How Many Social Workers Are There?

How many social workers are there? It depends on the definition of “social worker.” There is, for example, a proposed difference between today’s social work profession and genuine social work. That difference is not examined here. Staying within the limited scope of today’s profession, this post looks at several sources of information on the question.

Current Population Survey

The U.S. Statistical Abstract drew upon the Census Bureau’s Current Population Survey to provide estimates of the numbers of social workers in the United States. For purposes of that survey, social workers apparently fell into 2010 Census occupation classification code no. 21-1020. The numbers reported by various years’ Statistical Abstracts (S.A.) were as follows:

1983: 407,000 (64% female, 18% black, 6% Hispanic) (2001 S.A. Table 593)

1999: 813,000 (71% female, 24% black, 7% Hispanic) (2000 S.A. Table 669)

2000: 828,000 (72% female, 23% black, 9% Hispanic) (2001 S.A. Table 593)

2002: 848,000 (74% female, 23% black, 9% Hispanic) (2003 S.A. Table 615)

2003: 673,000 (80% female, 20% black, 9% Hispanic) (2004-2005 S.A. Table 597)

2004: 687,000 (78% female, 20% black, 11% Hispanic) (2006 S.A. Table 604)

2005: 670,000 (80% female, 19% black, 10% Hispanic) (2007 S.A. Table 602)

2006: 698,000 (83% female, 23% black, 10% Hispanic) (2008 S.A. Table 598)

2007: 673,000 (82% female, 23% black, 12% Hispanic) (2009 S.A. Table 596)

2008: 729,000 (79% female, 25% black, 10% Hispanic) (2010 S.A. Table 603)

2009: 725,000 (81% female, 23% black, 10% Hispanic) (2011 S.A. Table 615)

2010: 771,000 (81% female, 23% black, 11% Hispanic) (2012 S.A. Table 616)

These Statistical Abstract numbers did not appear highly reliable. For example, it did not seem likely that the number of social workers leaped by nearly 100,000 in three years, from 2007 to 2010. What appeared more likely was that different methods and definitions were being used in different years.

Even so, these numbers did seem to offer some insight. First, they confirmed that the social work profession had become significantly more female since the early 1980s. Also, from 1983 to 2000, it appeared that the number of social workers had doubled.

These numbers reported a severe drop from 2002 (848,000 social workers) to 2003 (673,000 social workers). My cursory investigation did not explain this 21% drop. One possibility was that, at that time, title protection laws came into effect in multiple and/or major states, making it illegal to call oneself a social worker without a license. It appeared unlikely, however, that this effect would be confined to a single year. An alternative explanation would be that the Census Bureau simply changed its working definition of “social worker” or added new job categories that drew off some individuals previously labeled as social workers. Either way, it tentatively appeared that retroactive use of today’s Current Population Survey methodology would have produced a lower estimate of perhaps 350,000 social workers in 1983.

It appeared, in addition, that the profession went from an already high proportion of females in 1983 to become overwhelmingly female by the mid-2000s. The foregoing figures suggested that this trend might have plateaued by 2009 if not earlier. Those figures also say some thing about minorities in social work, as discussed in another post.

The Statistical Abstract ceased publication after 2012, so it was not clear what numbers might have been reported for 2013 and 2014. The rise from 2003 or 2004 to 2009 or 2010 suggests growth of around 1% to 2% annually. Estimating compound annual growth of 1.5% since 2003 would produce values of 747,000 for 2010, 770,000 for 2012, and 793,000 for 2014. The 747,000 figure for 2010 is somewhat less than the 771,000 reported for that year in the 2012 Statistical Abstract, but perhaps that is for the best: given the values from the preceding years, it appears likely that the 771,000 figure would have been downsized in 2011.

It may seem odd that these simple numerical questions had not long since been sorted out and settled. Unfortunately, what was true when Hodge (2004, p. 261) quoted Gibelman and Schervish (1997) apparently remains true at this writing: “We still know very little about the larger population of social workers in the United States.”

Bureau of Labor Statistics

The U.S. Bureau of Labor Statistics (BLS) indicated that there were 607,300 social worker jobs in 2012. This was profoundly different from the 2012 projection of 770,000 social workers (above), based on the values reported in the Statistical Abstract.

It was not immediately obvious how these numbers could diverge so dramatically. After all, like the Statistical Abstracts, BLS claimed that it, too, drew upon Current Population Survey data. My limited investigation suggested that the Survey presented a fill-in-the-blank question of occupation, and then census workers would try to interpret what participants wrote in response. If someone described him/herself as a social worker, it did not appear that census workers would second-guess that. That is, it would presumably not matter if the person lacked a license or a degree in social work: if they described themselves as social workers, that is how they would be reported. Interpretation presumably occurred only when participants wrote something that did not fit within a preexisting Census employment category (e.g., “People tell me about their problems”). Census workers’ interpretations of participants’ ambiguous responses may have been guided in part by (a) a list of similar occupations that would have provided alternative classifications and (b) identification of several subgroups within BLS’s social worker category.

Looking forward, BLS expected the total number of social worker jobs in the U.S. to increase by 19% (more precisely, 18.8%) in the ten-year period between 2012 and 2022. From the BLS estimate of 607,300 social workers in 2012, that projection implied a rise to 721,400 jobs, for a net increase of 114,100 jobs, or about 11,400 per year on average over the 2012-2022 period. That calculation of the year-to-year change was very close to the estimate offered at the end of the preceding section, where the number of social workers was projected to rise by about 23,000 every two years, between 2010 and 2014. It seemed that BLS was suggesting 607,300 social workers in 2012 (above) and a projected total of about 630,000 in 2014.


The U.S. Department of Health and Human Services (HHS) Health Resources and Services Administration (HRSA) estimated (2013, p. 10) that 517,628 Americans reported their occupation as “social worker.” This estimate was based on analysis of American Community Survey data from 2008 to 2010. It seemed that the lower number may have been due in part to those somewhat older data. I did not investigate the question of whether the ACS differed significantly from the Current Population Survey in the questions it posed or its interpretation of responses.


The National Association of Social Workers (NASW) (2006, p. 9) indicated that there were about 310,000 licensed social workers in the U.S. in 2004, and that this number constituted about 38% of all self-identified social workers in the U.S. in 2004. It appeared that this information might be somewhat confused, however. NASW’s cited source for this number was a Census Bureau publication reporting data from 1999. Allowing for rounding error, it appears that this information matched the Statistical Abstract number for 1999 (above). It thus appeared that NASW was misstating the situation as of 2004, and might also be erroneously comparing a 2004 count of licensed social workers (totaling 310,000) against a 1999 count of all social workers, licensed or not.

The numbers shown above indicate that the number of American social workers increased dramatically after 1983. Those numbers also suggest that the growth may have somewhat leveled off by 1999. For purposes of discussion, it is assumed here that NASW meant to say that, as of 1999, there were an estimated 813,000 Americans who identified themselves as social workers, and that 310,000 (~38%) of those people had licenses to practice social work.

The downward adjustment from 2002 to 2003 (above) implies that the Census Bureau figures for years before 2003 would have been lower if the Census Bureau had been applying its revised rules in those years. That is, there was apparently a change significantly reducing the estimated number of social workers in 2003; presumably that change could theoretically have been implemented before 2003; and if it had been so implemented, the figures reported above for the years 2003 would have been reduced to be more in line with the 2003 figures.

Two possible explanations were suggested (above) for the drop from 2002 to 2003. It is unlikely that either of those explanations (i.e., new title protection laws or a tightening of the Census Bureau’s definition of “social worker”) would have removed licensed social workers from the estimate. In other words, it appears likely that the drop in the number of estimated social workers, from 2002 to 2003, occurred largely because some kinds of unlicensed social workers were no longer included in the count. Applying that 21% year-to-year reduction retroactively to 1999, it seems that NASW would have reported 310,000 licensed social workers out of a total of perhaps 642,000 (i.e., 100% – 21% of 813,000) who would have called themselves social workers (or would have been so called by the Census Bureau). In that case, one would also revise NASW’s statement that 38% of so-called social workers had a license. The reported figure of 310,000, in 1999, would constitute 48% of this revised total of 642,000.

After 2002, the Census Bureau totals (above) remained fairly consistent. There did not appear to be any further major adjustments. The post-2002 numbers are not entirely logical, but they do suggest a relatively smooth growth in the number of social workers. As noted above, it could be estimated that there was a total of 793,000 social workers in 2014. If the 48% figure just calculated was approximately accurate and continued to apply, then one might guess that the number of licensed social workers in 2014 would be about 381,000 (i.e., 48% of 793,000).

That, however, might be unlikely. Going back to 1999, it is not clear where NASW got its figure of 310,000. Its citation pointed toward a Census Bureau document. I was not able to obtain a copy of that document on short notice. Its title did not suggest a particular focus on social work. I doubted that the Census Bureau had undertaken a detailed analysis of the number of licensed social workers in the U.S.

Other Sources

There appeared to be difficulties in obtaining accurate numbers. For example, a recent state-by-state count by Donaldson et al. (2014, pp. 54-56), seeking to obtain total numbers of licensed clinical social workers (LCSWs) and their equivalents (given varying state names for that credential), was unable to obtain good counts from four states, including Illinois. As of 2013 or thereabouts, Donaldson et al. counted 201,368 LCSWs. With the addition of data from those four states, the total would probably be nearer to 210,000 licensed MSW holders.

I did not have an immediate source to indicate how many other licensed social workers there might be, in addition to those LCSWs. The National Center for Education Statistics (2012, p. 486) indicated that, in the 2010-2011 school year, there were 15,136 new BSW and 21,173 new MSW degrees awarded. It appeared unlikely that BSWs, generally lacking an objective as appealing as LCSW licensure, would seek licensure at the same rate as MSWs. So one would probably not expect the roughly 1:10 ratio of new MSW graduates to LCSWs (i.e., 21,173 : 210,000) to hold on the baccalaureate level. In other words, the 210,000 licensed MSWs mentioned in the previous paragraph would probably not be joined by another ~150,000 licensed BSWs. Hence, total (master’s plus bachelor’s) license holders would probably number well below 360,000 (i.e., 210,000 + 150,000). This surmise seemed to be supported by ASWB’s report that, at least in 2010, it had administered a total of only 34,085 exams of all types (similarly 2012): assuming substantial numbers of MSWs did take LMSW and LCSW exams, there might not be much room for 15,136 BSWs to LBSW exams.

Yet it was not clear how there could be well below 360,000 licensed social workers. It appeared that the title protection effort had continued over the years – it had apparently become harder to call oneself a social worker without having a license to that effect – so presumably we were not continuing to operate in the circumstances of 1999, when it was estimated (above) that 48% of all social workers (using the Census Bureau’s apparent post-2002 interpretation of the term) were licensed. Instead, as these title protection laws came into force, one would expect a steadily larger percentage of the Census Bureau figures to consist of licensed social workers. By 2014, it seems, perhaps 55% or 60% of the people counted as social workers by the Census Bureau would have licenses. Given a projection of 793,000 social workers that would have been reported in a Statistical Abstract for 2014 (above), it seems that, as of this writing, there could be 476,000 licensed social workers. Alternately, if a much lower estimate of current licensed social workers (e.g., 350,000) was accurate, one might expect the Census Bureau to say that, in 2014, there were only about 580,000 people who would call themselves social workers.


The foregoing estimate of total social workers would be close to the foregoing BLS-based estimate of 630,000 social workers in 2014. If approximately 50-60% of social workers are licensed nowadays, that BLS-based estimate would suggest that, as of 2014, there were roughly 350,000 persons holding some kind of social work license. Assuming NASW’s estimate of 310,000 in 1999 (above) was accurate, the number of persons holding social work licenses would have been growing at a rate of less than 1% per year over the past 15 years.

In short, the best I could do in this brief look was to estimate that, in 2014, there were about 630,000 people who could be called social workers, within BLS’s concept of that term; that about 350,000 of those people had some kind of social work license; and thus that an estimated 56% of self-described social workers were licensed. In addition, the work by Donaldson et al. (above) suggested that perhaps 60% of the licensed social workers were LCSWs, and that only one-third (or less) of self-described social workers were LCSWs. Those last estimates were consistent with a separate post discussing the barriers to obtaining LCSW licensure. Numbers like these, more carefully investigated, could raise questions about such matters as the desirability and form of licensure and the availability of good data on social work career trajectories.

Incidentally, while none of this had anything to do with NASW membership, it seemed that some might be confused about that. Briefly, NASW was an organization that purported to speak for social workers. It was not clear how representative it might be in that capacity. According to Hodge (2004), NASW had 151,000 members (in, presumably, 2003); but at this writing, in 2014, it reported having only 130,000. The organization thus seemed to have slipped, apparently representing about 22% of self-identified social workers in 2003 but (even with an apparent narrowing of “social worker” status due to title protection laws) only about 16% in 2014.

Trackbacks are closed, but you can post a comment.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: