The Growing Field of Nonprofit Accounting Research: 21st Century Data Sources, Topics, and Opportunities

G ordon, Greenlee, and Nitterhouse (1999) reported on a significant development that had the potential to broadly expand research in nonprofit accounting beyond what was, until that time, often difficult and daunting work. Financial data from Form 990, an annual filing submitted to the Internal Revenue Service (IRS) and the primary financial reporting mechanism for most nonprofit organizations, would now be available in digitized form, thanks to the National Center for Charitable Statistics (NCCS).¹ Heretofore, accounting researchers interested in examining nonprofit reports had to manually collect copies of 990 filings directly from nonprofits or state charity offices (often the office of the state Attorney General), then create a database on their own from the collected reports.

The data collection process was frustrating, time consuming, and laborious. As individual researchers/teams were required to create their own project-specific data files, there was no common dataset, collection efforts were duplicative, and the resulting datasets had limited years, sectors, and variables. According to Gordon et al., the NCCS's new digitized databases created an opportunity that would “enable researchers to perform analysis far beyond what is currently possible” (Gordon et al. 1999, 121) by “eliminating data access problems” (Gordon et al. 1999, 118). Additional good news was that the data would be available to academic researchers upon request without charge, so long as they included a description of their research study and agreed not to distribute the data or use it for additional projects.²

Since the 1999 publication of the Gordon et al. (1999) paper, additional sources of nonprofit data have become accessible, allowing scholars to broaden the overall scope of research questions examined and better understand the ramifications of a wide variety of factors on organizational performance, managers' behavior, and donor decision making. For example, many states make available healthcare organizations' reported information pertaining to patient satisfaction, charitable care, and governance, and nonprofit ratings agencies have undertaken efforts in recent years to assess organizations' transparency and ability to accomplish their stated mission.

This article examines nonprofit accounting research published in the two decades since Gordon et al. (1999, 118) first predicted a “real revolution” in oversight and examination of nonprofit organizations. We document trends in nonprofit accounting research spanning the past 21 years, specifically identifying data sources used, research topics investigated, and journals that published the studies. Our findings indicate that the availability of accessible and affordable digitized data opened the field to scholars who may have previously perceived data collection as a barrier to entry in the field. Nonprofit accounting research output has increased significantly and is now published broadly in accounting journals that are not focused specifically on nonprofit issues. As the field has developed, those who conduct nonprofit research have diversified the data sources they integrate into their research designs and the research methods they use. The field has grown and matured significantly over the past two decades, offering insights into many aspects of nonprofit performance—broadly defined to include organizational, financial, and management performance—and stakeholders' responses thereto.

In order to examine contributions to the nonprofit accounting literature, we compile a list of every nonprofit accounting paper published from 2000 to 2020 in a variety of journals. We start with six journals (listed alphabetically) that are almost unanimously considered to be the top-ranked accounting journals (Top 6): Accounting, Organizations and Society (AOS); The Accounting Review (TAR); Contemporary Accounting Research (CAR); Journal of Accounting and Economics (JAE); Journal of Accounting Research (JAR); and Review of Accounting Studies (RAST) (Moon, Summers, Waddoups, and Wood 2021). We also include all journals published by the American Accounting Association, both the association-wide general topics journals and all 14 section journals.

Finally, we include a selection of accounting and cross-disciplinary journals that have traditionally been open to nonprofit accounting research. Journal of Accounting Literature (JAL) welcomes literature reviews on any accounting topic. Journal of Accounting and Public Policy (JAPP) is a natural fit for nonprofit research as many nonprofit issues have a policy component. JAPP sponsored special issues in 2009 and forthcoming in 2022 dedicated exclusively to governmental and nonprofit papers, another indication that nonprofit research is important to the journal. Journal of Accounting, Auditing & Finance (JAAF) is an accounting journal that has also welcomed nonprofit accounting research. Journal of Public Budgeting, Accounting and Financial Management (JPBAFM) is described as an outlet for work in “the field of accounting, management and governance in organizations operating in the public square or public-private sphere.”³Financial Accountability & Management (FAM) is an international journal that is “interdisciplinary in approach” for studies “of all types of governmental and other nonprofit organizations and services.”⁴Journal of Business Ethics (JBE) is a multidisciplinary business journal open to any business-related study, including those focused on nonprofit organizations. Nonprofit and Voluntary Sector Quarterly (NVSQ) is a cross-disciplinary journal that focuses on “nonprofit, philanthropic, and civil society studies” in any discipline (business and non-business).⁵Research in Governmental and Nonprofit Accounting (RIGNA), last published in 2009, was a periodical devoted entirely to government and nonprofit accounting topics. See Appendix A for a list of all journals included in the publication search.

To identify articles that examine nonprofit accounting, the authors independently undertook the following protocols. First, we obtained the tables of contents for each issue of the 29 selected journals for the years 2000–2020 and (except for papers that are clearly unrelated to nonprofit accounting) read the title and abstract of each paper to determine if it is appropriate for inclusion in our list. Because some articles' titles and abstracts do not clearly indicate a nonprofit setting, we next conducted a series of keyword searches in our selected journals. For accounting journals (e.g., JAL, JAAF) we used the search terms nonprofit, nongovernmental, NGO, charity(ies), charitable, donor, donation, healthcare, hospital, and 990. For non-accounting journals (e.g., NVSQ) we searched for accounting, financial, audit, and 990. Once we completed the search process, we all reviewed and combined the separate lists to create a full listing of nonprofit accounting papers published in the selected journals during our examination period. Editors' comments, opinion columns, and book reviews are not included in our list of research papers. We exclude papers that do not examine issues associated with nonprofit organizations or the sector. Examples of excluded papers are studies that investigate corporate entities indirectly associated with healthcare (e.g., pharmaceutical companies, device manufacturers) rather than healthcare providers, research specifically focused on for-profit hospitals, studies of health benefits provided by employers, and papers that focus on governmental oversight agencies rather than the nonprofit organizations they regulate.

While JOGNA readers are interested in both nonprofit and governmental topics, this analysis is limited to nongovernmental, nonprofit (i.e., private charitable) organizations. The motivation for this study is to examine the state of nonprofit accounting research after digital databases became widely available, reducing the need for hand collection of data. There has been no similar development involving governmental accounting reports, and governmental accounting researchers still experience the need for laborious acquisition of individual financial reports and hand collection of governmental accounting data therein. We do, however, make two exceptions to include studies that examine public sector (governmental) entities: healthcare and higher education. Because hospitals and colleges/universities provide comparable services and experience similar reporting and governance challenges regardless of their organizational form (nonprofit or governmental), we include all studies that investigate accounting questions for healthcare or higher education organizations that do not operate for profit.

To show how readily accessible data are being used for research in the nonprofit accounting field, we ascertain the data source(s) used in each study. We then identify the accounting topic examined in order to catalog the breadth of the types of studies conducted. We partition the publications into two decades—2000 to 2010 and 2011 to 2020—to help detect notable changes in data sources or topics examined over the 21-year span.⁶ The Online Appendix provides a list of all published papers included in our analyses, including the assigned topics (see the link in Appendix C for the downloadable file).

Over the past two decades, nonprofit accounting researchers have enjoyed the proliferation of and access to new data like never before. While much of the prior century's accounting research involving nonprofit organizations focused on information obtained from individual IRS filings and/or hand-collected financial data, today there is great diversity in the sources and types of nonprofit data available to researchers and, in many cases, the data are freely available. From the expanded IRS Form 990 to healthcare filings and required audit reports, to ratings agency information and results from surveys and experiments, options abound for ways to answer many of the important research questions that nonprofit accounting scholars will ask as the field advances and the world around us continues to change.

As predicted in Gordon et al. (1999), data made available by the NCCS and others in recent years have allowed nonprofit researchers to perform analyses far beyond that which was possible a mere twenty, or even ten, years ago. In Table 1, we identify every data source included in the 447 nonprofit accounting papers published in the journals we selected and report the number of papers that use each data source. Research studies often obtain data from multiple sources (e.g., combining NCCS SOI, Core, and Digitized Data files with audit or healthcare data). When a paper in our sample uses multiple data sources, it is included in the table multiple times. Therefore, the column totals exceed the unique number of papers published in each time period. Our examination of the data sources used in nonprofit accounting papers published from 2000 to 2020 reveals some distinct and familiar data categories, plus a variety of additional sources.

Not surprisingly, the primary source of data used in nonprofit accounting research over the past two decades is IRS Form 990 (hereafter, 990). This is not different from the period before 2000; what has changed most dramatically is how nonprofit scholars can now obtain 990 data. Prior to the creation of the NCCS databases, nonprofit accounting researchers had to request 990s from state charity offices (often through their websites or by visiting state offices) or directly from nonprofit organizations in order to create a project-specific database. We identified 26 papers where the authors requested 990s directly from organizations, only four of which were published after 2010. Only 14 studies relied on 990s provided by states or obtained directly from the IRS during our sample period.

After the NCCS made 990 data available for research, nonprofit accounting researchers began to use these databases rather than create their own. NCCS databases include the Statistics of Income (SOI), Core and Digitized Data files, and the Business Master File (BMF).⁷ There are 61 papers (14 percent of all selected published papers) that used the SOI files, 16 that used the Core files, and 24 that obtained 990 data from the Digitized Data files. Published papers relying on the NCCS files increased substantially in the second decade of the 21st century compared to the first. In sum, the NCCS data files appear to have made an important contribution to the nonprofit accounting field.

More recently, other sources of 990 data have surfaced. GuideStar, a leading repository of nonprofit information, sells a 990 dataset similar to that provided by NCCS. Sixteen papers have used 990 data provided by GuideStar, primarily in the 2011–2020 timeframe. The E-Filer database, which is made available by the IRS and is now accessible through the IRS Tax Exempt Organization Search (TEOS), is a relatively new source of 990 data. In 2010, the IRS began accepting electronically filed (“e-filed”) Forms 990, 990-EZ, and 990-PF, then began making those electronic filings available as XML files through Amazon Web Services (AWS) in 2016. As of early 2022, the IRS is phasing out usage of AWS and moving access to the E-Filer database to its TEOS search engine.⁸ While this data can be difficult to access and use, its advantage is that it has nearly all fields from Form 990 accessible in machine-readable form.⁹ This data source was used in just four published studies in our sample, which is likely a function of it only recently being made available.

Healthcare research is a significant portion of the nonprofit accounting field, in no small part because of the size of the healthcare sector and its relevance to many facets of life. Another factor is the availability of data pertaining to a variety of areas, including governance, management, operations, and financial performance of healthcare organizations. The healthcare sector is highly regulated, and many states require healthcare facilities to report information that goes far beyond that which is typically contained in 990s or other financial disclosures. Additionally, there are several types of ownership structures in healthcare systems—including nonprofit, for-profit, and government-operated—which provides opportunities for scholars to examine healthcare-related research questions in a variety of contexts.

Healthcare-related data are available from a number of sources, and those too have changed over the past 21 years. Early in our examination period, healthcare accounting researchers often used reports filed in individual states or relied on private data sources. One source used exclusively in the 2000–2010 timeframe is Merritt System^®, a private credit and investment analysis database containing audited financial statement information, socio-economic information, and supplementary operational statistics for hospitals in all 50 states and Washington, D.C. Another data source is the American Hospital Association (AHA), which was more commonly used in the 2000–2010 period (seven papers) compared to the 2011–2020 period (only three papers). Some AHA data has recently become part of the widely used Wharton Research Data Services (WRDS) platform, with more than 40 years of hospital information now available through this system. For researchers who have access to WRDS, it is possible that the AHA will again be a relevant data source in the future.

Several states (in the U.S. and internationally) require healthcare organizations to report data that is then made available to the public, and these state hospital reports continue to be a source of data for healthcare research. There were 27 studies that relied on data from state filings, and the frequency was similar in both decades. California's Office of Statewide Health Planning and Development (OSHPD) is one of the most commonly cited data sources in healthcare accounting research. The state's size alone is good reason to use this data, but the OSHPD database provides an additional unique advantage in that the state has required healthcare organizations to report detailed governance information for years. Owner and board of director characteristics such as name, occupation, and overall board size are provided, along with owners' and board members' compensation and percent of ownership (if any).

In Texas, nonprofit hospitals must provide a minimum threshold of charitable care and community benefits in order to maintain their tax-exempt status, so they are required to submit an annual community benefit report. The Texas Health Care Information Collective (THCIC), established in 1995, provides the community benefit reports and other hospital data to the public. While not as extensive as the OSHPD database, THCIC reports do include common metrics like number of beds, utilization, bad debt charges, charitable care, and net and gross patient revenue. Many other states, including Arizona, Connecticut, Florida, Illinois, Oregon, Pennsylvania, and several non-U.S. governments, also collect and disseminate healthcare data, sometimes for a fee.

Recently, a more commonly used source of healthcare data is from the Centers for Medicare & Medicaid Services (CMS). CMS makes a large assortment of data available at their website, https://www.cms.gov/, from broad national health expenditures to specific zip code-level public use files and many things in between. As shown in Table 1, use of the CMS Medicare/Medicaid database increased in the 2011–2020 timeframe, from 1 to 11 papers. This may be due, at least in part, to the expansion of Medicaid in many states following the passage of the Patient Protection and Affordable Care Act (ACA) in 2010. The ACA requires more performance data about healthcare providers and suppliers to be provided to the public, and in 2011 CMS adjusted its Availability of Medicare Performance Measurement in response.¹⁰

Several nonprofit accounting studies combine 990 financial data with audit-related data from Single Audits (formerly known as OMB Circular A-133 audits). Single Audits are required for non-federal entities—including nonprofits, Native American tribes, and state and local governments—that expend more than $750,000 of federal funds in a given year. The audit report includes information about the auditor, auditee, type of audit, federal awards, and audit findings, and is available through either the Federal Audit Clearinghouse (FAC) A-133 database or Audit Analytics. Audit Analytics is a costly addition to the WRDS database while the FAC is free. Feng, Ling, Neely, and Roberts (2014) describe the FAC and Audit Analytics databases, noting that Audit Analytics includes cleansed auditor data from the FAC database.

Single Audit data are used in 22 nonprofit accounting papers that span our study's timeframe, 18 of which were published in the most recent decade, showing a sizable increase in interest in nonprofit audit research. Likely due to the additional cost, only 3 of the 22 audit papers in our sample use Audit Analytics to access Single Audit data.

Charity ratings agencies use 990 data to synthesize individual nonprofits' financial information into a performance score that can be a useful means of comparison for potential donors (Harris and Neely 2016). The three largest ratings agencies are Charity Navigator, the Better Business Bureau Wise Giving Alliance (BBB), and Charity Watch (formerly the American Institute of Philanthropy) (Lowell, Trelstad, and Meehan 2005). The agencies identify different metrics and use different rating scales: Charity Navigator gives an overall score of 0 to 4 stars based on separate “Financial Health” and “Accountability & Transparency” scores; BBB issues a Pass/Fail based on 20 standards for accountability; and Charity Watch issues a letter grade from A to F based on the program and fundraising efficiency ratios (Harris and Neely 2016). Researchers who incorporate charity ratings data into their analyses tend to use one or more of the ratings as a measure of organizational quality. Credit rating agencies, such as Moody's and S&P, provide ratings for bonds issued by charitable organizations (often hospitals). Studies using ratings data represent a relatively small portion of our overall sample with only nine total papers, but six of those were published in the most recent decade, which may signal an opportunity for additional studies that incorporate this data.

To access information that is relevant to nonprofit organizations but not included in 990s or audit reports, nonprofit accounting researchers often seek out other sources of information to use in combination with 990 data. From 2000 to 2020, 115 nonprofit accounting papers (26 percent of published papers) included at least one “other” archival source in their research design. This includes 66 papers over the past decade, suggesting that novel data from non-traditional sources can provide fruitful avenues for nonprofit research. Examples include college and university rankings, corporate giving directories, U.S. census and labor data, international audit data, state tax assessor data, news reports discovered through LexisNexis searches, and an online database from a national newspaper listing frauds in nonprofit organizations.

Many researchers use behavioral methods such as experiments, surveys, interviews, and case studies to gather data related to nonprofit accounting questions. These methods allow researchers to directly question nonprofit stakeholders, such as managers or donors, to obtain insight into how they prepare or use nonprofit accounting information. While surveys, interviews, and case studies are used with similar frequency throughout the timeframe we examine, experiments are becoming more common, with five experiments published in the early decade, and 15 published in the most recent decade. This trend may be enhanced in the future due to the creation and ease of using online survey distribution platforms such as Prolific and Amazon's Mechanical Turk (MTurk), which can host various types of experiments (e.g., timed, multi-round, open-access, invitation-only) and can connect researchers with participants who are pre-screened for eligibility criteria in areas such as education, occupation, geographic location, and demographic information.¹¹

Table 1 shows that however we access data, there is growth in the quantity of nonprofit accounting studies, with 197 papers published in the first decade of our sample period and 250 published in the second, an increase of 27 percent. Figure 1 shows a clear upward trend in nonprofit accounting publications over the past two decades, culminating in 40 publications in 2020. The increase is largely driven by papers that use 990 data, often paired with data from other sources such as healthcare, audit, or other archival databases. Studies using the NCCS SOI database increased from a total of 19 published papers from 2000 to 2010 to 42 published papers from 2011 to 2020. Studies using data obtained from GuideStar grew from 2 to 14 over the same timeframe. Medicare and Medicaid data available from CMS was used in 11 published papers from 2011 to 2020 compared to just 1 in the previous decade. Overall, the expansion of availability and variety of data bode well for the future of nonprofit accounting research.

Table 2 reports the various topics for nonprofit accounting papers published from 2000 to 2020, partitioned by decade. For each publication, we categorize the research topics into at least 1 of 17 broad categories. A single publication can address multiple topics, such as the impact of governance on fraud, executive compensation in higher education, or international healthcare. If a study addresses more than one topic, we include it in the count for each classification. For this reason, the column totals are greater than the number of unique publications.

Unsurprisingly, the two most common specific topics for nonprofit accounting papers across the entire examination period are healthcare and financial reporting and disclosure. Healthcare is a significant part of the nonprofit sector, and 141 papers (32 percent of all publications in our sample) involve an investigation of hospitals or other healthcare entities. Accounting research generally focuses on how various stakeholders use financial information, and the accounting research in nonprofit organizations is no exception. Financial reporting and disclosures accounted for 142 of the 447 published papers (32 percent) in our sample. There were 20 percent more papers addressing financial reporting and disclosure in the 2011–2020 period compared to the previous decade, likely due to the availability and accessibility of large databases with 990s recorded in digital format.¹²

Like financial reporting and disclosure, accounting studies in general, and nonprofit accounting studies specifically, often examine governance and managerial issues. During our sample period, there were 114 papers that focused on governance and accountability and 109 that included managerial concerns (26 percent and 24 percent, respectively, of our total sample).

The two topics that experienced the most growth from the first decade of our analysis to the second are governance and compensation. There were almost twice the number of studies examining either governance or compensation in the second decade compared to the first. The substantial increase is primarily due to new data that resulted from the 2008 expansion of Form 990, which involved the addition of questions and disclosures related to governance, management, and compensation. This new governance section includes over twenty questions about such items as the existence of certain policies (e.g., policies regarding whistleblowers or conflicts of interest) and details about board members and their relationships with key employees of the organization (Boland, Harris, Petrovits, and Yetman 2020). The new supplemental Schedule J provides details about the various forms of executive compensation.

A substantial area of nonprofit study involves organizations and/or stakeholders outside the United States. The international nonprofit studies in our sample (144 studies, representing 32 percent of published papers) examine charitable organizations and hospitals in a variety of countries, including Taiwan, Germany, Australia, Turkey, Ghana, and China. Often this research uses data from state regulatory filings. There is likely a great opportunity to continue examining international nonprofit organizations, especially where accounting standards and state support of nonprofits vary from those in the U.S.

The 27 percent increase in the number of papers in our sample from the first decade to the second includes a 58 percent uptick in the number of nonprofit accounting papers published in the Top 6 accounting journals.¹³ However, the representation of research topics included in the Top 6 journals differs somewhat from that of the sample as a whole. While 32 percent of our sample papers examine financial reporting and disclosures, only 8 percent of these papers are in Top 6 journals, representing 18 percent of the 67 nonprofit papers published in Top 6 journals during our sample period. Similarly, nearly one-third of our sample papers involve healthcare contexts, and only 17 percent of those papers are published in the highest-ranked journals. However, healthcare papers comprise 36 percent of all nonprofit papers in Top 6 journals from 2000 to 2021. Although only 7 percent of papers in our study address compensation issues, nearly half (43 percent) of those are in Top 6 journals, constituting 19 percent of the nonprofit papers published in Top 6 journals. Earnings management, including the management of spending ratios, is the subject of only 6 percent of our total sample, but 38 percent of the nonprofit earnings management papers published between 2000 and 2021 are in the highest-ranked accounting journals, representing 16 percent of our Top 6 sample. Although papers related to governance and accountability in nonprofits comprise 26 percent of our total sample, this is the most common topic for nonprofit papers in Top 6 journals, accounting for 42 percent of the total. Differences between the topics of published papers in our overall sample and those of papers in Top 6 journals may indicate that these journals are primarily open to traditional accounting topics with appeal across various organizational types, or perhaps that researchers who focus on these topics are more likely to submit their work to the Top 6 journals.

Figure 2 shows the topics of the papers in our sample categorized by the journals in which they appear. From 2000 to 2020, there were 447 nonprofit accounting papers published in a wide range of accounting and cross-disciplinary journals. Nonprofit research has been published in each of the Top 6 accounting journals, with a total of 67 nonprofit publications appearing in these journals over the past 21 years. Over one-third of these papers (25) are in Accounting, Organizations and Society, where the focus of nonprofit research is primarily on international healthcare, governance, and managerial topics. The Accounting Review is the Top 6 journal with second-most nonprofit publications, with its 20 papers being largely concentrated in the areas of compensation, earnings management, and other forms of financial reporting. The top-ranked journals most often publish articles investigating governance (28 papers), healthcare (24 papers), managerial issues (23 papers), compensation (13 papers), financial reporting and disclosures (12 papers), and earnings management and misreporting (11 papers), many of which have an international focus (24 papers).

Nonprofit accounting research is published in every AAA journal except the Journal of International Accounting Research. AAA has 14 section journals, several of which are relatively new. Six of the 14 AAA section journals were established and issued for the first time during our sample period. These include Journal of Legal Tax Research (JLTR) in 2003, Journal of Emerging Technologies in Accounting (JETA) in 2004, Current Issues in Auditing (Current Issues) in 2007, Journal of Governmental and Nonprofit Accounting (JOGNA) in 2012, and both the Journal of Forensic Accounting Research (Forensic) and Journal of Financial Reporting (JFR) in 2016. These new journals account for 33 publications (7 percent) in our sample, 18 of which are published in JOGNA, making it the section journal with the second-most nonprofit publications during the timeframe we examine. With 21 papers, Accounting Horizons is the AAA journal with the largest number of nonprofit research papers published during our sample period.

Over the past two decades, FAM is the journal that has published the most nonprofit accounting articles (88), followed by NVSQ (68), JPBAFM (41), JBE (34) and JAPP (28). There are, on average, approximately 13 articles per year for the past 21 years in journals that comprise our “other” category, which do not cater specifically to either nonprofit topics (FAM, JPBAFM, JAPP, JAAF, JAL) or accounting research (NVSQ, JBE). Multiple nonprofit accounting papers representing each of our 17 topic areas have been published in these “other” journals, indicating that the range of nonprofit accounting research topics is broad and becoming more diverse. FAM, which contains more international nonprofit accounting research (74 papers) than the total from all other journals in our list (70 papers), also publishes more papers examining healthcare (40), managerial issues (37), governance (31), regulation (15), and higher education (11) than the other journals. JPBAFM and JAPP frequently publish articles investigating healthcare (15 and 14, respectively) as well, with other common topics in those journals being financial reporting and disclosures (13 and 15, respectively) and managerial (8 each). NVSQ publishes more studies pertaining to financial reporting and disclosures (43), research methodology (6), and fraud (3) than any other journal in our list, while JBE has published 16 of the 18 papers related to corporate giving.

Altogether, the results in Figure 2 indicate there is a wide audience for the various types of research being conducted by nonprofit accounting scholars. Both the availability of data and the increased number of publication outlets that are receptive to nonprofit accounting research have contributed to the increase in the number of published papers in this area, from 197 in the 2000–2010 period (an average of 18 per year) to 250 in the 2011–2020 period (an average of 25 per year). A simple t-test shows the yearly average is significantly larger (p < 0.01) in the second period compared to the first.

The future of nonprofit accounting research will likely see a continuation of some of the trends of the past 21 years, especially as it relates to the increasing availability of data. Future opportunities could involve both improved access to existing data and the collection or creation of new information about nonprofits. In the realm of providing improved access to existing data, the 2019 Taxpayer First Act requires electronic filing of all 990s for tax years that begin after July 1, 2019, making the relatively new E-Filer database the most complete source of 990s moving forward. While the data available through AWS is somewhat difficult to use, the Nonprofit Open Data Collective (NODC) and Open990¹⁴ are working to create more accessible datasets derived from the E-Filer database. While we identify only four published studies to date that use E-Filer data, it is likely to be widely used in future nonprofit research, especially because the NCCS data files have not been updated to contain the most recent reporting years.

Accounting researchers will also need to investigate the impact of substantial changes in the political, economic, and/or cultural landscapes surrounding nonprofit organizations. For example, The Tax Cuts and Jobs Act of 2017 implemented changes, such as increasing the standard deduction, that greatly decrease the number of taxpayers who itemize deductions, in turn reducing the number of people who receive a tax deduction for their charitable giving. It remains to be seen—because the data are only just starting to become available—whether those changes have led to a change in donations and, if so, if they affected all nonprofit sectors similarly.

Additionally, the global COVID-19 pandemic has affected demand for services from and availability of donations to nonprofit organizations. As different countries have responded to the pandemic in a variety of ways, nonprofit accounting scholars will have many context-specific research questions to explore. New sources of data related to the pandemic may provide several avenues for researchers to pursue in the future. For example, Charity Navigator, a prominent charity ratings agency, has curated a list of highly rated nonprofits that specialize in providing support to people affected by the COVID-19 fallout.¹⁵ It groups the nonprofits by categories that include Medical Services, Relief Supplies, Education & Awareness, and Funding Local Organizations. While data are not yet available relating to donors' use of this resource, it shows one of the benefits of having ratings agencies in place that can assess the work of many nonprofit organizations and then use their analyses to provide donors with information that they otherwise might not be able to find on their own without investing much time and effort into the process.

A recent trend in the nonprofit sector is the effective altruism movement, which seeks to maximize the societal benefit of philanthropic dollars. GiveWell, a pioneer in this area, examines the impact of charities and recommends a list of organizations that maximize the lives saved or improved with each dollar of giving. GiveWell reports that its donors contributed $150 million to recommended charities in 2019 (Dey 2020). Other agencies have taken notice of the effective altruism movement and are starting to emphasize program outcomes and results alongside traditional financial metrics. GuideStar recently introduced its Platinum status, the highest level of recognition it offers to acknowledge organizations that report actual progress and results (i.e., outputs and outcomes).

In a similar vein, Charity Navigator recently acquired Impact Matters, an organization focused on evaluating high-impact nonprofits. In October 2020, Charity Navigator began a beta test of the Impact and Results Beacon, which will display information about the impact of the nonprofit organizations it rates. For example, the Impact and Results Beacon for Evidence Action finds that $0.50 provides clean water to a person for a year. It will also provide a score out of 100, which will make impact comparable across organizations, even when their units of measurement for outcomes are different. These and other sources of impact data will likely be useful for donors who wish to measure the true influence of their giving. Future researchers may study this information to determine its incremental value to a variety of stakeholders.

To aid nonprofit accounting scholars in their efforts to continue the positive trends this research area has seen over the past twenty-plus years, we include links to some resources pertaining to the information we have discussed. Appendix B summarizes information about the 990 datasets discussed in this paper. As these datasets are not held in a centralized location, we provide links to the data sources and summarize the years available for download, the types of 990s included (e.g., 990, 990-EZ, 990-PF), the number of observations and variables, the schedules available, and information regarding specialized data. See also Feng et al. (2014) for specific guidance on using many of the available databases, including a discussion of benefits and limitations of each.

In this article we review a significant portion of the nonprofit accounting research published in the years since Gordon et al. (1999) first encouraged scholars in the field to take advantage of the newly available digitized data from NCCS. Our results are summarized using descriptive statistics. What we find demonstrates that work in the field has increased and diversified over the past two decades, likely due to several factors. First, it appears nonprofit scholars answered the call from Gordon et al. and took advantage of readily available digitized data to address questions on a broad range of topics. Second, the amount and type of accessible information has expanded during the timeframe we examine. The IRS revised Form 990 in 2008 to include new governance and compensation information, and CMS increased the amount of healthcare data reported beginning in 2011. Third, nonprofit scholars had new outlets for their accounting work as several AAA journals receptive to nonprofit research were established after 2000. With 14 section journals now available, nonprofit studies can find a home in a wide range of special topics journals. Finally, nonprofit accounting research appears to be no longer destined exclusively for nonprofit journals. Each of the top-ranked accounting journals have published nonprofit research, as have the AAA association-wide and section journals, general accounting journals that do not focus on nonprofits, and nonprofit journals that do not focus on accounting. The field has grown and matured, and opportunities to continue examining new areas in a changing world abound.

JOGNA-2021-011_Online Appendix: http://dx.doi.org/10.2308/JOGNA-2021-011.s01

JOGNA-2021-011_Figure 2: http://dx.doi.org/10.2308/JOGNA-2021-011.s02