized, and transparent disclosures, backed up by consequential enforcement for misreporting. Methods Scope 1 source data Our source data are all available CSR reports for our sample, defined as all firms included at least once in the S&P 500 Index from 2010-2020. We began by downloading all CSR reports that were available on every company’s corporate website. We then augmented this effort by using the Internet Archive Wayback Machine to examine archived versions of firms’ websites for historical reports. This first effort was completed in August 2021. We then continuously monitored firms’ websites to identify and download new reports, and augmented that by contracting with SustainabilityReports.com to search for additional reports published after 2021. Ultimately, we ended up with 3,657 reports for 610 companies during our sample period. Although we cannot guarantee that the set of reports is comprehensive, it provides a significant panel of data that allows us to compare the scope 1 reporting practices of a large number of firms across more than a decade. Accounting restatement data Data on firms’ restatements of revenue and earnings are taken from Capital IQ. We identified all instances from 2010 to 2020 where firms’ originally reported earnings or revenue numbers differed from the currently reported number for a given year. Sample description Our sample consists of all 610 companies that were included in the S&P 500 Index at least once between 2010 and 2020. This sample is among the largest and most sophisticated firms in the world, meaning that they are the most likely to have the resources to measure and report accurate information on their environmental impact. We believe that, by analyzing these firms, as opposed to a broader sample, we are providing a conservative estimate of scope 1 misstatements as these firms have high institutional ownership and are heavily monitored by various stakeholders, in addition to being most likely to have the ability to accurately report. Identifying and measuring scope 1 restatements We take a multi-step approach to identifying and measuring scope 1 restatements. We begin by identifying all pages in all CSR reports that discuss firms’ scope 1 emissions. We then read those pages and identify only instances where firms report total scope 1 emissions for the year of the report. For each CSR report, we collect the scope 1 emissions numbers and the scale (e.g., “thousands of metric tons) for all reported years. We then convert all numbers to ensure that the scales are comparable. In order for a firm-year observation to be in our final sample, we require that the scope 1 emissions number for that year be reported in the given year and at least once in a future report. We next identify all instances where the reported emissions for a given year do not match those reported in another year. For example, in 2020 American Tower reported total scope 1 emissions of 596,969 metric tons of CO2e (Fig. S1). In 2021 and 2022, it increased the 2020 number to 635,108 and 662,428, respectively. To ensure accuracy, we conducted a second check of all identified restatements, returning to the source data to confirm that an emissions numbers do not match. This process, though, is not always straightforward and requires some judgement. We aimed to take as conservative approach as possible. Firms often restate emissions numbers multiple times across multiple years. For example, as shown in Fig. S3, AES, a utility and power generation company, in 2018 reported total scope 1 emissions of “51,878” for that year without any description of the scale of that number. In 2019, it increased the 2018 number to 54,154, while adding to the table that the scale of emissions is thousand metric tons. While that represents a 4.4% increase in emissions, the absolute amount of underreported emissions, 2.3 million tons, is equivalent to the annual emissions of nearly 500,000 cars. In 2020 and 2021, the last years in which it reported for 2018, AES reported 2018 scope 1 emissions of 50,291 thousand metric tons. For consistency, we define the restatement value as that in the last reporting year, so for AES, it would be 50,291. One issue we identified was instances where firms reported contradictory scope 1 emissions numbers in the same report. For example, Chipotle Mexican Grill reported scope 1 emissions in two locations within its 2022 report, and the numbers differed by 6,000 metric tons, 137,178 and 131,178 (Fig. S4). In instances where scope 1 emissions numbers differ within the same report, we report the number that most closely matches that reported in the report in the year immediately before or after the conflicted reporting. In 2023, Chipotle reported 2022 emissions of 131,178, so it was not identified as a restatement. Another issue was instances of obvious typographical errors. It is not uncommon for companies to report their emissions using incorrect scaling that could greatly increase the amount of restated emissions. For example, Kimberly-Clark’s 2013 report provided a table with its historical emissions and stated that they were reported in millions of metric tons of CO2e (Fig. S5). The number reported in the table for 2013 was “2,243,926.” If the scale was correct, that would mean that the company had scope 1 emissions of more than 2 trillion tons, more than the total estimated emissions since the beginning of the industrial revolution. To address such issues and ensure that we are providing a conservative estimate of the amount of under- and over-reported emissions, we winsorize the reported emissions numbers, as well as the difference between the restated and original numbers at the 1% and 99% levels. Declarations Acknowledgments: We acknowledge Robert Kaplan, George Serafeim, and participants of the Harvard Business School brownbag series for helpful feedback. Abhay Duggirala, Sam Karasik, and Dean Kim provided excellent research assistance. Funding: We acknowledge financial support from Harvard Business School and the Ross School of Business. Author contributions: All authors contributed equally to the work. Competing interests: Authors declare that they have no competing interests. Materials and Correspondence: All correspondences should be directed to Ethan Rouen. References United States Environmental Protection Agency, “Sources of Greenhouse Gas Emissions” (EPA, 2024); https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions. Securities and Exchange Commission, “SEC Adopts Rules to Enhance and Standardize Climate-Related Disclosures for Investors” (SEC, 2023); https://www.sec.gov/newsroom/press-releases/2024-31 Greenstone, C. Leuz, P. Breuer, Science 381, 837 (2023). Freeman, C.D. Kolstad, Ed. Moving to Markets in Environmental Regulation: Lessons from Twenty Years of Experince (Oxford University Press, Oxford, UK, 2007). Grossman, O. Hart, J. Finance. 35, 323 (1980). Milgrom, Bell J. of Econ. 12, 380 (1981). Rouen, K. Sachdeva, A. Yoon. “Sustainability Meets Substance: Evaluating ESG Reports in the Context of 10-Ks and Firm Performance” 2024; https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4227934. Bourveau, M. Chowdhury, A. Lee. E. Rouen, “Human Capital Disclosures” (2023); https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4138543. Burghstahler, I. Dichev. Earnings Management to Avoid Earnings Decreases and Losses. J. Account. Econ. 24, 99 (1997). Erb, B. Perciasepe, V. Radulovic, M. Niland, Corporate Climate Commitments: The Trend Towards Net Zero. In: Handbook of Climate Change Mitigation and Adaptation (Springer Nature, Switzerland, 2022). Additional Declarations There is NO Competing Interest. Supplementary Files ExtendedDataFigures.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5640071","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Social Sciences - Article","associatedPublications":[],"authors":[{"id":412748321,"identity":"f6c41b11-5045-4b0a-ad00-b5f5ca44e76d","order_by":0,"name":"Lauren Cohen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAp0lEQVRIiWNgGAWjYBACPgYGNiBlk2AA4vEQo4UNoiWNdC2HSdHCfvjYg487zueZ8x9gfPC2jRgtPGnphjPP3C62nJHAbDiXKC0SPGbSvG23EzfcYGADMojV8rftXOKG8wfYfxOvhbHtQOKGAwlszMRp4UlLk+xtS07cOSOxWXLOOSK08ANDTOJnm13idv7DBz+8KSNCCxJgbCBN/SgYBaNgFIwC3AAAgWgx9UHvmXUAAAAASUVORK5CYII=","orcid":"","institution":"Harvard Business School","correspondingAuthor":true,"prefix":"","firstName":"Lauren","middleName":"","lastName":"Cohen","suffix":""},{"id":412748322,"identity":"e6afa3c2-2a2c-4917-a06f-db90ab59e70d","order_by":1,"name":"Ethan Rouen","email":"","orcid":"","institution":"Harvard Business School","correspondingAuthor":false,"prefix":"","firstName":"Ethan","middleName":"","lastName":"Rouen","suffix":""},{"id":412748323,"identity":"ed935fc2-fd97-4b8d-ab34-9155aa421d88","order_by":2,"name":"Kunal Sachdeva","email":"","orcid":"","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Kunal","middleName":"","lastName":"Sachdeva","suffix":""}],"badges":[],"createdAt":"2024-12-13 18:30:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5640071/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5640071/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":75984555,"identity":"ec4f8860-a936-4a36-aef2-97a3c6010737","added_by":"auto","created_at":"2025-02-11 08:09:10","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":47242,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePercentage of firms restating scope 1 emissions and revenue or net income. \u003c/strong\u003eThe percentage of firms restating their scope 1 emissions (red line) and revenue or net income (blue line) are reported from 2010 to 2020.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5640071/v1/eccb9cfed5037eeb4b49b096.png"},{"id":75984554,"identity":"5c34cf5a-284d-4e29-a2ee-4964a07e4e38","added_by":"auto","created_at":"2025-02-11 08:09:10","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":186517,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistributions of unreported emissions.\u003c/strong\u003e The gross amount of emissions understated (red line) and overstated (blue dashed line) from 2010-2020. Flags for countries with comparable emissions in a given year (Norway-2013, Chile-2014, Belgium-2015, Philippines-2016, Netherlands-2017, Pakistan-2018, Spain-2019, and Taiwan-2020) are shown to provide a comparison.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5640071/v1/9305f2a9b00b886caf683886.png"},{"id":77884915,"identity":"d561b30c-58eb-4327-b89a-f7ed5b6a3abe","added_by":"auto","created_at":"2025-03-06 12:53:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":646185,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5640071/v1/6081609f-cb92-4fe4-9a98-0af79ba402d3.pdf"},{"id":75984556,"identity":"c194fd1b-14eb-4187-83d4-b48c99294470","added_by":"auto","created_at":"2025-02-11 08:09:11","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":825346,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-5640071/v1/bb2b0f2190a02a5e15489e5e.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Widespread Misstatements of Emissions by Major U.S. Corporations","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCorporations are among the largest contributors to global greenhouse gas emissions. In the United States, the industrial sector (defined as the producer of goods and raw materials) was responsible for 30 percent of the country\u0026rsquo;s 2022 total emissions, or approximately 2 billion metric tons of CO2 equivalents (\u003cem\u003e1)\u003c/em\u003e. Worried in part about the potential regulatory threats that companies face, investors and asset managers have called for greater transparency of publicly traded firms\u0026rsquo; climate impacts. These calls have sparked a debate about whether government agencies should require such disclosures, or whether market pressures are sufficient to drive disclosure necessary for investors to make informed decisions \u003cem\u003e(2)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eTo address investor demand, standard setting bodies such as the Securities and Exchange Commission in the United States and the International Accounting Standards Board are actively contemplating the mandating of climate disclosures such as firms\u0026rsquo; greenhouse gas (GHG) emissions since policies to reduce emissions cannot be developed without credible data \u003cem\u003e(3)\u003c/em\u003e. Still, some argue that market forces will be sufficient to pressure firms to act on environmental issues - including reporting the relevant information - obviating the need for regulation \u003cem\u003e(4)\u003c/em\u003e. It is possible that market pressure will indeed lead firms to disclose all important information, as the top-performing companies will do so, prompting others to follow to avoid signaling themselves as lower quality\u0026mdash;a process known as unraveling (\u003cem\u003e5-6)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eWe conduct the first large-scale examination of reported scope 1 emissions for a sizeable set of the largest and most sophisticated corporations in the world. This analysis paints a clear \u0026mdash; but bleak \u0026mdash; picture of the voluntary disclosure mechanism in this landscape. We examine the CSR reports of 610 of the largest public firms and find that 276 report historical emissions data at least once from 2010 to 2024, a total of 1,800 unique disclosure years. We find that, across more than a decade of data, the majority (60%) of originally reported scope 1 emissions are changed in the future. In addition, the evidence suggests that at least some of these misstatements are intentional, with understatements (i.e., initially reporting a lower emissions number and increasing it) greater than overstatements in both frequency and magnitude. That this inaccurate information is so widespread and that the accuracy of the reported information does not appear to be meaningfully improving over time suggests that the current disclosure regime is insufficient, and, more importantly, that corporate contributions to global GHG emissions are distorted when incorporating this sizeable \u0026mdash; yet misreported \u0026mdash; piece.\u0026nbsp;\u003c/p\u003e"},{"header":"Data","content":"\u003cp\u003eOur analysis is based on a hand collected sample of all Corporate Social Responsibility (CSR) Reports for 610 firms included in the Standard \u0026amp; Poor\u0026rsquo;s 500 index at least once from 2010 to September 2020. \u0026nbsp;We require that the report contain a total scope 1 emissions number for the year of the report, and we require that this emissions number is reported at least once in the future in order to identify misstatements.\u003c/p\u003e\n\u003cp\u003eThese reports are voluntary, although there has been a rapid increase in their publication among this sample, with more than 85% of firms publishing one by 2020 \u003cem\u003e(7)\u003c/em\u003e. This sample is also one that is ideal to study when examining whether market forces have led to effective environmental disclosures. These companies are among the largest in the world with sophisticated shareholders and reporting mechanisms to communicate financially material information to their stakeholders. While almost all of these firms voluntarily disclosed a CSR report in recent years, only 20% of all other publicly traded companies did so \u003cem\u003e(8)\u003c/em\u003e. This group also is among the largest emitters of GHGs.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe focus of our analysis is firms\u0026rsquo; scope 1 emissions reporting behavior. These emissions are among the easiest to measure with established best practices. In addition, scope 1 emissions is a common data point, and it is unlikely that firms would report other climate-related data without reporting scope 1 emissions.\u003c/p\u003e\n\u003cp\u003eIn the United States and much of the world, the information in CSR reports remains unaudited with no enforcement body tasked with ensuring the accuracy of the information. Our analysis shows that these unaudited reports are rife with inaccuracies. From 2010 to 2020, 60% of all scope 1 emissions were restated by the company at some future date. For example, American Tower Corporation, a developer and operator of wireless and broadband infrastructure, in 2020 reported total scope 1 emissions of 596,969 metric tons of CO2e (Fig. S1). It restated its 2020 emissions in both 2021 and 2022, increasing it to 635,108 in 2021 and then 662,428 in 2022, while reporting the information in near-identical tabular format with no explanation for the restatement. While that represents an 11% increase in emissions, the absolute amount of underreported emissions, 65,459 tons, is equivalent to the annual emissions of 8,100 U.S. homes.\u003c/p\u003e\n\u003cp\u003eThe percentage of restatements has also been relatively stable over time, ranging between roughly 50-75% - and never being below 50% - suggesting that firms are not becoming more accurate in their measures over time (Fig. 1). To put these emissions restatements into perspective, during our sample period, restatements of financial performance metrics like revenue or net income\u0026mdash;the most common indicators\u0026mdash;occurred annually in less than 2% of public firms\u0026rsquo; reports to the SEC.\u003c/p\u003e\n\u003cp\u003eThe magnitudes and direction of these restatements are also consequential and provide suggestive evidence that, to some extent, the misstatements are intentional. While it is impossible to identify the reasons for the changes, the accounting literature has examined the distribution of data and their deviation from a normal distribution as a way to highlight the possibility of intentional manipulation \u003cem\u003e(9)\u003c/em\u003e. If the misstatements were made in error or changed solely for methodological reasons, it is likely that the distribution of the percent difference between the original and restated numbers would be symmetric around 0. This is not the case. The distribution of the logged difference between the last restated and original scope 1 number for all restatements is skewed to the right of 0, with the median slightly above 0 (Fig. S2). While the distribution shows that a majority of restatements are small, it also reveals that firms are more likely to understate their scope 1 emissions \u0026ndash; 60% of all restatements are understatements. In addition, when firms understate, the understatements tend to be larger than overstatements.\u003c/p\u003e\n\u003cp\u003eOf great concern is the distortion these restatements create for stakeholders seeking to understand corporations\u0026rsquo; contributions to GHG emissions. The net aggregate amount of unreported emissions (i.e., the sum of the difference between restated and originally reported scope 1 numbers) from 2010 to 2020 is more than 150 million tons, but that disguises the magnitude of unreported emissions from those that underreport (Fig. 2). During the sample period, the total amount of scope 1 emissions that went underreported was 265 million tons, approximately equal to the total emissions of Taiwan in 2020. The total amount of emissions reductions due to initial overstatements was 115 tons, or less than half the value of understated emissions. It is important to reiterate that scope 1 emissions are among the easiest emissions-related measures that firms can quantify. Therefore, it is likely that these numbers represent a lower bound since there are no mechanisms requiring firms to systematically restate these numbers (i.e., there are likely to be unobservable misstatements), and they do not take into account harder-to-measure scope 2 and 3 emissions. In addition, our stringent data requirements exclude all firms that report scope 1 emissions in a given year but do not report that year\u0026rsquo;s emissions in future reports.\u003c/p\u003e"},{"header":"Policy Implications \u0026 Conclusion","content":"\u003cp\u003eThis analysis reveals significant flaws in the current information environment related to firms\u0026rsquo; environmental disclosures. It suggests that the information available to stakeholders and decision makers is inaccurate, and, despite the rapid increase in disclosure, market forces have failed to produce reliable information on corporate environmental impact. The persistence of these inaccuracies is particularly concerning given the rapid increase in the number of large companies that have set \u0026ldquo;net-zero\u0026rdquo; targets, which will be necessary to build lower-carbon economies, and yet depend on misstated emissions \u003cem\u003e(10)\u003c/em\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis concern is exacerbated by the fact that contracting on environmental performance is becoming an increasingly common practice. In addition to making public emissions-reduction commitments based on these data, firms are relying on environmental performance to write contracts with shareholders by tying executive compensation to emissions reduction targets and with lenders through green bonds that connect borrowing costs to environmental impact. The current system essentially asks firms to grade themselves on an honor system, with no explicit or perceived consequences for dishonesty within that reporting.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur analysis suggests a potential policy solution. Restatements of accounting information such as earnings and revenue are exceedingly rare, even though these measures are complex and require numerous estimates to calculate. The most glaring difference between these disclosures and scope 1 emissions is that financial reporting is regulated, and the consequences for misreporting are explicit and enforced.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe long-term stakes of accurately reporting environmental impact are, arguably, far greater than that of precise quarterly or annual reporting of financial performance. Given the growing standardization of emissions measurement, it is feasible that firms\u0026rsquo; environmental disclosures can achieve such levels of accuracy, too. The path forward, though, requires a system similar to that in place for financial reporting, with regulated, standardized, and transparent disclosures, backed up by consequential enforcement for misreporting.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eScope 1 source data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur source data are all available CSR reports for our sample, defined as all firms included at least once in the S\u0026amp;P 500 Index from 2010-2020. We began by downloading all CSR reports that were available on every company\u0026rsquo;s corporate website. We then augmented this effort by using the Internet Archive Wayback Machine to examine archived versions of firms\u0026rsquo; websites for historical reports. This first effort was completed in August 2021. We then continuously monitored firms\u0026rsquo; websites to identify and download new reports, and augmented that by contracting with SustainabilityReports.com to search for additional reports published after 2021.\u003c/p\u003e\n\u003cp\u003eUltimately, we ended up with 3,657 reports for 610 companies during our sample period. Although we cannot guarantee that the set of reports is comprehensive, it provides a significant panel of data that allows us to compare the scope 1 reporting practices of a large number of firms across more than a decade.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccounting restatement data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData on firms\u0026rsquo; restatements of revenue and earnings are taken from Capital IQ. We identified all instances from 2010 to 2020 where firms\u0026rsquo; originally reported earnings or revenue numbers differed from the currently reported number for a given year.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSample description\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur sample consists of all 610 companies that were included in the S\u0026amp;P 500 Index at least once between 2010 and 2020. This sample is among the largest and most sophisticated firms in the world, meaning that they are the most likely to have the resources to measure and report accurate information on their environmental impact. We believe that, by analyzing these firms, as opposed to a broader sample, we are providing a conservative estimate of scope 1 misstatements as these firms have high institutional ownership and are heavily monitored by various stakeholders, in addition to being most likely to have the ability to accurately report.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentifying and measuring scope 1 restatements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe take a multi-step approach to identifying and measuring scope 1 restatements. We begin by identifying all pages in all CSR reports that discuss firms\u0026rsquo; scope 1 emissions. We then read those pages and identify only instances where firms report total scope 1 emissions for the year of the report. For each CSR report, we collect the scope 1 emissions numbers and the scale (e.g., \u0026ldquo;thousands of metric tons) for all reported years. We then convert all numbers to ensure that the scales are comparable. In order for a firm-year observation to be in our final sample, we require that the scope 1 emissions number for that year be reported in the given year and at least once in a future report.\u003c/p\u003e\n\u003cp\u003eWe next identify all instances where the reported emissions for a given year do not match those reported in another year. For example, in 2020 American Tower reported total scope 1 emissions of 596,969 metric tons of CO2e (Fig. S1). In 2021 and 2022, it increased the 2020 number to 635,108 and 662,428, respectively. To ensure accuracy, we conducted a second check of all identified restatements, returning to the source data to confirm that an emissions numbers do not match.\u003c/p\u003e\n\u003cp\u003eThis process, though, is not always straightforward and requires some judgement. We aimed to take as conservative approach as possible. Firms often restate emissions numbers\u0026nbsp;multiple times across multiple years. For example, as shown in Fig. S3, AES, a utility and power generation company, in 2018 reported total scope 1 emissions of \u0026ldquo;51,878\u0026rdquo; for that year without any description of the scale of that number. In 2019, it increased the 2018 number to 54,154, while adding to the table that the scale of emissions is thousand metric tons. While that represents a 4.4% increase in emissions, the absolute amount of underreported emissions, 2.3 million tons, is equivalent to the annual emissions of nearly 500,000 cars. In 2020 and 2021, the last years in which it reported for 2018, AES reported 2018 scope 1 emissions of 50,291 thousand metric tons. For consistency, we define the restatement value as that in the last reporting year, so for AES, it would be 50,291.\u003c/p\u003e\n\u003cp\u003eOne issue we identified was instances where firms reported contradictory scope 1 emissions numbers in the same report. For example, Chipotle Mexican Grill reported scope 1 emissions in two locations within its 2022 report, and the numbers differed by 6,000 metric tons, 137,178 and 131,178 (Fig. S4). In instances where scope 1 emissions numbers differ within the same report, we report the number that most closely matches that reported in the report in the year immediately before or after the conflicted reporting. In 2023, Chipotle reported 2022 emissions of 131,178, so it was not identified as a restatement.\u003c/p\u003e\n\u003cp\u003eAnother issue was instances of obvious typographical errors. It is not uncommon for companies to report their emissions using incorrect scaling that could greatly increase the amount of restated emissions. For example, Kimberly-Clark\u0026rsquo;s 2013 report provided a table with its historical emissions and stated that they were reported in millions of metric tons of CO2e (Fig. S5). The number reported in the table for 2013 was \u0026ldquo;2,243,926.\u0026rdquo; If the scale was correct, that would mean that the company had scope 1 emissions of more than 2 trillion tons, more than the total estimated emissions since the beginning of the industrial revolution. To address such issues and ensure that we are providing a conservative estimate of the amount of under- and over-reported emissions, we winsorize the reported emissions numbers, as well as the difference between the restated and original numbers at the 1% and 99% levels.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u003c/strong\u003e We acknowledge Robert Kaplan, George Serafeim, and participants of the Harvard Business School brownbag series for helpful feedback. Abhay Duggirala, Sam Karasik, and Dean Kim provided excellent research assistance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e We acknowledge financial support from Harvard Business School and the Ross School of Business.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u0026nbsp;\u003c/strong\u003eAll authors contributed equally to the work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e Authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterials and Correspondence:\u003c/strong\u003e All correspondences should be directed to Ethan Rouen.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eUnited States Environmental Protection Agency, \u0026ldquo;Sources of Greenhouse Gas Emissions\u0026rdquo; (EPA, 2024); https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions.\u003c/li\u003e\n\u003cli\u003eSecurities and Exchange Commission, \u0026ldquo;SEC Adopts Rules to Enhance and Standardize Climate-Related Disclosures for Investors\u0026rdquo; (SEC, 2023); https://www.sec.gov/newsroom/press-releases/2024-31\u003c/li\u003e\n\u003cli\u003eGreenstone, C. Leuz, P. Breuer, \u003cem\u003eScience\u003c/em\u003e 381, 837 (2023).\u003c/li\u003e\n\u003cli\u003eFreeman, C.D. Kolstad, Ed. Moving to Markets in Environmental Regulation: Lessons from Twenty Years of Experince (Oxford University Press, Oxford, UK, 2007).\u003c/li\u003e\n\u003cli\u003eGrossman, O. Hart, \u003cem\u003eJ. Finance.\u003c/em\u003e 35, 323 (1980).\u003c/li\u003e\n\u003cli\u003eMilgrom, \u003cem\u003eBell J. of Econ.\u003c/em\u003e 12, 380 (1981).\u003c/li\u003e\n\u003cli\u003eRouen, K. Sachdeva, A. Yoon. \u0026ldquo;Sustainability Meets Substance: Evaluating ESG Reports in the Context of 10-Ks and Firm Performance\u0026rdquo; 2024; https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4227934.\u003c/li\u003e\n\u003cli\u003eBourveau, M. Chowdhury, A. Lee. E. Rouen, \u0026ldquo;Human Capital Disclosures\u0026rdquo; (2023); https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4138543.\u003c/li\u003e\n\u003cli\u003eBurghstahler, I. Dichev. Earnings Management to Avoid Earnings Decreases and Losses. J. Account. Econ. 24, 99 (1997).\u003c/li\u003e\n\u003cli\u003eErb, B. Perciasepe, V. Radulovic, M. Niland, Corporate Climate Commitments: The Trend Towards Net Zero. In: Handbook of Climate Change Mitigation and Adaptation (Springer Nature, Switzerland, 2022).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5640071/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5640071/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"In the United States, corporations were directly responsible for more than 30 percent of the country’s 2022 total emissions, or approximately 2 billion metric tons of CO2 equivalents (1). Worried in part about the potential regulatory threats that companies face, investors and asset managers have called for greater transparency of publicly traded firms’ climate impacts. These calls have sparked a debate about whether government agencies should require such disclosures, or whether market pressures are sufficient to drive disclosure necessary for investors to make informed decisions (2). We show that these disclosures are, in fact, unreliable. Examining 900 emissions disclosures from 276 of the largest U.S. firms with available data, we find that 60% of these disclosures are eventually restated by the firm, a rate of restatement that has been consistent during the sample period. Furthermore, firms are more likely to understate than overstate, and the value of the understated emissions, on average, is more than twice the value of overstated emissions, suggesting that at least some of the misstatements are not errors. During the decade that we examine, the total gross understated public-firm emissions is 265 million tons, equivalent to the total emissions of Taiwan in 2020. Our results demonstrate the failure of market forces to produce reliable corporate environmental data.","manuscriptTitle":"Widespread Misstatements of Emissions by Major U.S. Corporations","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-11 08:09:05","doi":"10.21203/rs.3.rs-5640071/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4a7891e1-ca70-430d-a5e3-5689084a604f","owner":[],"postedDate":"February 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":44018843,"name":"Earth and environmental sciences/Environmental social sciences/Energy and society/Energy economics"},{"id":44018844,"name":"Earth and environmental sciences/Environmental social sciences/Energy and society/Political economy of energy"},{"id":44018845,"name":"Scientific community and society/Social sciences/Carbon and energy/Energy modelling"},{"id":44018846,"name":"Scientific community and society/Social sciences/Climate change/Projection and prediction"}],"tags":[],"updatedAt":"2025-03-06T12:45:52+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-11 08:09:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5640071","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5640071","identity":"rs-5640071","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}