The deindustrialization of the United States has been widely considered to be a major force in shaping the economy. It’s one thing to measure where decline has been greatest but where has manufacturing survived or even grown? I use Bureau of Labor Statistics data on manufacturing jobs by county for 1967 and 2014. The results were so surprising that I at first could not believe it.
In 1967 the US had 19,423,000 manufacturing jobs, 25% of an employed labor force of 76 million, while in 2014 there were 11,900,000 such jobs, constituting only 8.3 % (that is one-third of the 1967 share). Almost 12 million is still a lot of jobs, and higher productivity probably means that the sheer amount of stuff produced may not have fallen, but the role of manufacturing in employment has certainly shrunk and as we shall see, greatly relocated.
I reproduce a large table, because it is so interesting, indeed so astounding. There are three sections, first counties with over 25,000 manufacturing jobs in 2014 ( there were far more in 1967), then counties with over 50,000 jobs in 1967, but under 25,000 in 2014, and third, a few counties with over 4000 manufacturing jobs in 2014, and where these were a high share (over 40%) of the local labor force. These were the some of the winners from geographic relocation. I also map these changes. The maps include three additional sets of counties: counties with between 10 and 25,000 jobs in 2014, counties with between 25 and 50,000 jobs in 1967, and counties from 33 to 40% in manufacturing in 2014. These groups are summarized in Table 1.
Table 1: Manufacturing Change 1967-2014 (Measured in 1,000s)
Set
# of Counties
Character
2014 jobs
%
1967 jobs
%
Change
%
% Change
1A
19
> 25k in 2014, gain
1,102
718
385
54
1B
50
> 25k in 2014 loss
2,616
6,698
-4,082
-61
2
26
> 50k in 1967
435
2,828
-2,403
-85
3 & 6
58
> 33% manuf in 2014
343
232
111
48
4A
65
10 to 25K in 2014, gain
1,164
682
482
71
4B
71
10 to 25k in 2014, loss
1,018
1,909
-841
-44
5
26
25 to 50k, 1967
355
1,029
-674
-66
Mapped
315
7,083
60
13,555
70
-6,472
87
-48
Unmapped
2,835
4,822
40
5,758
30
-976
13
-17
US
3,170
ALL
11,900
19,323
-7,423
-38
The 315 mapped counties include 60% of the 2014 manufacturing jobs and some 70% of the jobs in 1967. It is evident that the counties with high numbers of manufacturing jobs in 1967 bore the brunt of losses from 1967 to 2014. In contrast, the smaller, mostly unmapped counties lost only modestly as a set. Many larger counties did gain or hold steady, largely outside the traditional manufacturing belt of the north, or from older core counties into new growing suburbs, as we shall see. Since the losses in the larger mapped counties are so much higher a share of the total jobs in 1967 than in 2014, we have a yet stronger indication of de-concentration.
I’ll begin with the biggest losers, who are on table 2. Now New York City may be thriving in 2014, but it has utterly transformed from an industrial dominance to a minor backwater -- the four boroughs dropping their industrial employment from almost 900,000 to a paltry 67,000 jobs, a drop of 92.5%. In New York County (Manhattan) the fall was even more precipitous: 96%. This is not a misprint. Do not turn off your computer! These are joined by an 84% decline for the New Jersey suburbs: 416,000 to 65,000. Philadelphia, greater Boston, St. Louis, and, yes, especially Baltimore, city and county, experienced the same kind of precipitous decline. Can we begin to understand the basis for riot and unrest in these core cities, whose manufacturing departed as soon as integration opened manufacturing jobs to black workers! As a set, these counties lost 2.83 million manufacturing jobs, a drop of 85%.
Table 2
Set 1: More than 25,000 Manufacturing Jobs in 1967
County
Manuf Jobs 1967
Manuf Jobs 2014
Change
% Change
United States
19,323,000
11,900,000
-7,423,000
-38.2
Snohomish County, Washington
16,000
60,156
44,156
276.0
Harris County, Texas
123,000
164,479
41,479
33.7
San Diego County, California
64,000
97,346
33,346
52.1
Maricopa County, Arizona
59,300
91,348
32,048
54.0
DuPage County, Illinois
24,500
53,913
29,413
120.1
Riverside County, California
17,000
41,519
24,519
144.2
Orange County, California
126,000
150,020
24,020
19.1
Waukesha County, Wisconsin
20,000
43,232
23,232
116.2
Elkhart County, Indiana
31,300
53,705
22,405
71.6
Salt Lake County, Utah
26,000
46,402
20,402
78.5
San Bernardino County, California
30,000
46,822
16,822
56.1
Washington County, Oregon
12,000
27,919
15,919
132.7
Ottawa County, Michigan
16,000
31,831
15,831
98.9
El Paso County, Texas
19,000
31,000
12,000
63.2
Pinellas County, Florida
18,000
28,305
10,305
57.3
Fresno County, California
15,500
25,269
9,769
63.0
Bexar County, Texas
26,000
30,474
4,474
17.2
Suffolk County, New York
49,000
51,967
2,967
6.1
Newport News city, Virginia
25,000
26,503
1,503
6.0
Sum of gaining counties
717,600
1,102,210
384,610
54.0
Tulsa County, Oklahoma
39,000
37,197
-1,803
-4.6
Kent County, Michigan
60,000
57,371
-2,629
-4.4
Tarrant County, Texas
76,000
70,421
-5,579
-7.3
Lake County, Illinois
41,000
35,174
-5,826
-14.2
Kane County, Illinois
39,000
30,327
-8,673
-22.2
Bucks County, Pennsylvania
40,000
27,061
-12,939
-32.3
Greenville County, South Carolina
41,000
26,782
-14,218
-34.7
Hillsborough County, New Hampshire
40,000
25,287
-14,713
-36.8
Sedgwick County, Kansas
56,000
40,629
-15,371
-27.4
Alameda County, California
80,000
63,679
-16,321
-20.4
Multnomah County, Oregon
49,000
32,206
-16,794
-34.3
Santa Clara County, California
120,000
100,981
-19,019
-15.8
York County, Pennsylvania
51,000
31,890
-19,110
-37.5
Lancaster County, Pennsylvania
54,000
33,212
-20,788
-38.5
Guilford County, North Carolina
54,000
32,428
-21,572
-39.9
Winnebago County, Illinois
49,000
25,024
-23,976
-48.9
Berks County, Pennsylvania
56,000
29,439
-26,561
-47.4
Miami-Dade County, Florida
58,000
30,387
-27,613
-47.6
Macomb County, Michigan
94,000
59,114
-34,886
-37.1
Hennepin County, Minnesota
109,000
72,307
-36,693
-33.7
Dallas County, Texas
138,000
94,078
-43,922
-31.8
Oakland County, Michigan
94,000
47,243
-46,757
-49.7
Franklin County, Ohio
76,000
28,991
-47,009
-61.9
Jefferson County, Kentucky
90,000
40,666
-49,334
-54.8
Bristol County, Massachusetts
78,000
26,935
-51,065
-65.5
Middlesex County, New Jersey
82,000
28,277
-53,723
-65.5
Essex County, Massachusetts
94,000
38,451
-55,549
-59.1
Jackson County, Missouri
85,000
25,870
-59,130
-69.6
St. Louis County, Missouri
97,000
35,884
-61,116
-63.0
Summit County, Ohio
93,000
27,965
-65,035
-69.9
King County, Washington
146,000
79,631
-66,369
-45.5
Montgomery County, Pennsylvania
106,000
39,566
-66,434
-62.7
Hamilton County, Tennessee
95,000
25,092
-69,908
-73.6
Bergen County, New Jersey
107,000
33,434
-73,566
-68.8
Marion County, Indiana
120,000
42,808
-77,192
-64.3
New Haven County, Connecticut
115,000
31,792
-83,208
-72.4
Montgomery County, Ohio
110,000
26,188
-83,812
-76.2
Erie County, New York
134,000
42,606
-91,394
-68.2
Hartford County, Connecticut
151,000
57,332
-93,668
-62.0
Monroe County, New York
133,000
38,958
-94,042
-70.7
Fairfield County, Connecticut
130,000
35,507
-94,493
-72.7
Hamilton County, Ohio
152,000
45,901
-106,099
-69.8
Middlesex County, Massachusetts
166,000
59,454
-106,546
-64.2
Worcester County, Massachusetts
165,000
34,677
-130,323
-79.0
Milwaukee County, Wisconsin
181,000
48,963
-132,037
-72.9
Allegheny County, Pennsylvania
195,000
36,428
-158,572
-81.3
Cuyahoga County, Ohio
277,000
69,606
-207,394
-74.9
Wayne County, Michigan
396,000
71,526
-324,474
-81.9
Los Angeles County, California
855,000
359,532
-495,468
-57.9
Cook County, Illinois
831,000
181,315
-649,685
-78.2
Sum of losing counties
6,698,000
2,615,592
-4,082,408
-61.0
Table 2, set 2: Over 50,000 in 1967 and Under 25,000 in 2014
Manuf Jobs 1967
Manuf Jobs 2014
Change
% Change
Bronx
NY
59,000
6,000
-53,000
-89.8
Kings
NY
220,000
18,000
-202,000
-91.8
Onondaga
NY
59,000
19,000
-40,000
-67.8
Queens
NY
132,000
22,000
-110,000
-83.3
Westcheste
NY
73,000
12,000
-61,000
-83.6
New York, NY
NY
482,000
21,000
-461,000
-95.6
Lucas
OH
62,000
16,000
-46,000
-74.2
Stark
OH
63,000
23,000
-40,000
-63.5
Philadelphia
PA
264,000
23,000
-241,000
-91.3
Providence
RI
93,000
22,000
-71,000
-76.3
Fulton
GA
65,000
18,000
-47,000
-72.3
Nwcastle
DE
53,000
13,000
-40,000
-75.5
Lake
IN
98,000
23,000
-75,000
-76.5
Baltimore
MD
68,000
11,000
-57,000
-83.8
Baltimoecity
MD
107,000
12,000
-95,000
-88.8
Hampden
MA
65,000
21,000
-44,000
-67.7
Norfolk
MA
58,000
21,000
-37,000
-63.8
Suffolkk
MA
85,000
8,000
-77,000
-90.6
Ramsey
MN
72,000
23,000
-49,000
-68.1
Essex
NJ
124,000
18,000
-106,000
-85.5
Hudson
NJ
107,000
8,000
-99,000
-92.5
Passaic
NJ
83,000
18,000
-65,000
-78.3
Union
NJ
102,000
21,000
-81,000
-79.4
StLouis city
MO
132,000
17,000
-115,000
-87.1
San Francisco
CA
52,100
7,500
-44,600
-85.6
Delaware
PA
59,600
13,000
-46,600
-78.2
2,837,700
434,500
-2,403,200
-85
Table 2, set 3: High Manufacturing Share, Over 4,000 Jobs
Manuf Jobs 1967
Manuf Jobs 2014
Change
% Change
Jackson County, Alabama
3,200
5,196
1,996
62.4
Boone County, Illinois
8,300
7,619
-681
-8.2
DeKalb County, Indiana
4,200
8,128
3,928
93.5
LaGrange County, Indiana
1,200
5,141
3,941
328.4
Noble County, Indiana
4,700
8,351
3,651
77.7
Whitley County, Indiana
2,000
4,541
2,541
127.1
Marion County, Iowa
1,400
6,128
4,728
337.7
Ford County, Kansas
1,000
6,272
5,272
527.2
Pontotoc County, Mississippi
1,100
6,199
5,099
463.5
Scott County, Mississippi
2,000
4,883
2,883
144.2
Alexander County, North Carolina
2,600
3,284
684
26.3
Bladen County, North Carolina
1,000
5,565
4,565
456.5
Auglaize County, Ohio
5,300
7,339
2,039
38.5
Shelby County, Ohio
7,900
10,052
2,152
27.2
Williams County, Ohio
5,900
6,337
437
7.4
Elk County, Pennsylvania
9,400
6,587
-2,813
-29.9
Newberry County, South Carolina
3,700
4,831
1,131
30.6
Titus County, Texas
800
5,865
5,065
633.1
Box Elder County, Utah
2,300
6,206
3,906
169.8
Trempealeau County, Wisconsin
1,200
6,418
5,218
434.8
69,200
124,942
55,742
80.0
The first set of counties include some winner and more losers. The winners grew from 718,000 to 1,102,000 jobs, or up 54%, but this is dwarfed by the colossal loss of 4.1 million out of 6.7 million jobs, a loss of 61% in manufacturing jobs. The losers are somewhat like set 2, just not quite so extreme. Included are the two counties which lost the most—Cook (Chicago) and Los Angeles- 650,000 and 500,000! Other big losses include Wayne (Detroit), 324,000, Cuyahoga (Cleveland), 207,000, Milwaukee, 132,000, Hamilton (Cincinnati), 106,000, Allegheny (Pittsburgh), 159,000, and Worcester, MA, 136,000.
The gaining larger counties are the beneficiaries of two forms of de-concentration – from the north to the south and west, and from older core counties to their suburbs. Growing industrial centers in the south and west include Harris (Houston), San Diego, Maricopa (Phoenix), Fresno, Bexar (San Antonio), Salt Lake, and Pinellas, FL (St. Petersburg), but as or more important is the growth of suburbs, notably Orange, CA, Suffolk, NY (way out there), San Bernardino-Riverside, Waukesha, WI, Washington, OR, and the biggest winner of all, Snohomish, WA, where Boeing builds big jets, and the home of the late Senator Henry Jackson. This leaves two growing smaller metro areas of the north: Ottawa, MI, and Elkhart, IN, one of the fastest growing and most successful examples of manufacturing and income growth.
Sets 1 and 2 represent the larger manufacturing cores of 1967, 2014 or both. But in 1967 they held 57% of all manufacturing jobs, while in 2014, their share dropped to 35% (10.3 million versus 4.2 million), again illustrating the basic geography of de-concentration.
Sets 3 and set 6 counties, with high manufacturing shares in 2014, include many successful micropolitan or suburban counties in all regions. A few counties with high manufacturing shares in 2014 are suburban, often to smaller metropolitan areas, e.g. Scott, KY, (Lexington). Many more are exurban to medium sized metro areas, as to Springfield, MO, Raleigh, NC, Des Moines, IA or Jackson, MS, and especially 3 counties in northeastern IN, in exurban territory beyond Ft. Wayne and Elkhart.
Some success stories are in more remote small town areas, as AL, AR, TN, MO, OH, SD, NC, SD, TX, and KS, for example, Ford County (Dodge City) and McPherson (Hutchinson Space Center).
Set 4 counties, with 10,000 to 25,000 manufacturing jobs in 2014, again include both losers (71 counties, losing 841,000 jobs, or 44%) and winners, gaining 682,000 jobs, or 71%. Losses are not so severe as for the sets 1, 2, and 5 counties, but are still significant, as in PA, 6 more counties, OH, 3 more, NJ, 3 more, MI, 3 more, and 1 each in MN, CT, IN, KS, CO (Denver), and also several in the south, as in AL (Jefferson-Birmingham), TN, (Shelby-Memphis) and Knox, and NC, 2 counties.
Counties gaining the most include 6 TX counties, Travis (Austin), 2 Houston suburbs, 2 Dallas suburbs, and Potter (Amarillo), 5 CA counties, Kern and Merced in the central valley, and suburban Sonoma, Ventura and Napa, 3 Atlanta suburban counties, 3 UT counties, suburban or exurban to Salt Lake, 2 in CO, Weld (Greeley) and Larimer (Ft. Collins), 2 in LA, and in OH (exurban and small town in the west of the state). Thus almost all are large metro suburbs or smaller independent metro counties. From the list it is clear that these counties well represent the twin trends of suburban-exurban spillover or relocation, as well as the broader de-concentration from the north to the south and west.
Several set 5 counties, 25,000 to 50,000 jobs in 1967, are also in the set 4 list (10,000 to 25,000 job in 2014), often with significant losses. Some with even higher losses, to under 10,000 manufacturing jobs in 2014, include counties in IL, IN, LA (Orleans), MI, NJ, NY (4 more), and PA (4 more).
What do the maps tell us?
The preceding discussion has probably induced the curious reader to peruse the maps to find places of decline versus growth. The maps show data for only 10 percent of US counties, 315 of 3170. Yet these contained 70% of manufacturing jobs in 1967, and 60% in 2014.
The 1967 and 2014 maps of jobs in manufacturing depict the broad distribution of loss. Although the sheer density and size of places in the traditional industrial belt of the north stand out, a few big losses in the west, notably Los Angeles and San Francisco, appear. But the rests of the south, the plains and the west suggest a widespread if modest expansion, often in proximity to larger declining counties.
The pattern of change from 1967 to 2014 displays the patterns of change in numbers and rates of growth versus decline. Losses are largest and almost continuous from Detroit east to Boston, while in the south, the Midwest and west, the big losers are in older, long standing large centers, Like, LA, SF, New Orleans, St. Louis, Minneapolis, Chicago, Cincinnati, and Indianapolis. These are interspersed with growing centers of manufacturing in TX, across the west, but also quite prominent in suburban and exurban and new industrial places in the south and Midwest, e.g. MS, AL, GA, TN, LA and AR, but in substantial numbers in different areas of OH, IN, MI, WI, IL and MN, KS and MO. While the growth in the burgeoning west and TX might be an expected product of sheer population, and located both in suburbs, as around LA, SF, Portland and Seattle, much was in new independent place such as Boise, Phoenix, Tucson, Salt Lake, Greeley and Ft. Collins, Reno and Las Vegas. In contrast, a pattern of core decline but impressive suburban-exurban growth occurred in parts of the Midwest, in MI, IN, OH, WI, MN, and MO.
Conclusion
Yes, the decline of manufacturing as a dominant part of the labor force is large and rea. But America still makes a lot of stuff, much in quite different places, so that there is no longer a distinctive industrial belt, but in a more dispersed pattern. Many of the older centers of manufacturing, like NY, Boston, Philadelphia, and Chicago and Los Angeles, have long since transformed into world centers of services, while others, like Pittsburgh, Detroit and Cleveland appear to be in a process of transformation.
Some may view this transformation and the huge decline in manufacturing jobs as a benign market effect in which the US specializes in services while much of making things is out-sourced to lower cost countries. But in much of the real America far too few equivalent middle class jobs have replaced the lost jobs. Perhaps what the US needs now is serious innovation in making new kinds of things, and bring manufacturing up to 19 million and beyond! Instead I suspect the ever-wise market will innovate with robots, presaging a time when the country will complete its transformation to an owner and servant society.
Richard Morrill is Professor Emeritus of Geography and Environmental Studies, University of Washington. His research interests include: political geography (voting behavior, redistricting, local governance), population/demography/settlement/migration, urban geography and planning, urban transportation (i.e., old fashioned generalist).