NALP Bulletin+
July/August 2022

NALP's research on graduate employment data in the legal profession has helped keep a spotlight on trends in legal hiring and employment over the past three decades. This article analyzes 30 years of data collected in NALP's Employment Report & Salary Survey, comparing it with U.S. Census data from 1990 to 2020.

The combination of 30 years of granular and accessible graduate employment data, for the Classes of 1991-2020 — book-ended nicely by four decennial U.S. Censuses, 1990-2020 — provides a new opportunity to assess city-level trends in law firm job opportunities, and population changes from various perspectives:

• Changes in the number of law firm job opportunities, by city, to see if the number has gone up, down, or stayed about the same. All these scenarios are represented in the 43 cities included in this study, some in perhaps unexpected places. We can also determine if changes have been generally positive, generally negative, or something in between.

• Population changes in these cities, which run the gamut: from up 150% to down by more than 30%.

• Job density — that is, the number of law firm jobs taken relative to the population size, and how that has changed over 30 years. Again, the changes range from large decreases in "density," to very little change and large increases. As we shall see, density measures and changes in density have little to do with population size, and how that has changed.

This study examines 43 cities, each of which meets at least one of the following criteria:

• The city was among the 25 largest in the U.S. based on the 1990 Census.
• The city was among the 25 largest in the U.S. based on the 2020 Census.
• The city provided at least 90 law firm jobs to Class of 1991 graduates.
• The city provided at least 90 law firm jobs to Class of 2020 graduates.

The Technical Appendix lists the 43 cities according to the criteria/criterion met.

Net Job Changes

Looking first at net job changes over the 30-year period, there is a large range, from numbers more than tripling, to losses of more than half (see Table 1). Of course, impressive gains can result from small numbers, such as in Irvine and Wilmington, DE. Wilmington presents a unique case: the number of jobs more than doubled, even as the population numbers decreased just a bit. However, even the largest market by far, New York City, saw job numbers increase by 44%. Numbers in the second-largest market in 2020, Washington, DC, were up by 25%. In contrast, the number of jobs taken in Chicago and Cleveland are quite flat over the period, after an increase leading up to 2000. Notably, these two cities are alone in sustaining job numbers, even as population fell, in the case of Cleveland by 25%.

Sixteen cities experienced a net job loss of more than 10% over the period. Cities experiencing the greatest job losses in percentage terms - Baltimore and Hartford — have both experienced population losses since 1990. But population loss is neither a necessary nor a sufficient condition for job loss. Thus, despite substantial population growth (by half or more), job numbers fell in Seattle, Denver, and Phoenix. In San Francisco and Los Angeles, job numbers fell even as population increased by 20% and 14%, respectively. [Table 2 ranks the 43 cities by change in population from 1990 to 2020.]

Just looking at change over the entire period obscures considerable variation in the roads from 1991 to 2020. Relatively few cities had a steady increase, at least as measured by these 10-year segments. Likewise, relatively few cities experienced a steady decline. Table 3 groups cities according to patterns of change from 1991-2020.

The first group of cities saw a net increase in jobs over each period, or at least a flat number, with the 2020 count being the highest of the four end-of-decade counts. In general, the greatest increases occurred in the 1991-2000 period, with much smaller increases in each of the next two time periods. Houston is the exception here, with essentially no net change from 1991-2000.

In the next group of cities, the 2020 count was also the highest of the four end-of-decade counts but came after a net drop from 2000 to 2010, followed by strong growth from 2010 to 2020. For example, Charlotte NC and Wilmington DE both showed large gains during the 2010-2020 period. Job numbers in these two cities more than doubled over the 30-year span.

The next, much larger group of cities (12) experienced a net job gain from 1991-2020, but the gains were neither steady, nor was the 2020 count the highest of the four end-of-decade counts. Except for San Diego, all these cities saw healthy gains from 1991-2000, with Palo Alto leading the pack, followed by Memphis, Jacksonville, and Austin, albeit these are based on smaller numbers. Larger employment centers — Washington, DC, Boston, and Atlanta — all saw net gains of one-third to one-half from 1991-2000. Each of the cities in this group experienced a net loss over at least one of the next two time periods (2000-2010 and 2010-2020), and some — Memphis, Austin, and Fort Worth — did in both time periods.

Job numbers were essentially flat — that is, changing by three jobs or less, or less than 2% — in four cities, despite measurable job growth from 1991 to 2000.

A group of 11 cities experienced net job loss over the period, but the lowest job number was not necessarily in 2020. All these cities experienced net job growth from 1991 to 2000, except Denver and Los Angeles. Figures were essentially flat in Seattle from 1991 to 2000. Note that each of these cities saw population growth, in the case of Seattle and Denver of about 50%. Few cities saw job growth from 2000 to 2010, and in each case the gain was wiped out over the next decade. Every city in this group, except Kansas City and New Orleans, experienced a double-digit job loss over the 30-year period. In Indianapolis, the number of jobs fell by one-third.

In the final group of cities, job numbers fell in each time period, or were essentially flat. Baltimore and Hartford fared the worst, with job numbers down by almost two-thirds from 1991 to 2020 in each city. Explosive population growth in Phoenix was not accompanied by job growth. As will be discussed later, such opposite forces have resulted in big shifts in job numbers relative to population size.

First, however, it should be noted that all 43 cities experienced ups and down in job counts in the years in between the decade markers. For example, most cities experienced their highest job counts in the 2000-2009 period, with a decline in 2010, the year the job market largely reacted to the Great Recession. However, this was not the case in Miami, Tampa, New Orleans, Houston, San Antonio, Irvine, and Seattle, all of which maintained at least essentially flat numbers or some growth in their 2010 numbers. Cities with the largest reported job numbers in 2011 or later include Miami, Tampa, Nashville, San Antonio, and Wilmington. (Table 4 shows the high mark for reported job numbers and the year in which it occurred.)

Let's look at a few case studies of year-by-year counts. Houston and Miami both experienced substantial growth from 1991-2020 (see Figure 1). Houston notched high job counts in 2003 and 2019 (about 545) followed closely by 2013 and 2020 (about 530). Job numbers in Miami followed a general upward trajectory through 2011, reaching a high in 2011. Figures then declined, but have come up in recent years, close to pre-recession levels. Compared to some rather dramatic ups and downs in Houston, those in Miami are far more modulated.

Philadelphia PA and Wilmington DE are geographic neighbors, with very different job number arcs (see Figure 2). Philadelphia's high job counts occurred in 1991 and 2006 and have been on a generally downward trajectory since, despite upticks at time in the 2010's that were not sustained. The 2020 reported count is the lowest of the thirty-year period. A much smaller market overall, reported jobs in Wilmington climbed fairly steadily before dropping during the recession, down to 1990s levels. That was short-lived, however, and the trajectory in the past 10 years has been generally upward, with the highest number recorded in 2020.

Reported job counts in Chicago reached a high in 2007, followed by several years of decline (see Figure 3). A mid-decade growth spurt was not sustained, and in 2020 the reported number of jobs was essentially where it was 30 years prior. Jobs reported in Los Angeles dropped considerably in 2010, as they did in many other places, and have only partially recovered since. Reported job numbers were highest by far in 1991.

Additional year-by-year comparisons can be found in Figure 3a and Figure 3b below.

Job Density

Finally, we can look at a measure of "job density" — the number of law firm jobs taken by new law school graduates per 10,000 population in the city — first, with cities grouped by population change over the 30-year period, and second with cities grouped by job change over the same period. Either way of grouping cities provides interesting findings.

• Among 11 cities that lost population between 1990 and 2020, the majority maintained a relatively steady job density, or even gained a little — e.g., Chicago, Cleveland, Detroit, Pittsburgh. Baltimore and Philadelphia both lost jobs relative to their population, and Hartford did even more so. Wilmington is an outlier, with job density more than doubling (see Figure 4a).

• Among the cities that saw population grow by up to 25%, job numbers in Boston have kept up, whereas they have not in Los Angeles, as discussed above. Palo Alto stands out here, despite 2020 being far from the high-water mark in terms of reported jobs, which occurred in 2001 (see Figure 4b).

• Among the 21 cities experiencing the most population growth, in only four — Miami, Irvine, Orlando/Winter Park, and Tampa — did job growth substantially outperform population growth. Five areas under-performed based on this measure — Atlanta, Denver, Las Vegas, Phoenix, and Seattle. The rest of the cities experienced smaller or essentially no change. El Paso, Fort Worth, and Jacksonville, FL do not seem destined to ever be major job centers for grads taking law firm jobs, despite being large and growing population centers (ranking 23, 12, and 13, respectively, on population size in 2020). (See Figure 4c and Figure 4d.)

• Grouping cities according to net job change shows that among the cities experiencing the steeped job loss in percentage terms, job density declined in all of them, to varying degrees, and was accompanied by population loss except in Indianapolis and San Jose (see Figure 5a).

• A number of cities experiencing net job loss of 10-25%, nonetheless maintained job density, whereas cities that were growing in population — Denver, Los Angeles, San Francisco, Seattle — did not (see Figure 5b.)

• Cities where job numbers changed by less than 10% net generally saw relatively small changes in density over the period. Atlanta is an exception, where population increase by about 30% but job growth was about 5.5% (see Figure 5c).

• Finally, among cities with the greatest net job growth, job density was relatively steady in many. Palo Alto and Wilmington are exceptions, as is, to a lesser extent, Miami. Job density in Las Vegas, never high, dropped by about half (see Figure 5d and Figure 5e).

Whether any of these trends offer a prediction as to future outcomes is unknown. It is likely Washington, DC, New York, Chicago, and other large markets will continue to be major employment centers, whereas El Paso won't be, and places such as Hartford and Baltimore could continue a downward trend, or stabilize. Will Wilmington maintain its recent growth? If remote work continues to be reported according to the location of the office employing the graduate, options for remote work should not materially disrupt counts of jobs by city.

An additional wild card is population change, both short-term and long-term. For example, the U.S. Census Bureau recently reported that almost three-quarters of U.S. counties - about 2,300 - experienced natural decrease in 2021, that is the net effect of migration in, migration out, births and deaths. (See "Over Two-Thirds of the Nation's Counties Had Natural Decrease in 2021," U.S. Census Bureau, March 2022.) A number of metropolitan statistical areas (MSAs) that encompass the cities noted here as increasing in population from 1990 to 2020 have recently lost population, including Boston (down 36,500); Los Angeles (down 176,000); Miami (down 35,000); New York (down 328,000); San Francisco (down 116,000); San Jose (down 43,000); Seattle (down 113,000); and Washington, DC (down 29,000). (See Table: "Annual Estimates of the Resident Population in the United States and Puerto Rico: April 1, 2020 to July 1, 2021", U.S. Census Bureau, March 2022.)

Whether such changes signify the start of a longer-term shift or a shorter-term reaction to the COVID-19 pandemic, and a desire to escape crowded cities for some period of time, remains to be seen. As we have seen, population changes do not always track with changes in job numbers.

Table 1. Law Firm Jobs Taken in Selected Cities — 1991-2020
By Change in Number of Jobs, 1991-2020

Note: Orlando figures include Winter Park starting with the Class of 2011. Population figures for 2020 combine the two cities. Job figures for New York City include Bronx and Brooklyn, and so do not match those reported in Jobs & JDs reports, which track Brooklyn separately in all the years shown, and the Bronx separately in 2000, 2010, and 2020. This was done to make the job count coverage correspond to the Census Bureau coverage. Typically, the two boroughs provide 40-45 law firm jobs. Figures in yellow are estimates, not the reported number. See the technical appendix for details.
Source: Compiled from Employment Report and Salary Survey data for 1991-2020.

Table 2. Law Firm Jobs Taken in Selected Cities — 1991-2020
By Change in Population, 1990-2020

Note: Orlando figures include Winter Park starting with the Class of 2011. Population figures for 2020 combine the two cities. Job figures for New York City include Bronx and Brooklyn, and so do not match those reported in Jobs & JDs reports, which track Brooklyn separately in all the years shown, and the Bronx separately in 2000, 2010, and 2020. This was done to make the job count coverage correspond to the Census Bureau coverage. Typically, the two boroughs provide 40-45 law firm jobs. Figures in yellow are estimates, not the reported number. See the technical appendix for details.
Source: Compiled from Employment Report and Salary Survey data for 1991-2020.

Table 3. Law Firm Jobs Taken in Selected Cities — 1991-2020
By Job Change Category

1. Reported job numbers were essentially flat from 1991 to 2000.
2. Essentially the number of jobs reported is unchanged from 1991 to 2020.
3. Number of jobs reported increased by 3 from 2010-2020, from 34 to 37, essentially flat.
4. Decrease in number of reported jobs from 2010 to 2020 was less than 2%, or 5 jobs.
5. Number of jobs reported essentially flat 1991-2000; job numbers estimated for 2010 and 2020, but suggest a drop of approximately 20% in the 2000's.
6. Number of jobs reported essentially flat from 2000 to 2010.
7. Number of jobs reported decreased by 3, less than 2%, from 1991 to 2000, essentially flat.
8. The number of jobs reported was essentially flat from 2000-2010.

Note: Orlando figures include Winter Park starting with the Class of 2011. Population figures for 2020 combine the two cities. Job figures for New York City include Bronx and Brooklyn, and so do not match those reported in Jobs & JDs reports, which track Brooklyn separately in all the years shown, and the Bronx separately in 2000, 2010, and 2020. This was done to make the job count coverage correspond to the Census Bureau coverage. Typically, the two boroughs provide 40-45 law firm jobs. Figures in yellow are estimates, not the reported number. See the technical appendix for details.
Source: Compiled from Employment Report and Salary Survey data for 1991-2020.

Table 4. Year in Which Largest Number of Law Firm Jobs Were Reported

Note: A runner-up is reported if the difference in reported job numbers was 5 or less.
*Numbers for El Paso generally varied by one or two from year to year. Hence every year is essentially a runner-up.
**Estimated. See Technical Appendix for details.

Figure 1. Number of Jobs — 1991-2020 — Houston and Miami

Source: Employment Report and Salary Survey data, 1991-2020

Figure 2. Number of Jobs — 1991-2020 — Philadelphia and Wilmington, DE

Source: Employment Report and Salary Survey data, 1991-2020

Figure 3. Number of Jobs — 1991-2020 — Chicago and Los Angeles

Source: Employment Report and Salary Survey data, 1991-2020

Figure 3a. Number of Jobs — 1991-2020 — Seattle and San Diego

Source: Employment Report and Salary Survey data, 1991-2020

Figure 3b. Number of Jobs — 1991-2020 — Nashville and Tampa

Source: Employment Report and Salary Survey data, 1991-2020

Figure 4a. Job Density — Cities with population loss from 1990-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population. Job counts for St. Louis for 2020 were imputed. See text for details.

Figure 4b. Job Density — Cities with population growth up to 25% from 1990-2020

Note: Job density is defined as the number or law firm job jobs taken by new law school graduates per 10,000 population. 1991 figures for New York were imputed. See text for details. New York job figures also include Brooklyn and Bronx.

Figure 4c. Job Density — Cities with population growth of 25% to 40% from 1990-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population.

Fig 4d. Job Density — Cities with population growth exceeding 40% from 1990-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population. 2020 figures for Las Vegas were imputed. See text for details.

Figure 5a. Job Density — Cities with job loss greater than 25% from 1991-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population.

Figure 5b. Job Density — Cities with job loss 10-25% from 1991-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population. Detroit figures for 1991 and St. Louis figures for 2020 were imputed. See text for details.

Figure 5c. Job Density — Cities with Job Change of less than 10% from 1991-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population.

Figure 5d. Job Density — Cities with job growth 20-90% from 1991 -2020

Note: Las Vegas figures for 2020 and New York figures for 1991 were imputed. See text for details. Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population.

Figure 5e. Job Density — Cities with job growth more than double from 1991-2020

Note: Job density is defined as the number of law firm job jobs taken by new law school graduates per 10,000 population.

Technical Appendix

Methodology for Selection of Cities

Cities selected for this study meet at least one of the following criteria:

• The city was among the 25 largest in the US based on the 1990 Census.
• The city was among the 25 largest in the US based on the 2020 Census.
• The city provided at least 90 law firm jobs to Class of 1991 graduates.
• The city provided at least 90 law firm jobs to Class of 2020 graduates.

Study cities, by criteria met

Note: Because job counts for Minneapolis and St. Paul are combined, populations for the two cities are combined. This put the combined entity in a population top 25 list, whereas Minneapolis by itself is not. Therefore, the top population list includes the 25 cities as reported by the Census Bureau, plus Minneapolis/St. Paul, for a total of 26 cities. Cities highlighted in green lost population from 1990 to 2020. In the case of Philadelphia, the change was just -0.4%.

*Winter Park, FL was included with Orlando starting with the Class of 2011.

Sources for population figures:

"(SU-99-1) Population Estimates for Cities with Populations of 100,000 and Greater (Sorted by 1999 Population Size Rank in U.S.): July 1, 1999 (includes April 1, 1990 Population Estimates Base)", Population Estimates Program, Population Division, U.S. Census Bureau, Washington, DC 20233, Internet Release Date: October 20, 2000.

"SUB-EST2020: Subcounty Resident Population Estimates: April 1, 2010 to July 1, 2019; April 1, 2020; and July 1, 2020", U.S. Census Bureau, Population Division, Release date: May 2021. Updated July 2021 with April 1, 2020 Estimates.

Methodology for Imputing Job Counts

For some cities in some years, lack of participation by a school can have a measurable effect in depressing job counts for that city and year. To correct for this, job counts were imputed as follows: The percentage of law firm jobs in the city accounted for by the school was calculated based on years around the years that the school did not report. For example, if 2000 reporting was missing, the percentage of law firm jobs accounted by the school was calculated for 1996-1999 and 2001-2006. This percentage was generally fairly consistent, and was then applied to the reported number for the year with missing data to determine an imputed number. For example, if 500 jobs were reported as taken in the city in 2000, but school x, which typically accounts for about 15% of the law firm jobs taken in the city, did not report in 2000, then 0.85x = 500 and solving for x results in an imputed number of 588 jobs taken in the city in 2000. Figures that have been imputed in this way are highlighted in yellow in the tables.

Other adjustments were made to conform job count area to the population count area. Orlando figures include Winter Park starting with the Class of 2011, therefore population figures for 2020 combine the two cities. Job figures for New York City include Bronx and Brooklyn, and so do not match those reported in NALP's Jobs & JDs: Employment and Salaries of New Law Graduates reports, which track Brooklyn separately in all the years shown, and the Bronx separately in 2000, 2010, and 2020. Combining the job counts for the three boroughs was done to make the job count coverage correspond to the Census Bureau coverage, which includes all five boroughs. Typically, Bronx and Brooklyn provide 40-45 law firm jobs.

Caution on Attributing Change Based on Small Numbers or Percentages

It is likely that any given count of jobs reported by city is off by a small amount, for any number reasons, including the graduate not revealing their job location, or any information at all, and simple inadvertent errors. For that reason, the job count is referred to the "reported" number of jobs, and small changes in count from one period to the next, for example 3, or on a larger base, 1.5% are described as "flat" or "essentially flat."