Under the current definitions of labor force status, one person can be working in a job (in retail sales, for example) and be counted as employed because he or she is being paid, while another person can be working in a nearly identical job (such as in retail sales for a charity organization’s consignment shop) and be counted as not in the labor force because he or she is not receiving pay. Retirees doing substantial amounts of volunteer work are classified as not in the labor force, even when their work is done in a traditional workplace. We do not suggest changes in the official statistical concepts of the U.S. Bureau of Labor Statistics; however, we discuss issues relating to an expanded measure of work to supplement the current definition to include volunteers, and we construct measures consistent with that definition from currently available surveys. These measures could be particularly useful for understanding the labor force activity of older people, especially those who are retired from regular employment and are volunteering. Further, in 2020, the COVID-19 pandemic has highlighted the importance of volunteering—both the kinds of volunteer work that are currently measured and those that are not. This article calls attention to this measurement issue and other matters relating to the understanding, collection, and presentation of data on volunteers.
The Current Population Survey (CPS) is one of the oldest, largest, and most well-recognized surveys in the United States. It is immensely important, providing information on many of the things that define us as individuals and as a society—our work, our earnings, our education.1—U.S. Census Bureau and U.S. Bureau of Labor Statistics
To understand how the economy is functioning and changing over time requires measuring and valuing all productive unpriced labor time.2—Henry M. Peskin and Burton A. Weisbrod
Including volunteering as a subset of work means that the tangible and invaluable contributions volunteering makes to individuals and society are being recognized as a force that should be tracked and measured so that it can be better supported and fostered.3—The Johns Hopkins Volunteer Measurement Project
As the economy changes over time, it is useful to reconsider fundamental labor force measurement issues. In 2020, the coronavirus disease 2019 (COVID-19) pandemic has raised many such issues; for instance, it has highlighted the impact of volunteering, including forms of volunteering that are not currently included in the available statistics on volunteers. In this article, we examine an issue of language and measurement. The definition of work has important implications at both the micro- and macrolevels. At the microlevel, the definition of work affects our understanding of how people spend their time. As the opening quotation about the Current Population Survey (CPS) suggests, how work is defined also affects how people define themselves. At the macrolevel, it affects our understanding of the aggregate amount of productive activity people are undertaking in the labor force. In this article, we discuss the definition of work in labor force statistics and analyze issues relating to a supplementary measure that includes volunteering as a subset of work.
Under the current definitions of labor force status, one person can be working in a job (in retail sales, for example) and be counted as employed because he or she is being paid, while another person can be working in a nearly identical job (such as in retail sales for a consignment shop of a charitable organization) and be counted as not in the labor force because he or she is not receiving pay.
The COVID-19 pandemic has highlighted the importance of volunteering, especially activities done individually rather than through organizations. Examples include grocery shopping for older neighbors or friends and making face masks to give to people who could not obtain them. The people doing this type of volunteering—termed “direct volunteers”—are not counted as employed or as volunteers in U.S. statistics because they are not doing this work through an organization. They are counted as not in the labor force. The 2020 Presidential Proclamation announcing National Volunteer Week specifically mentions some of the types of volunteering that are “invisible” in current statistics.4 Although this article focuses on the labor force status of measured volunteers, it also calls attention to matters involving the understanding, collection, and presentation of data on volunteers.
Currently, the U.S. Bureau of Labor Statistics (BLS) has six measures of labor underutilization (U-1 to U-6), one of which (U-3) is the official unemployment rate.5 This framework of measures recognizes that one definition alone does not suit all data users. We consider issues relating to whether there should be additional measures of labor utilization. In particular, we consider the case of whether there should be an expanded definition of employment that is based on the nature of the activity people are doing and not solely on monetary compensation. Under the current definition, people are considered employed only if they worked for pay or profit.
An expanded definition could supplement the current definition and provide an additional measure of work that could be useful in enhancing our understanding of the transition from work to retirement and the productive contributions of older people. Such a definition would recognize that many people, particularly retirees with pension benefits, do not require or need monetary compensation but will engage in volunteer work because of its nonmonetary benefits, including psychological and health benefits.
According to BLS statistics, in 2015, among the U.S. civilian population ages 16 and over, 63 million people, or 25 percent, volunteered. Among people ages 65 and over, 11 million volunteered, or about 24 percent of the population in that age group.6 Although its volunteer rate was about the same as that of the total population, the older population accounted for median annual hours of volunteering that were close to double the median for the total population ages 16 and over. Overall, most volunteers combined paid work with volunteer work; however, for the older population, three-quarters of the volunteers were not in the paid labor force.
The concept of work that we analyze in this article would not replace existing concepts or measures of labor force activity but would supplement them, providing a different and possibly fuller picture of work in the labor force. Our focus is on volunteers as defined by BLS, particularly those volunteers who are ages 65 and over, in order to capture volunteer work done by retirees who are mostly out of the officially defined labor force. We recognize, but do not attempt to measure, a wider concept of volunteer work (informal volunteer work carried out on one’s own directly for individuals and not for an organization), as well as other forms of unpaid work.
The official BLS definition of the labor force includes some unpaid workers—unpaid family workers who work at least 15 hours per week in a family farm or business. The number of such workers has dwindled from more than 1 million in 1970 to fewer than 100,000 in December 2019, trending downward with the decline in traditional family farms.7 Such family workers are counted as employed, in recognition of their economic contribution, while millions of people who make similar contributions by performing unpaid volunteer work are not counted as employed.
Research conducted by Yvon H. Pho for the U.S. Bureau of Economic Analysis has parallels with the concept we consider here.8 Just as Pho gives a national accounting perspective on volunteers and estimates the value of their work, our article offers a labor force activity perspective on volunteers. Pho estimates the value of volunteer work at around 1 percent of gross domestic product; we show how an alternative labor force participation rate increases—and quite substantially for the older population—when volunteers who are currently classified as not in the labor force are counted as part of the labor force. Much like Pho, we are looking outside the scope of the official definitions in order to acknowledge volunteers’ contributions as workers.
The article begins with a discussion of a theory of economic measurement, illustrating why the current definition of work may bear reexamination in a changing economy. Next, we discuss the concept and definition of work. We then consider an expanded concept based on the nature of the activity rather than on the nature of the compensation. This concept for a definition of work could supplement the existing definition and is in close accord with recommendations made by the International Labour Organization.9 We then briefly introduce some of the research on the nonmonetary aspects of compensation. We proceed to the definition of volunteer work and some of the BLS data on volunteering. Using two BLS sources, we then estimate the magnitude of the effect of including volunteers in an expanded concept of labor force participation. These estimates have advantages and disadvantages that we describe. We provide several suggestions for improving the presentation and measurement of the data on volunteers, and we follow with a few concluding comments.
A 2002 article cowritten by staff from BLS and the U.S. Census Bureau includes the following statement: “The labor force concepts and definitions used in the CPS have undergone only slight modification since the survey’s inception in 1940.”10 Although this kind of historical continuity is desirable in many ways, statistical measures need to be revised from time to time, as demographics and the economy change, economic theory progresses, and new concepts are developed. In the current era, changes in the economy include the aging of the population, increased life expectancy, better health at older ages, and a much larger number of people who are able to perform work without compensation because of Social Security and pension benefits. Economic theory is progressing as it increasingly incorporates concepts from psychology and behavioral economics, with growing recognition that work often provides psychological benefits as a form of compensation.11
In addition, different statistics may be needed for different age groups. For example, Leslie A. Muller and John A. Turner argue that measures of poverty for older age groups should differ from those for younger age groups because of the selectivity bias at older ages that results from the positive relationship between income and life expectancy.12 In this article, we consider whether a different measure of work may be needed for older people because, given that they are receiving Social Security and pension benefits, they are able to work for nonmonetary compensation.
In this section, we discuss the definitions of work used by various statistical agencies.
The Current Population Survey (CPS), which is cosponsored by BLS and the Census Bureau, is the nation’s primary source of labor force statistics for the working-age population.13 According to explanatory information about the CPS on the BLS website, “The basic concepts involved in identifying the employed and unemployed are simple: People with jobs are employed.”14 Other CPS documentation states the following: “People are considered employed if they did any work at all for pay or profit during the survey reference week. This includes all part-time and temporary work, as well as regular full-time, year-round employment.” It also notes that unpaid family workers are included among the employed if they worked in a family business or farm for at least 15 hours per week. The CPS definition of employment specifically excludes people “whose only activity consisted of work around their own house (painting, repairing, cleaning, or other home-related housework)” and those performing “volunteer work for religious, charitable, or other organizations.” Although the documentation does not mention them, family caregivers are also excluded.15
The International Labour Organization (ILO) is a United Nations agency whose primary mandate is to set international labor standards in order to promote economic and social justice. The ILO conducts its activities through a permanent secretariat known as the International Labour Office.16 One function of the ILO is to recommend standards for defining and measuring “work statistics,” which it defines as “productive activities of people.” Among these activities, “employment” is defined by the ILO as work done for compensation or for profit and “volunteer work” as noncompulsory work done for others without compensation.17 In addition, in 2011 the ILO published a manual on measuring volunteering that further defines volunteer work as “time individuals give without pay to activities performed either through an organization or directly for others outside their own household,” including “activities that produce goods and/or services which contribute something of potential value to its recipients.”18 The ILO definition of work is independent of the form of compensation (i.e., monetary or nonmonetary), independent of the nature of the employer (for profit or nonprofit), and independent of whether it is done voluntarily or under compulsion, although the ILO has refined the recommended definition of volunteers to exclude compulsory volunteering.
Furthermore, at the October 2013 International Conference of Labour Statisticians, the ILO adopted a resolution that introduced a new statistical framework for a broad range of human productive activities. The resolution significantly expanded the conceptualization of work, and it separately identified five mutually exclusive forms of work. This was the first time that a comprehensive set of operational concepts, definitions, and guidelines had been agreed upon internationally that attempted to cover all forms of work, whether paid or unpaid.19 This new framework on different forms of work—with volunteer work identified as one form—has increased countries’ interest in measuring volunteer work, as evidenced by many new data sources on volunteering that are being introduced worldwide.20 The new framework supplements current measures of employment, rather than replacing them.
In this section, we explain an aspect of the concept of work that we are proposing. This concept of work can be divided into two broad categories: work in the labor force and unpaid work in the person’s own home. Work in the labor force can be divided into work for pay or profit (employed and self-employed) and unpaid work. We are focusing on unpaid work. Unpaid work can be divided into noncompulsory volunteer work for nonmonetary benefits, compulsory “volunteer” work (such as high school graduation requirements), and unpaid work with the expectation of future pay based on the work experience, such as an unpaid internship. We are focusing on volunteer work, including compulsory “volunteer” work, which is within the scope of the BLS definition of volunteers.21 Thus, we are focusing on volunteer work done for an organization and not considering volunteer work done directly for another person.
We consider the option that, for the purpose of calculating an expanded measure of labor force participation, people who are currently classified as not in the labor force but are doing volunteer work of at least 1 hour per week be placed in a separate category. Thus, the basic framework we consider would divide people into the following four categories with respect to labor force participation: working for pay or profit, working not for pay or profit (i.e., volunteering), unemployed, and not in the labor force. The current framework includes three categories: working for pay or profit, unemployed, and not in the labor force. People working without pay or profit can be in any of the other three categories. They can be employed and also volunteering, they can be unemployed and also volunteering, and they can be not in the labor force and also volunteering. In order to avoid double counting, we focus on those who are currently classified as not in the labor force. Under current BLS definitions, the labor force consists of people who are employed and those who are unemployed, the latter defined as people who are actively seeking employment but are unable to find it. According to BLS classification rules, if people are in more than one of the official categories at the same time—such as a person who currently has a job while also seeking another one—employment always takes precedence over unemployment so that double counting is avoided.
As the CPS documentation states, “The CPS focuses on the labor force status (employed, unemployed, not in labor force) of the working-age population and the demographic characteristics of workers and nonworkers.”22 Under the current definitions, people who are actively working in a volunteer capacity—but not in family businesses and not for pay—are classified as being not in the labor force. For example, a person can work for more than 1,000 hours a year as a volunteer without pay and still be classified as someone who is not in the labor force.
Our concept of work would include some “unpaid work”—that is, work not done for pay or profit. For example, under this approach, if a person worked in retail sales for a nonprofit thrift store as a volunteer, that person would be considered as working (or employed) under the expanded definition, just as if he or she were working in retail sales at a store for pay. Thus, the expanded definition of work that we propose is based on the nature of the activity, rather than on the nature of the compensation. Our definition would include the current group of employed workers plus some workers who are currently classified as not in the labor force—namely, those who are working as volunteers. (Volunteers who are currently employed or actively seeking employment are already classified as being in the labor force.) This expanded definition would provide a more complete picture of the productive activity that people perform.
A similarity between paid work and unpaid work is that generally both provide nonmonetary forms of compensation. The current exclusion of unpaid work seems to be based on the concept that work is something that people would only be willing to do for pay because of the marginal disutility of the activity. Although that concept applies to many work-for-pay situations, it does not apply to all work situations. A Gallup survey conducted in 2015 and 2016 found that 33 percent of workers are actively engaged in their work, while 51 percent are disengaged and 16 percent have negative feelings toward their work.23 One possible reason volunteer work is less likely to have disutility at the margin is that the hours are generally flexible, and people can adjust their hours so that they do not reach the point at which the last hour results in disutility.
Historically, economists have tended to focus on the monetary compensation associated with work, but more recently, with the incorporation of concepts from psychology into economics, economists have increasingly come to recognize the nonmonetary compensation associated with work. For example, in a 2018 study, Isabel V. Sawhill and Christopher Pulliam write, “lack of work likely leads to social isolation, diminished self-worth, and too much unstructured time. In short, work provides more than income. It provides self-respect, a sense of contributing, an identity, and connection to others.” The authors report that the General Social Survey, a nationally representative survey conducted by the National Opinion Research Center at the University of Chicago, asks respondents the following question: “If you were to get enough money to live as comfortably as you would like for the rest of your life, would you continue to work or would you stop working?” Seventy percent of Americans reported that they would continue to work. Thus, this survey provides evidence against the definition of work as an activity that a person would only do for pay because of its disutility. It supports the concept that nonmonetary compensation can also be an important reason to work.24 Supporting evidence is provided by the RAND 2015 American Working Conditions Survey, which shows that 56 percent of workers report that they have very good friends at work.25
This section looks more closely at definitions of volunteer work. BLS defines volunteers as follows: “Volunteers are defined as persons who did unpaid work (except for expenses) through or for an organization.”26 Thus, BLS recognizes volunteering as a type of work, but it does not include volunteers as part of the labor force. The BLS definition also does not include volunteering that is done directly, such as helping an older neighbor or one with a disability, as opposed to volunteering through an organization. Volunteers do a wide range of work activities that contribute to society, including serving food, mentoring and tutoring students, and coaching youth sports teams. In addition, many people engage in more than one kind of volunteer activity. Figure 1 shows CPS data collected in September 2017, covering the previous 12 months, according to types of volunteer activity by people ages 16 and over.
The ILO Manual on the Measurement of Volunteer Work, published in 2011, states the following: “The objective [of collecting and publishing data on volunteers] is to make available comparative cross-national data on a significant form of work which is growing in importance but is often ignored or rarely captured in traditional economic statistics.” The manual defines volunteer work as follows: “Unpaid non-compulsory work; that is, time individuals give without pay to activities performed either through an organization or directly for others outside their own household.”27 Subsequently, this definition was modified so that it would exclude persons volunteering to help family members living in other households.28 The wording was amended again at the 2013 ILO International Conference of Labour Statisticians to that shown in exhibit 1. This exhibit includes key features of the ILO-recommended definition and its measurement. This is a broader definition of volunteers than the one BLS uses in its surveys on volunteering because it includes volunteer work performed directly for others, rather than only that done through an organization. At the same time, however, it is also a narrower definition in that the ILO excludes all forms of “compulsory” volunteering, whereas BLS includes people who volunteer under court order as well as students who engage in volunteer activities as a requirement for high school graduation.29
Exhibit 1. ILO recommendations for defining and measuring volunteer work
Definition of volunteers: All those of working age who, during a short reference period, performed any unpaid, noncompulsory activity to produce goods or provide services for others.
Key elements of measurement:
Sources: International Labour Office, Manual on the Measurement of Volunteer Work (Geneva, Switzerland: International Labour Organization, March 2011), pp. 11–20, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/publication/wcms_162119.pdf; and “Report of the Conference,” Report III, 19th International Conference of Labour Statisticians, Geneva, October 2–11, 2013 (Geneva, Switzerland: International Labour Organization, October 2013), pp. 54–55, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/publication/wcms_234124.pdf. The latter source provided some amendments to the definition cited in the Manual.
The ILO has established a partnership with the United Nations Volunteers (UNV) program to advance survey methods for producing statistics on volunteer work. Between 2018 and 2021, The ILO and UNV are partnering with interested national statistical offices to test a module on volunteer work suitable for attachment to national labor force surveys in order to provide practical guidance and to help countries implement the module in their national labor force surveys.30 The ILO reports that as of 2019, nearly 60 countries had measured volunteer work via modules attached to their national surveys. National data collected on volunteers are available from the ILO Department of Statistics website. However, the ILO warns that because countries use different approaches and tools for collecting these data, they do not recommend direct cross-country comparisons.31
Data on volunteers in the United States come primarily from two sources: (1) a September supplement to the CPS and (2) the American Time Use Survey (ATUS), which uses a sample derived from the CPS. Both surveys use the same definition of volunteering, but they collect the data over different timeframes; the September CPS supplement collects data for a “reference year,” whereas the ATUS collects data for a single day, based on the entries to a diary kept by the respondent. The September supplement has undergone some recent changes that we describe below. We then highlight some of the results from the 2015 September supplement to the CPS.
Beginning in 2002, BLS published data annually on volunteering through a special September supplement to the CPS.32 After the 2015 survey, however, BLS stopped publishing data on volunteers. Instead, beginning in 2017, the sponsor of the September supplement, the Corporation for National and Community Service (CNCS), an independent federal agency that supports national volunteering efforts, took responsibility for publishing these data. After 2017, the September supplement to the CPS became biennial.33 New questions were added that combine data collection for volunteers with data collection on civic engagement. Along with volunteering, civic engagement includes such activities as talking to and spending time with friends, family, and neighbors; posting views about political, social, or local issues on the internet or on social media; voting in local elections; attending public meetings; and belonging to civic organizations and clubs. The questions on the other types of civic engagement come before the questions on volunteering. This change, along with others discussed later in the article, has affected the historical continuity of the series. The changes in the survey were the result of recommendations made by the National Academies of Sciences, Engineering, and Medicine Committee on National Statistics, which was commissioned by the CNCS in 2010 to improve understanding of civic engagement, including measures of volunteers. The committee issued their report and recommendations in 2014.34 The Census Bureau continues to collect data from the September supplement on volunteering, but the data are published by the CNCS on its website. The CNCS publishes these data for all 50 states and the District of Columbia.35
The technical note to the 2016 BLS news release on volunteering includes the following language from the survey referencing September 2015:
Following this introduction, respondents were asked the first supplement question: “Since September 1st of last year, have you done any volunteer activities through or for an organization?” If respondents did not answer “yes” to the first question, they were asked the following question: “Sometimes people don’t think of activities they do infrequently or activities they do for children’s schools or youth organizations as volunteer activities. Since September 1st of last year, have you done any of these types of volunteer activities?”36
In 2017, the survey began asking a more probing second question that includes a long list of volunteer organizations that are read to the respondent, as follows: civic, political, professional or international; educational or youth service; environmental or animal care; hospital or other health; public safety; religious; social or community service; sport, hobby, cultural or arts; other.37 This question, along with other changes in the questionnaire, seem to have created a discontinuity in the data, probably explaining an increase in the volunteer rate from 24.9 percent in 2015 to 30.3 percent in 2017, as well as an increase in the number of volunteers over that period, from 63 million to 77 million.38 Over the period from 2011 to 2015, the survey showed a stable volunteering rate of about 24 or 25 percent.
According to the September 2015 volunteer supplement, about 62.6 million people volunteered through or for an organization at least once between September 2014 and September 2015. Overall, the volunteer rate for the population ages 16 and over was 24.9 percent in 2015, and the rate for those ages 65 and over was not much different, at 23.5 percent. However, these figures obscure a much higher intensity of volunteer work among older workers. Median annual hours spent on volunteer activities ranged from a high of 94 hours for those ages 65 and over to a low of 36 hours for those under the age of 35.
A relatively early study found that older people accounted for 19 percent of total volunteer hours in 1989.39 That percentage is higher now with the retirement of the baby-boom generation: the 2015 volunteers supplement indicates that people ages 65 and over accounted for 25 percent of total volunteer hours. The total number of hours volunteered by the population ages 16 and over in 2015 was 8.9 billion, with workers ages 65 and over contributing 2.2 billion of those hours. In fact, a 2015 study finds that volunteering by people ages 65 and over accounts for 45 percent of volunteer hours among adults ages 25 and over.40
Although additional age cohorts for the older population are not published for the latest surveys, these cohorts are available in a 2003 Monthly Labor Review article based on the 2002 survey.41 Among people ages 65 and over, the volunteer rate declined with age. For example, 26.3 percent of people ages 65 to 69 volunteered, compared with 25.0 percent of people ages 70 to 74, 22.9 percent of people ages 75 to 79, and 16.1 percent of people ages 80 and over.
Among people who volunteered in 2015, 6.8 percent of those ages 55 to 64 volunteered for 500 or more hours per year, and 9.4 percent of those ages 65 and over volunteered for 500 or more hours per year. People ages 65 and over had the highest percentage for any age group, and we suspect the percentage would be even higher for people ages 65 to 74.
Thus, these data on volunteer rates and labor force participation rates suggest that an expanded definition of work to include unpaid work would have its largest effect on the statistics for people ages 65 and over, a rapidly growing segment of the population because of the aging of the baby-boom generation. Data from the ATUS discussed in the next section show similar results for volunteering among the population ages 65 and over.
Since 2003, the ATUS has provided national estimates of how Americans ages 15 and over spend their time, covering the full range of market and nonmarket activities, including volunteering.42 The ATUS is a continuous survey conducted for BLS by the Census Bureau. The sample is drawn from participants in the CPS, with respondents assigned a day about which they report their activities over the entire 24 hours. Thus, ATUS involves less recall bias than the CPS September supplements, which ask respondents to remember their volunteer activity over the course of an entire year.
The ATUS reports data in terms of average hours per day per activity. Figure 2 shows averages for the period from 2011 to 2015 on the percent of the population ages 15 and over who volunteered on an average day. The highest rates of volunteer activity are found among those ages 65 and over. About 9 percent of the population in this age group volunteered on an average day, compared with 6 percent of the entire population ages 15 and over.43
What if the productive activity of volunteers were recognized by including volunteers in a supplemental measure of the labor force? In the following exercise, we investigate the impact of including volunteers in labor force participation rates and employment–population ratios in 2015. The purpose is to show whether including volunteers as workers would change our perception of the economic contribution of the population generally or of older people. We make estimates for the total working-age population (ages 16 and over) and for the older population (ages 65 and over).
We tried two different approaches. Our first approach was an attempt to replicate the CPS definition and reference period by using other data sources. However, we were unable to arrive at supportable estimates compatible with the monthly CPS because such calculations would have involved mixing surveys with different reference periods, requiring numerous and often questionable assumptions. Instead, we used an approach that combines the results of two CPS supplements, which allows us to make reasonable calculations without additional data collection.
The method adopted used a “work experience” approach. That is, we calculated yearlong participation rates and employment–population ratios. The data sources are two CPS supplements: (1) the “work experience” portion of the Annual Social and Economic Supplement (ASEC) to the CPS and (2) the CPS September supplement on volunteers. An explanation of the methodology follows. Table 1 shows the calculations and sources.
|Characteristics||Ages 16 and over||Ages 65 and over|
Work experience data
Civilian noninstitutional population
Total who worked or looked for work
Percent of population
Total who worked during the year
Percent of population
Not in the labor force during the year
Data on all volunteers
Total who volunteered during the year
Percent of population
Employment status of volunteers
Civilian labor force
Not in the labor force
Adjusted after adding volunteers who were not in the labor force
Percent of population
Percent of population
Note: The work experience data are collected in the Annual Social and Economic Supplement to the Current Population Survey (CPS) and refer to work activity performed during the entire 2015 calendar year. The data on volunteers are collected in a September supplement to the CPS and cover the period from September 1, 2014, through the survey period in September 2015. The population figures for the two measures differ accordingly. The adjusted levels are calculated by adding volunteers who were counted as not in the labor force to the work experience totals for the labor force (worked or looked for work) and the employed.
Source: U.S. Bureau of Labor Statistics, Current Population Survey.
We start with data reported in the work experience survey conducted from February to April 2016 for the 2015 calendar year. This survey collects information on employment and unemployment experienced during the previous calendar year. Overall, 65.5 percent of the population ages 16 and over worked or looked for work (i.e., were in the labor force at some time in 2015), while 64.2 percent worked at some time during the year (i.e., were employed). For the population ages 65 and over, the rates based on the work experience data are much lower, at 23.5 percent and 23.0 percent, respectively. These published work experience data are shown at the beginning of table 1.
Data from the CPS volunteers supplement were collected in September 2015; respondents were asked to recall their volunteer activities over the period from September 2015 back to September 1, 2014. Thus, both the work experience and volunteer surveys use reasonably comparable timeframes. Statistics from the supplement are also shown in table 1, indicating volunteer rates of 24.9 percent for the population ages 16 and over and 23.5 percent for those ages 65 and over.
To obtain a yearlong participation rate that includes volunteers, as well as to avoid double counting, we must exclude the volunteers who are already counted as in the labor force. The volunteers supplement reports data on the employment status of volunteers by age groups. The number of volunteers who are not in the labor force is the volunteers total to be added to the reported labor force and employment data from the work experience survey to arrive at work-experience-based, year-long participation rates and employment–population ratios.These estimates are not to be confused with the official CPS figures for these two measures, which are based on data collected according to labor force activity in reference weeks.
The calculations show a relatively moderate increase in the yearlong labor force participation rate for people ages 16 and over—from the reported figure of 65.5 percent to the adjusted figure of 73.4 percent. For those ages 65 and over, as expected, the increase resulting from adding volunteers to the labor force is much larger, from 23.5 percent to 41.0 percent. This sizeable increase for older people occurs because three-quarters of them are classified as not in the labor force, whereas most of the population ages 16 to 64 were working or actively seeking work and therefore were already included in the labor force. The results for employment–population ratios are similar. (See table 1.)
As mentioned previously, these calculations bring together data from two different CPS supplements that are collected for different purposes—one to enumerate volunteers and the other to present data on work experience. These surveys also have slightly different timeframes. Thus, the estimates provided here cannot be considered exact values, but rather as values that provide a rough order of magnitude. The advantage of these estimates is that they are based on data collected from the same CPS population and survey methodology; they are transparent calculations that can be verified with relative ease; and the timeframes are reasonably comparable. Another important advantage of these estimates is that they do not require additional data collections, but instead use existing and ongoing data sources based on the CPS. They do not affect data comparability over time because they do not affect existing labor force measures; rather, they provide a supplement to those measures.
Linking volunteers into a supplementary labor force statistics framework, as shown in the previous section, could change the perception of the economic contributions made by different groups of people, especially the older population. The exercise on estimating work-experience-based participation rates and employment–population ratios including volunteers shows that the impact is sizable for older workers. This calculation clearly makes volunteer work more visible.
Some other issues could be considered relating to the understanding, collection, and presentation of data on volunteers: recall bias, more detailed age breakdowns of economic statistics, and data that cover all volunteers, not only those who volunteer through organizations. These issues are considered in the next three subsections.
The September volunteer supplements use a reference year rather than a reference week (as in the monthly CPS). Because volunteer work is often sporadic and does not always adhere to a regular work schedule, the survey responses for the longer reference period in the volunteer survey are subject to a higher level of recall bias. Although all surveys that rely on a respondent’s memory are subject to recall bias, this may be more prevalent for sporadic activity among people of older ages. The longer the reference period, the more difficult it is to recall the requested information accurately, and thus the surveys need to provide additional prompts to try to help respondents remember. A strong additional prompt question has indeed been asked of respondents since 2017. Nevertheless, respondents may forget activities performed near the beginning of the reference year. A high level of recall bias may distort the results.
The ILO Manual recommends a 4-week recall period to take care of most of the issues of recall, noting that testing of the ILO survey module “suggests that the four-week reference period should be long enough to capture irregular activity but not so long as to make recall unduly difficult.”44 Even a quarterly survey that requires the recall of volunteer activities over the previous 3 months would involve fewer recall problems than an annual recall period. More frequent enumeration of volunteers over the course of a year would likely lead to a higher count of volunteers, making them more visible in economic statistics.
The Alzheimer Society of Canada makes the following point about age-related memory decline, which is relevant for recall bias at older ages: “Almost 40 per cent of people over the age of 65 experience some form of memory loss. When there is no underlying medical condition causing this memory loss, it is known as ‘age-associated memory impairment,’ which is considered a part of the normal aging process.”45 Age differences in recall have long been discussed in the scientific literature, with, for example, a laboratory investigation documenting them in 1987.46 Thus, recall bias, taking the form of not remembering volunteer activities over a recall period of 1 year, presumably is a more serious problem for older people than for younger ones. For this reason, the survey responses for older people may understate (or even overstate) the actual amount of volunteer work being performed. However, we believe that people are more likely to make an error of not remembering something that did happen than to make an error of imagining something that did not happen.
Changes in the age structure of the U.S. population may necessitate changes in the age groups for which data are routinely published. The percentage of the population ages 65 to 74 increased from 3.8 percent in 1930 to 7.0 in 2010 and is projected to increase to 10.8 percent by 2030.47 With the growth of this age group, standard age groups, which often have the 65-and-over group as the oldest one, could be further divided to include people ages 65 to 74 and those ages 75 and over. It would be particularly useful to have statistics for those ages 65 to 74 because people in that age group are often transitioning from full-time work to retirement by volunteering. As stated previously, data on volunteering for these age groups were published in 2003, but they have not been published since then. Publishing data for the older age groups would also help make older volunteers more visible. Since 2017, the CNCS has published data online by generational categories—Generation Y, millennials, Generation X, baby boomers, the Silent Generation—rather than by specific age groups.
In this section, we consider the option of collecting data for all volunteers, as opposed to collecting data only for people who volunteer for or through an organization. As mentioned previously, the COVID-19 pandemic has highlighted the contributions made by “direct volunteers”—those who see a need and respond to it directly, without working through an organization. They represent a mostly unmeasured component of volunteers in the United States.48 A study for the National Academy of Sciences states the following:
Time spent helping others informally is arguably as or more important than formal volunteering. As the population ages, we might expect an increase in the reliance on friends, family, and neighbors to help out with grocery shopping, yard work, and other activities of daily life. Some of these activities fall conceptually between household production and volunteer work. The practical problems of measuring and valuing volunteer labor activity not connected with a formal organization may be severe; still, we want to highlight its omission from traditional economic accounts and because such activity clearly contributes to real output, encourage attention to it.49
The Alzheimer’s Association notes that 18.4 billion hours of care annually, valued at $232 billion, are provided by family and other unpaid caregivers.50 In addition, the American Time Use Survey shows that each year 41.3 million people provide unpaid care to people ages 65 and over.51 According to an estimate prepared for the 2018 State of the World’s Volunteerism Report by the United Nations, 70 percent of global volunteer activity occurs through direct person-to-person engagement, while 30 percent takes place formally through organizations.52.
The ILO concluded its 2020 announcement of the celebration of International Volunteer Day with the following statement: “A key contribution of statisticians is to make the invisible visible through more and better data. This clearly applies to the important work done by volunteers throughout the world.”53
Expanding the basic framework for measuring work would have advantages and disadvantages. The advantages include creating a broader view of the labor force that incorporates volunteer workers within a familiar framework that enables better understanding of the labor force activity of older people. A potential disadvantage is that changing the definition of work would create a break in the historical data series—data from the period after the definition change would not be comparable to data from the period before the change. This disadvantage is somewhat mitigated, however, because the definitions would not be changed in the historical data. The expanded definition would create a new data series that would begin when the change was made. In addition, most of the changes could be made within the existing framework of the current CPS, although collecting data on all volunteers, including those who volunteer directly and not through organizations, would require adding new questions to the questionnaire.
In 2020, the COVID-19 pandemic has highlighted the importance of volunteering, not only activities done through organizations, but also those done by “direct volunteers”—such as grocery shopping for their older neighbors and friends or sewing face masks to give to people who otherwise might not be able to obtain them, including healthcare workers. This kind of volunteering is currently not counted in U.S. statistics on volunteers, leaving a wide gap in the data on this subject.
ACKNOWLEDGMENTS: The authors would like to thank the following people for their valuable contributions to this article: Anthony Nerino, formerly of the Corporation for National and Community Service, for providing guidance on and data from the 2017 volunteers survey; S. Wojtek Sokolowski and Megan Haddock of the Johns Hopkins University Center for Civil Studies for their comments on an early version of this manuscript and guidance toward additional sources of information; Vladimir Ganta of the International Labour Office for information on the ILO’s recent work on volunteer statistics; Anne Polivka of the Bureau of Labor Statistics for suggestions on BLS data sources; Lawrence Leith of the Bureau of Labor Statistics, who made numerous valuable suggestions; an anonymous reviewer from the Bureau of Labor Statistics, who also made a number of valuable suggestions; and CPS staff at the Census Bureau who provided information on the surveys of volunteers.
John A. Turner, Bruce W. Klein, and Constance Sorrentino, "Making volunteer work visible: supplementary measures of work in labor force statistics," Monthly Labor Review, U.S. Bureau of Labor Statistics, July 2020, https://doi.org/10.21916/mlr.2020.15.
1 Current Population Survey: Design and Methodology, Technical Paper 63RV (U.S. Bureau of Labor Statistics and U.S. Census Bureau, March 2002), Foreword, https://www.census.gov/prod/2002pubs/tp63rv.pdf.
2 Katherine G Abraham and Christopher Mackie, eds., Beyond the Market: Designing Nonmarket Accounts for the United States (Washington, DC: National Academies Press, 2005), p. 148, https://doi.org/10.17226/11181.
4 Donald J. Trump, Proclamation on National Volunteer Week, 2020 (The White House, April 17, 2020), https://www.whitehouse.gov/presidential-actions/proclamation-national-volunteer-week-2020/.
5 John E. Bregger and Steven E. Haugen, “BLS introduces new range of alternative unemployment measures,” Monthly Labor Review, October 1995, pp. 19–26, https://www.bls.gov/opub/mlr/1995/10/art3full.pdf.
7 The Employment Situation: December 2019, USDL-20-0010 (U.S. Department of Labor, January 10, 2020), table A-8, https://www.bls.gov/news.release/archives/empsit_01102020.htm; see also, Patricia A. Daly, “Unpaid family workers: long-term decline continues,” Monthly Labor Review, October 1982, pp. 3–5. https://www.bls.gov/opub/mlr/1982/10/art1full.pdf.
8 Yvon H. Pho, “Volunteer output and the national accounts: an empirical analysis,” U.S. Department of Commerce, Bureau of Economic Analysis WP2004–03, paper presented at the Eastern Economic Association Annual Meeting, Washington, DC, February 20–22, 2004, https://www.bea.gov/system/files/papers/WP2004-3.pdf. Pho subsequently updated and expanded upon her paper in “The value of volunteer labor and the factors influencing participation: evidence for the United States from 2002 through 2005,” Review of Income and Wealth, vol. 54, no. 2, June 2008, pp. 220–236, http://www.roiw.org/2008/2008-10.pdf.
9 See Manual on the Measurement of Volunteer Work (Geneva, Switzerland: International Labour Organization, Department of Statistics, March 2011), https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/publication/wcms_162119.pdf.
10 Current Population Survey: Design and Methodology, p. 1-2.
11 See, for example, Greg Kaplan and Sam Schulhofer-Wohl, “The Changing (Dis-)Utility of Work,” Journal of Economic Perspectives, vol. 32, no. 3, summer 2018, pp. 239–258.
12 Leslie A. Muller and John A. Turner, “Selectivity bias due to differential mortality: a new measure of old-age poverty,” Journal of Aging and Social Policy, 2020, forthcoming.
13 See Current Population Survey: Design and Methodology. The CPS includes only people who are ages 15 and over and not in the Armed Forces. It excludes people in institutions, such as prisons, nursing homes, and long-term care facilities. BLS generally publishes labor force data only for people who are ages 16 and over, because the labor market activity of those younger than 16 is limited by mandatory schooling and child labor laws. There is no upper age limit in the CPS.
14 Labor Force Statistics from the Current Population Survey, “How the government measures unemployment: what are the basic concepts of employment and unemployment?” (U.S. Bureau of Labor Statistics, 2015), https://www.bls.gov/cps/cps_htgm.htm#concepts.
15 See Current Population Survey: Design and Methodology, p. 5-3.
16 See “How the ILO works: departments and offices” (Geneva, Switzerland: International Labour Organization, 2020), https://www.ilo.org/global/about-the-ilo/how-the-ilo-works/departments-and-offices/lang--en/index.htm.
17 See “Forms of work and labour force statistics conceptual frameworks” (Geneva, Switzerland: International Labour Organization, 2020).
18 See Manual on the Measurement of Volunteer Work, pp. 13–15. See also International Labour Office, “National practices in measuring volunteer work: a critical review,” Room Document 12, 20th International Conference of Labour Statisticians, October 10–19, 2018 (Geneva, Switzerland: International Labour Organization, October 2018), https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/meetingdocument/wcms_636049.pdf.
19 “Exploring light time-use approaches for measuring productive activities, what counts,” Statistics on Work Series, brief no. 1 (Geneva, Switzerland: International Labour Organization, August 2018), https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/publication/wcms_635909.pdf.
20 International Labour Office, “National practices in measuring volunteer work: a critical review,” p. 11. For instance, from 2014 to 2017, in 73 countries, 93 data sources were implemented, with 76 sources being dedicated modules on volunteering.
21 Although “compulsory” volunteering is not explicitly mentioned in the definition of volunteers, unpublished tabulations from the September 2015 CPS supplement on volunteers enumerate a small number of people performing court-ordered community service or volunteering due to school requirements. The term “compulsory” is not used in the survey, and respondents are not asked if their volunteer work is mandatory.
22 Current Population Survey: Design and Methodology, p. 1-1.
23 State of the American Workplace (Washington, DC: Gallup, 2017), https://www.gallup.com/workplace/238085/state-american-workplace-report-2017.aspx.
24 Isabel V. Sawhill and Christopher Pulliam, “Money alone doesn’t buy happiness, work does,” Real Clear Markets, November 5, 2018, https://www.realclearmarkets.com/articles/2018/11/05/money_alone_doesnt_buy_happiness_work_does_103475.html. The General Social Survey (GSS), funded by the National Science Foundation, is conducted by the National Opinion Research Center at the University of Chicago. For more information, see the GSS website at https://www.norc.org/Research/Projects/Pages/general-social-survey.aspx.
25 Nicole Maestas, Kathleen J. Mullen, David Powell, Till von Wachter, and Jeffrey B. Wenger, “Working conditions in the United States: results of the 2015 American Working Conditions Survey” (RAND Corporation, 2017), https://www.rand.org/pubs/research_reports/RR2014.html.
26 See Volunteering in the United States—2015, p. 1. Also, the BLS American Time Use Survey (ATUS) uses the same definition. See American Time Use Survey User’s Guide: Understanding ATUS 2003–2017 (U.S. Bureau of Labor Statistics, June 2018), p. 52.
27 Manual on the Measurement of Volunteer Work, p. 13.
28 The exclusion of helping “related family members” in another household was subsequently added in instructions to countries using the manual in order to exclude such situations as a family member driving another family member who lives in a different household to a medical appointment. Information from discussions with Megan Haddock, coauthor of the ILO Manual on the Measurement of Volunteer Work.
29 See Volunteering in the United States—2015, “Technical Note.” See also endnote 21.
31 Ibid. Data on the number of volunteers and volunteer rates by country are available from the ILO Department of Statistics website.
32 See Volunteering in the United States—2015. See also, Volunteering in the United States, USDL-02-686 (U.S. Department of Labor, December 18, 2002), the first news release on volunteers with data from the September 2002 CPS supplement, at https://www.bls.gov/cps/volun.pdf.
33 A supplement was added to the CPS in September 2019, covering the period from September 2018 to September 2019. The data from this supplement have not yet been processed and are scheduled for release in September 2020, according to the U.S. Census Bureau. The next survey will be conducted in September 2021.
34 National Research Council, Civic Engagement and Social Cohesion: Measuring Dimensions of Social Capital to Inform Policy (Washington, DC: National Academies Press, 2014), https://www.nap.edu/catalog/18831/civic-engagement-and-social-cohesion-measuring-dimensions-of-social-capital. The recommendations are summarized on page 8 of the report.
37 See Current Population Survey, September 2017: Volunteering and Civic Life File, “Attachment 8: Volunteering and civic life supplement,” technical documentation, CPS–17, (U.S. Census Bureau, 2017), pp. 8-1–8-10, https://www2.census.gov/programs-surveys/cps/techdocs/cpssep17.pdf.
38 Conversations of coauthor Sorrentino with Anthony Nerino of the Corporation for National and Community Service (CNCS). For the 2017 volunteer rate from the CNCS data, see “Rankings” (Washington, DC: Corporation for National and Community Service), https://www.nationalservice.gov/serve/via/rankings.
39 Richard B. Freeman, “Working for nothing: the supply of volunteer labor,” Journal of Labor Economics, vol. 15, no. 1, part 2: essays in honor of Yoram Ben-Porath, January 1997, pp. S140–S166; see p. S143.
40 Giving in Retirement: America’s Longevity Bonus: A Merrill Lynch Retirement Study Conducted in Partnership with Age Wave (Charlotte, NC: Merrill Lynch, Bank of America Corporation, 2015), https://agewave.com/wp-content/uploads/2016/07/2015-ML-AW-Giving-in-Retirement_Americas-Longevity-Bonus.pdf.
41 Stephanie Boraas, “Volunteerism in the United States,” Monthly Labor Review, August 2003, pp. 3–11. Data on detailed age breakdowns from more recent surveys can be downloaded from the Census Bureau at https://www.census.gov/data.html.
43 The cited data for both sexes combined are from tabulations made available by ATUS staff. For additional charts on volunteers from the ATUS, see “Charts by topic: volunteer activities” (U.S. Bureau of Labor Statistics, 2015), https://www.bls.gov/tus/charts/volunteer.htm.
44 Manual on the Measurement of Volunteer Work, p. 20.
45 See “Normal aging vs. dementia” (Toronto, Ontario: Alzheimer Society of Canada, 2018), https://alzheimer.ca/en/Home/About-dementia/What-is-dementia/Normal-aging-vs-dementia.
46 Fergus I. M. Craik and Joan McDowd, “Age differences in recall and recognition,” Journal of Experimental Psychology: Learning, Memory, and Cognition, vol. 13, no. 3, July 1987, pp. 474–479.
47 “Helping the ‘old-old’—possible changes to Social Security to address the concerns of older Americans,” Issue Brief (American Academy of Actuaries, June 2016), https://www.actuary.org/sites/default/files/files/publications/OldOld.IB_.6.16.pdf.
49 Abraham and Mackie, Beyond the Market, p. 148.
50 “Alzheimer’s disease: facts and figures” (Chicago, IL: Alzheimer’s Association, 2018), https://www.alz.org/media/HomeOffice/Facts%20and%20Figures/facts-and-figures.pdf
51 Unpaid Elder Care in the United States—2015–16: Data from the American Time Use Survey, USDL 17-1292 (U.S. Department of Labor, September 20, 2017), https://www.bls.gov/news.release/archives/elcare_09202017.pdf.
52 “Current estimates and next steps: the scope and scale of global volunteering,” a background paper for the 2018 State of the World Volunteerism Report: The Thread that Binds, (United Nations Volunteers, 2018), Executive Summary, https://www.unv.org/sites/default/files/The%20Scope%20and%20Scale%20SWVR2018%20final.pdf.
53 ILO Department of Statistics, “Counting the volunteers the world counts on” (Geneva, Switzerland: International Labour Organization, 2020), https://ilostat.ilo.org/counting-the-volunteers-the-world-counts-on/.