EDIT

The Great Divergence

The Great Divergence refers to the ever growing levels of income inequality in the United States "title": "Great Divergence (Inequality)". In this post, we’re going to look at the extent of this divergence and possible capitalist-based causes. I expect this to be the first in a (non-consecutive) series of posts

Divergence

First, let’s see how we know there is a divergence.

One of the most popular measures of income inequality is the Gini index "title": "Gini coefficient". According to the Census Bureau, this has increased steadily "title": "Historical Income Tables: Income Inequality" since the late 1960s.

{ "type": "scatter", "series": [ { "x": "Year", "y": "Gini", "point_type": "disk", "point_size": 3, "connected": false, "color": "#FF0000" } ], "title": "Gini Index of the United States", "y_axis_label": "Gini", "min_x": 1960, "max_x": 2020, "min_y": 0.35, "max_y": 0.5, "x_tick": 1960, "y_tick": 0.4, "x_tick_size": 10, "y_tick_size": 0.05 }
YearGini
20150.479
20140.480
20130.482
20130.476
20120.477
20110.477
20100.470
20090.468
20080.466
20070.463
20060.470
20050.469
20040.466
20030.464
20020.462
20010.466
20000.462
19990.458
19980.456
19970.459
19960.455
19950.450
19940.456
19930.454
19920.433
19910.428
19900.428
19890.431
19880.426
19870.426
19860.425
19850.419
19840.415
19830.414
19820.412
19810.406
19800.403
19790.404
19780.402
19770.402
19760.398
19750.397
19740.395
19730.400
19720.401
19710.396
19700.394
19690.391
19680.386
19670.397
Note, the y-axis does not start at 0. I often don’t let the y-axis start at 0, because I believe that it’s important to notice how a change has related to past variation.

Of course, this isn’t the only measure. The Census Bureau provides other useful statistics: the ratio between the 95th percentile and the 50th percentile has increased from roughly 2.6 to nearly 3.7 over the last 46 years – an increase of 42% "title": "Historical Income Tables: Households".

{ "type": "scatter", "series": [ { "x": "Year", "y": "Ratio", "point_type": "disk", "point_size": 3, "connected": false, "color": "#FF0000" } ], "title": "Ratios in Income Between the 95th and 50th Percentiles", "title_size": 30, "y_axis_label": "Ratio", "min_x": 1960, "max_x": 2020, "min_y": 2, "max_y": 4, "x_tick": 1980, "y_tick": 3, "x_tick_size": 10, "y_tick_size": 0.5 }
YearMedianTop5Ratio
2015565162144623.795
2014536572065683.850
2013535852051283.828
2013519391960003.774
2012510171911563.747
2011500541860003.716
2010492761804853.663
2009497771800013.616
2008503031800003.578
2007502331770003.524
2006482011740123.610
2005463261660003.583
2004443341571523.545
2003433181541203.558
2002424091500023.537
2001422281504993.564
2000419901452203.458
1999406961420003.489
1998388851321993.400
1997370051265503.420
1996354921195403.368
1995340761130003.316
1994322641098213.404
1993312411046393.349
199230636990203.232
199130126964003.200
199029943947483.164
198928906917503.174
198827225856403.146
198726061809283.105
198624897771063.097
198523618720043.049
198422415685003.056
198320885635003.040
198220171600862.979
198119074552002.894
198017710506612.861
197916461468602.847
197815064420552.792
197713572380002.800
197612686350002.759
197511800321292.723
197411197306002.733
197310512289502.754
19729697265552.738
19719028241382.674
19708734231752.653
19698389218002.599
19687743198502.564
19677143190002.660
This y-axis doesn’t start at 0 either.

As we examine the extremes, the growth in this discrepancy only increases. Between 1979 and 2007, even after accounting for federal taxes and income transfers, the top 1% gained 275% while the bottom 18% gained only 20% "title": "Great Divergence (Inequality)" - a divergence of 218%. The aforementioned Census Bureau data demonstrates that the median income has risen by 8% over the bottom 20th percentile's income. Taken together, this demonstrates that the incomes of the top 1% have increased by 194% more than the median income.

One last graph. The percent of income earned by the top 1% has more than doubled over the past 40 years "title": "Who exactly are the 1%":

I’m going to stop pointing out when the y-axis doesn’t start at 0. You can see it.

Okay, so this divergence exists. Why?

Greedy Corporations

One naïve guess might be that its because corporations are keeping the money for themselves. This can’t be the cause for two reasons.

First, this is just not true. According to the Bureau of Economic Analysis, the proportion of GDP going towards personal income has not reliably changed over time "title": "National Data: GDP & Personal Income":

{ "type": "scatter", "series": [ { "x": "Year", "y": "Ratio", "point_type": "disk", "point_size": 3, "connected": false, "color": "#FF0000" } ], "title": "Labor Compensation / GDP", "y_axis_label": "Labor Comp %", "min_x": 1940, "max_x": 2020, "min_y": 0.5, "max_y": 0.6, "x_tick": 1980, "y_tick": 3, "x_tick_size": 10, "y_tick_size": 0.02 }
YearCompensationGDPRatio
1947129.1243.10.5311
1947.25130.9246.30.5315
1947.5132.9250.10.5314
1947.75137260.30.5263
1948140.6266.20.5282
1948.25142.2272.90.5211
1948.5147279.50.5259
1948.75148.1280.70.5276
1949146.4275.40.5316
1949.25144.9271.70.5333
1949.5143.8273.30.5262
1949.751432710.5277
1950147.2281.20.5235
1950.25153.8290.70.5291
1950.5162.5308.50.5267
1950.75170.3320.30.5317
1951178.8336.40.5315
1951.25185.1344.50.5373
1951.5188.1351.80.5347
1951.75191.7356.60.5376
1952196.4360.20.5453
1952.25197.7361.40.5470
1952.5201.8368.10.5482
1952.75209.6381.20.5498
1953213.4388.50.5493
1953.25216.5392.30.5519
1953.5216.8391.70.5535
1953.75215.4386.50.5573
1954213.4385.90.5530
1954.25213386.70.5508
1954.5213.5391.60.5452
1954.75217.8400.30.5441
1955222.2413.80.5370
1955.25228.4422.20.5410
1955.5234.2430.90.5435
1955.75238.7437.80.5452
1956243.2440.50.5521
1956.25247.9446.80.5548
1956.5250.84520.5549
1956.75256.7461.30.5565
1957260.6470.60.5538
1957.25262.5472.80.5552
1957.5265.1480.30.5519
1957.75263.7475.70.5543
1958260.2468.40.5555
1958.25259.9472.80.5497
1958.5267.7486.70.5500
1958.75272.7500.40.5450
1959280.2511.10.5482
1959.25286.9524.20.5473
1959.5287.4525.20.5472
1959.75290.7529.30.5492
1960299.3543.30.5509
1960.25302.3542.70.5570
1960.5303.25460.5553
1960.75302.7541.10.5594
1961303.7545.90.5563
1961.25308557.40.5526
1961.5313568.20.5509
1961.75319.7581.60.5497
1962326.2595.20.5481
1962.25332.1602.60.5511
1962.5334.9609.60.5494
1962.75338.4613.10.5519
1963343.4622.70.5515
1963.25348.3631.80.5513
1963.5353.46450.5479
1963.75359.6654.80.5492
1964366671.10.5454
1964.25373.5680.80.5486
1964.5380.9692.80.5498
1964.75387698.40.5541
1965393.8719.20.5476
1965.25400.8732.40.5472
1965.5409.3750.20.5456
1965.75421.4773.10.5451
1966434.1797.30.5445
1966.25445.4807.20.5518
1966.5456.7820.80.5564
1966.75464.9834.90.5568
1967471.18460.5569
1967.25476.8851.10.5602
1967.5486.4866.60.5613
1967.75497.4883.20.5632
1968512.2911.10.5622
1968.25525.3936.30.5610
1968.5539.2952.30.5662
1968.75551.7970.10.5687
1969564.2995.40.5668
1969.25578.21011.40.5717
1969.559510320.5766
1969.75606.61040.70.5829
1970616.71053.50.5854
1970.256221070.10.5813
1970.56301088.50.5788
1970.75631.81091.50.5788
1971649.91137.80.5712
1971.25661.81159.40.5708
1971.5672.21180.30.5695
1971.75684.21193.60.5732
1972710.61233.80.5759
1972.25724.71270.10.5706
1972.5738.11293.80.5705
1972.75761.213320.5715
1973787.81380.70.5706
1973.25805.61417.60.5683
1973.5822.81436.80.5727
1973.75844.11479.10.5707
1974860.91494.70.5760
1974.25881.11534.20.5743
1974.5902.81563.40.5775
1974.75916.516030.5717
1975921.21619.60.5688
1975.25934.11656.40.5639
1975.59591713.80.5596
1975.75986.41765.90.5586
19761017.61824.50.5577
1976.251038.41856.90.5592
1976.51061.21890.50.5613
1976.751087.71938.40.5611
19771116.31992.50.5603
1977.251153.42060.20.5598
1977.51185.32122.40.5585
1977.7512212168.70.5630
19781255.82208.70.5686
1978.251302.92336.60.5576
1978.51339.22398.90.5583
1978.7513832482.20.5572
19791427.32531.60.5638
1979.251458.32595.90.5618
1979.51498.82670.40.5613
1979.751539.72730.70.5638
19801579.22796.50.5647
1980.251600.52799.90.5716
1980.51631.228600.5703
1980.7516942993.50.5659
19811745.23131.80.5573
1981.251777.53167.30.5612
1981.51815.43261.20.5567
1981.7518433283.50.5613
19821866.43273.80.5701
1982.251885.13331.30.5659
1982.51905.33367.10.5659
1982.751920.43407.80.5635
19831949.93480.30.5603
1983.251988.63583.80.5549
1983.52031.23692.30.5501
1983.752085.93796.10.5495
19842147.13912.80.5487
1984.252196.840150.5471
1984.52242.64087.40.5487
1984.752283.14147.60.5505
1985232742370.5492
1985.2523654302.30.5497
1985.52406.74394.60.5476
1985.752457.14453.10.5518
19862493.34516.30.5521
1986.252519.14555.20.5530
1986.52557.14619.60.5535
1986.752605.94669.40.5581
198726514736.20.5597
1987.252694.24821.50.5588
1987.52741.84900.50.5595
1987.752810.35022.70.5595
198828615090.60.5620
1988.252926.15207.70.5619
1988.529785299.50.5619
1988.753035.15412.70.5607
198930845527.40.5579
1989.253117.55628.40.5539
1989.53157.65711.60.5528
1989.753211.25763.40.5572
19903278.85890.80.5566
1990.253334.85974.70.5582
1990.53373.96029.50.5596
1990.753383.26023.30.5617
19913395.16054.90.5607
1991.253430.26143.60.5583
1991.53469.26218.40.5579
1991.753513.56279.30.5595
19923597.66380.80.5638
1992.253654.46492.30.5629
1992.53690.76586.50.5603
1992.753741.96697.60.5587
19933747.96748.20.5554
1993.253803.36829.60.5569
1993.538386904.20.5559
1993.753893.57032.80.5536
19943917.27136.30.5489
1994.253997.27269.80.5498
1994.54035.17352.30.5488
1994.7540917476.70.5472
19954141.37545.30.5489
1995.254177.67604.90.5493
1995.54223.77706.50.5481
1995.7542687799.50.5472
19964321.47893.10.5475
1996.254393.58061.50.5450
1996.54455.781590.5461
1996.754517.98287.10.5452
19974599.58402.10.5474
1997.254665.88551.90.5456
1997.54745.48691.80.5460
1997.754848.18788.30.5517
19984950.48889.70.5569
1998.255035.98994.70.5599
1998.55120.99146.50.5599
1998.755204.29325.70.5580
19995298.89447.10.5609
1999.255352.595570.5601
1999.55429.79712.30.5591
1999.755560.29926.10.5602
20005759.9100310.5742
2000.255800.810278.30.5644
2000.5591410357.40.5710
2000.755951.710472.30.5683
20016059.410508.10.5766
2001.256050.210638.40.5687
2001.56033.710639.50.5671
2001.756042.910701.30.5647
20026074.810834.40.5607
2002.256137.710934.80.5613
2002.56161.411037.10.5582
2002.756193.711103.80.5578
20036226.611230.10.5545
2003.256318.711370.70.5557
2003.56403.811625.10.5509
2003.756508.711816.80.5508
20046565.711988.40.5477
2004.256690.312181.40.5492
2004.56823.312367.70.5517
2004.756878.912562.20.5476
20056952.612813.70.5426
2005.257027.412974.10.5416
2005.57142.213205.40.5409
2005.75722513381.60.5399
20067406.913648.90.5427
2006.25745313799.80.5401
2006.57508.513908.50.5398
2006.75764114066.40.5432
20077832.114233.20.5503
2007.257866.814422.30.5455
2007.57902.114569.70.5424
2007.75799214685.30.5442
20088077.914668.40.5507
2008.258077.3148130.5453
2008.58094.9148430.5454
2008.758062.814549.90.5541
20097762.414383.90.5397
2009.257790.514340.40.5433
2009.5777414384.10.5405
2009.75782114566.50.5369
20107801.714681.10.5314
2010.257947.114888.60.5338
2010.58019.115057.70.5326
2010.758077.915230.20.5304
20118209.715238.40.5388
2011.258248.915460.90.5335
2011.5833215587.10.5345
2011.758285.415785.30.5249
2012850415973.90.5324
2012.25855016121.90.5303
2012.58587.616227.90.5292
2012.75879816297.30.5398
20138705.116475.40.5284
2013.258823.716541.40.5334
2013.58880.616749.30.5302
2013.758960.416999.90.5271
20149108.317025.20.5350
2014.259173.817285.60.5307
2014.59299.617569.40.5293
2014.759432.117692.20.5331
20159502.317783.60.5343
2015.259637.617998.30.5355
2015.5974018141.90.5369
2015.759892.418222.80.5429
20169892.618281.60.5411
2016.2510046.518450.10.5445
2016.51017618675.30.5449
2016.7510268.918860.80.5445

Second, even if the proportion of GDP going toward personal income was shrinking, we’re talking about a divergence within the population, not a divergence between the population and corporate incomes. This means we couldn’t use such a fact as a cause of income inequality. On the other hand, we could use such a fact (if it were true) to explain the stagnating median income "title": "Real Median Household Income in the United States".

Income Composition

Fine, so corporation aren’t just being greedy, maybe the problem is that corporations are paying Americans smaller salaries, and distributing those profits to their stockholders? This would still count as personal income, because dividends are income to people.

This is similar to the position popularized by Thomas Piketty’s relatively recent book Capital in the Twenty-First Century "title": "Capital in the Twenty-First Century". Piketty claims that the Great Divergence is a feature of capitalism that can only be solved with state intervention (such as a tax on wealth). This work has political implications, so it’s controversial "title": "Piketty's Numbers Don't Add Up".

If we examine the data provided by the BEA, we see that Piketty is correct in thinking that financial assets tend to grow as a ratio of overall income and GDP "title": "Integrated Macroeconomic Accounts for the United States Table".

{ "type": "scatter", "series": [ { "x": "Year", "y": "Ratio", "point_type": "disk", "point_size": 3, "connected": false, "color": "#FF0000" } ], "title": "Financial Asset / GDP", "y_axis_label": "Fin. Assets / GDP", "min_x": 1955, "max_x": 2020, "min_y": 4, "max_y": 11, "x_tick": 1980, "y_tick": 6, "x_tick_size": 10, "y_tick_size": 1 }
YearGDPFinancial AssetsRatio
1960544.62730.75.014
1961564.32970.45.264
1962605.23094.85.114
1963640.13307.65.167
1964685.83587.95.232
19657433906.45.258
1966809.94061.95.015
1967858.34492.15.234
1968939.34967.15.288
19691018.35117.95.026
19701070.55452.95.094
19711158.36078.35.248
19721275.36907.75.417
19731422.47316.55.144
19741541.476384.955
19751675.78560.35.108
19761857.19510.35.121
19772066.710384.45.025
19782333.411673.15.003
19792587.413245.75.119
19802818.614977.75.314
19813174.316344.15.149
19823338.217720.15.308
1983358419547.75.454
19844002.121558.45.387
19854295.524637.45.736
19864513.427522.46.098
19874829.7297366.157
19885253.132568.66.200
19895593.5354366.335
19905888.236904.76.268
19916085.739736.76.530
19926428.4421916.563
19936726.445588.96.778
19947171.947940.86.685
19957573.553278.37.035
19968043.658132.67.227
19978596.265075.27.570
19989149.373300.18.012
19999698.183277.48.587
200010384.386812.38.360
200110736.889183.68.306
200211050.389586.28.107
200311524.399200.48.608
200412283.5110616.99.005
200513129.21203259.165
200614073.21324899.414
200714460.1143316.79.911
200814619.2134611.19.208
200914343.4139372.99.717
201014915.2146326.59.811
201115556.3149287.99.597
201216358.5158849.29.710
201316829.5173427.510.305
201417651.118280810.357
201518290.3186732.710.209

This 100% increase in financial assets over GDP looks promising in explaining much of the divergence. The argument goes that if financial assets increased by 100%, then the income from these assets probably also increased 100% - contributing to income inequality, because the rich disproportionately derive income from financial assets.

However, let’s look at a different set of statistics (again, from the BEA again) "title": "National Data: GDP & Personal Income".

It turns out, that over the past 40 years, the proportion of personal income earned due to capital (i.e. ownership of stocks, bonds, and other financial assets) has increased from 23.7% to 26.0% - a 10% increase. Technically, this is an increase, but if we combine this with the finding that the proportion of wealth held by the top 1% is about 35% (I cite this later), we find that this accounts for only about 8% of the entire divergence.

What happened? How did a 75% increase in financial assets correspond with an increase of only 10% in capital income? I don’t have a completely solid explanation, but I suspect we are seeing a basic principle of economics: as the supply of funds go up, the interest (and dividends) paid goes down. This idea is supported by the fact that the ratio between nonfinancial assets and GDP didn't see this rise [see chart below]. I hope someone, somewhere analyzes this point in more detail.

{ "type": "scatter", "series": [ { "x": "Year", "y": "Ratio", "point_type": "disk", "point_size": 3, "connected": false, "color": "#FF0000" } ], "title": "Nonfinancial Asset / GDP", "y_axis_label": "NonFin. Assets / GDP", "min_x": 1955, "max_x": 2020, "min_y": 3, "max_y": 5, "x_tick": 1980, "y_tick": 4, "x_tick_size": 10, "y_tick_size": 0.5 }
YearGDPNonfinancial AssetsRatio
1960544.62209.84.058
1961564.32305.74.086
1962605.22417.83.995
1963640.12516.13.931
1964685.82654.23.870
19657432824.83.802
1966809.93061.23.780
1967858.33281.83.824
1968939.336073.840
19691018.33945.43.874
19701070.54263.93.983
19711158.34659.14.022
19721275.35168.94.053
19731422.45897.84.146
19741541.46788.64.404
19751675.77404.64.419
19761857.18136.64.381
19772066.79115.94.411
19782333.410388.14.452
19792587.412020.84.646
19802818.613635.74.838
19813174.315045.94.740
19823338.215815.24.738
1983358416454.74.591
19844002.117670.14.415
19854295.518928.94.407
19864513.420128.34.460
19874829.721436.24.438
19885253.122980.44.375
19895593.524438.44.369
19905888.225158.74.273
19916085.725302.84.158
19926428.4256193.985
19936726.426451.63.933
19947171.927767.33.872
19957573.528976.13.826
19968043.630074.93.739
19978596.232456.33.776
19989149.3348643.811
19999698.137306.63.847
200010384.341204.43.968
200110736.843241.34.027
200211050.346027.44.165
200311524.349147.64.265
200412283.555908.54.552
200513129.2628264.785
200614073.266580.74.731
200714460.167948.94.699
200814619.263088.34.315
200914343.458217.44.059
201014915.259793.74.009
201115556.362181.63.997
201216358.565748.24.019
201316829.571303.64.237
201417651.175320.44.267
201518290.379483.24.346

Capital-Based Income

We’re not quite done with Piketty. We showed that the increase in capital-based income can account for only 8% of the Great Divergence, but we also have to look at the distribution of that capital-based income.

Between 1983 and today, the share of wealth held by the top 1% grew by just 0.8 percentage points (from 33.8% "title": "A New Look at Long-Term Trends in Wealth Inequality in the United States" to 34.6%"title": "Occupy Wall Street And The Rhetoric of Equality"). Recall that 26.0% of personal income is derived from capital in the United States. Thus, this 0.8% change in capital income actually represents just a reallocation of 0.2% change in total income. This accounts for 0.4% of the 194% divergence.

Thus, overall, we’re looking at probably less than 9% of the Great Divergence being caused by the increase or distribution in financial capital.

Taxes

I’m not going to belabor this point, but the average percent of income paid in federal taxes by the top 1% has more-or-less remained steady over the past several decades "title": "In 2013, the Top 1% Will Pay Their Highest Total Tax Rate Since 1979".

Other Factors

This leaves us with just one conclusion. If capital-income can explain at most 9% of the Great Divergence, and taxes can explain essentially none of it, we must conclude that the distribution of labor-based income and demographic factors explain the remaining 91%. I'll be doing a separate post on those.

Works Cited [show]