Rising Corporate Concentration, Declining Trade Union Power, and the Growing Income Gap: American Prosperity in Historical PerspectiveJordan Brennan

February 2016

RisingCorporateConcentration,DecliningTradeUnionPower,andtheGrowingIncomeGap:

AmericanProsperityinHistoricalPerspective

JordanBrennan*

March2016

*JordanBrennanisaneconomistwithUnifor,Canada’slargestprivatesectorlaborunion,andaresearchassociate of the Canadian Centre for Policy Alternatives. E-mail: jordan.brennan@unifor.org.Website:www.jordanbrennan.org.

1

Contents

ExecutiveSummary 2

Acknowledgments 4

ListofFigures 5

PartI:CorporateConcentration,SecularStagnation,andtheGrowingIncomeGap 6

PartII:LaborUnions,Inflation,andtheMakingofanInclusiveProsperity 24

Appendix 48

References 51

2

ExecutiveSummary

The rise of income inequality amidst the deceleration of GDP growthmust rank as two of themost

perplexing and challenging problems in contemporary American capitalism. Comparing 1935–80 with

1980–2013—that is, the Keynesian-inspired welfare regime and, later, neoliberal globalization—the

averageannualrateofGDPgrowthwasmorethanhalvedandincomeinequalitywentfromapostwar

lowin1976toapostwarhighin2012.Howdoweaccountforthisdouble-sidedphenomenon?

The conventional explanations of secular stagnation and elevated inequality are inadequate,

largely because mainstream (“neoclassical”) economics rejects the notion that the amassment and

exercise of institutional power play a role in the normal functioning of markets and business. This

analytical inadequacy has left important causal elements outside the purview of researchers,

policymakers,andthepublicatlarge.

This two-partanalysis investigates someof thecausesandconsequencesof income inequality

andsecularstagnation intheUnitedStates.Usinganalytical tools fromearlyAmerican institutionalism

and Post Keynesianism, I find that two explanatory variables—institutional power and distributive

conflict—play critical causal roles in the shifting patterns of American economic growth and income

inequalitybeginninginthelate19thcenturyandcontinuinguptothepresentday.

In this context, “institutional power” takes two key forms: corporate power, which is a

commodifiedformofpower;andtradeunion,or“countervailing,”power.(Thesovereignpowerofthe

UnitedStatesgovernmentisnotthefocusofthisanalysis.)“Corporatepower”maybedefinedaslarge

firmsoperatinginoligopolisticmarketstructures,while“tradeunionpower”isthecapacityofworkers

toactinconcertthroughalaborunion.

Thekey finding is that the commodifiedpowerof large firmsdepresseseconomicgrowthand

exacerbates income inequality while the countervailing power of organized labormitigates inequality

andproducessignificantinflationarypressure.ThisanalysispresentsnewestimatesofUSmergeractivity

(1895–2013), corporate concentration (1950–2013), and the earnings margins and fixed asset

investmentofthe100largestAmerican-listedfirms(1950–2013).

Theanalysisbeginswithaninvestigationofthecommodifiedpoweroflargefirmsandfindsthat

mergersandacquisitions (M&A) lead to thecentralizationofcorporateownershipmanifested inasset

concentration.Theperiodbetween1990and2013waswitnesstothemostsustainedperiodofmerger

activity in American corporate history and, as a result, asset concentrationmore than doubled, rising

from9percentto21percent.Thereareroughly5.7millionregisteredcorporationsintheUnitedStates,

3

but the 100 largest firms account for one-fifth of total assets, which is a very high degree of

concentration.Increasedconcentrationisalsoshowntoreducecompetitivepressure,increaseearnings

margins(i.e.,marketpower),andinflatethenationalincomeshareoflargefirms.

Because investment in fixed assets is a key driver of GDP growth, the diversion of corporate

resources away from industrial expansion in favor of M&A puts downward pressure on growth and

leavesmorecorporate income inthehandsof largefirms.Withtheriseofstockoptions inthe1980s,

executives were given an institutional incentive to divert income into share price–inflating stock

repurchase,whichincreasestheearningsofexecutivesandexacerbatespersonalincomeinequality.

In the decades between 1950 and the 1970s, investment by the 100 largest firmsmore than

doubled,risingfrom6percentto13percentofrevenue,onlytotrenddownward inthedecadesafter

1980.Thissuggests that large firmsmaybe leadingthestagnationtendenciesof recent timesthrough

fixed asset underinvestment. Stock repurchase was nearly nonexistent in the 1970s but grew in

significance in each subsequent decade, rising from less than 1 percent of revenue in the 1970s to 7

percentin2007.

AnotherfirstinAmericancorporatehistorywasseenin2005,whenthe100largestfirmsspent

moremoneyinstockbuybacks(inflatingtheirstockprice)thanonfixedassetinvestment(replenishing

theirindustrialbase).Largefirmshavealsobeenonabuyingspreeinrecentdecades,plowingenormous

resourcesintoacquisitions.Thus,thecreationofatop-heavycorporatedistributionsimultaneouslyputs

downwardpressureongrowthwhileelevatinginequality.

Thesecondpartof theanalysisdocuments the interplaybetween thecountervailingpowerof

organized labor, inflation, and income inequality from the late 19th century up to the present day.

Mainstreameconomics insists that “market forces”distribute income inaccordancewithproductivity,

but this assertion is rooted in deeply problematic assumptions, concepts, and measurements.

Historically,unionshaveplayedacrucialroleinredistributingfactorincomefromcapitaltolabor(profit

towages)andfromtheuppertothelowerstrataofthepersonalincomehierarchy.

The growth of American labor unions in tandem with the exercise of the strike weapon,

especially from the 1930s to the 1970s, helped create an inclusive prosperity, or “middle class.” The

erosionofunionssincethemid-1950sandthepacificationof theAmericanworkforcesincethe1970s

hascoincidedwithwagestagnation,ashrinkingnationalwagebill,andheightenedincomeinequality.

PostKeynesiantheoryviews inflationastheproductoftheexcessiveclaimsmadebydifferent

groupsover national income. It is in this context that inflationmay validly beunderstood as a power

process insofar as it is nourishedon social conflict and is closely associatedwith the redistributionof

4

incomebetweendifferentincomegroups.Overthepastcentury,USinflationhastendedtoredistribute

incomefromcapitaltolaborandfromtheuppertothelowerstrataofthepersonalincomehierarchy.If

this set of claims is true, then anti-inflationary monetary policy must not only be understood as a

political phenomenon; it must also be viewed as the use of state power to regressively redistribute

income.

Acknowledgments

TheauthorwouldliketothankJosephA.FrancisforsharinghisdatafilesontheHistoricalStatisticsof

the United States and for forwarding a copy of Simon Kuznets’s estimates on American business

investment.ScottAquannohelpedclarifymyunderstandingofheterodoxeconomicthinkingoninflation

andmonetarypolicy,whileJimStanford’sfeedbackwasinstrumentalinrefiningtheargument.

5

ListofFigures

Figure1:AmericanCorporateAmalgamation,1895–2013 10

Figure2:SecularRatesofGrowth:InvestmentandGDP,1920–2013 11

Figure3:BusinessInvestmentinIndustrialCapacity,1946–2014 12

Figure4:Top100Firms:InvestmentDecomposition,1950–2013 13

Figure5:AmalgamationandConcentration,1940–2013 15

Figure6:ConcentrationandProfitShare,1950–2013 16

Figure7:ProfitShareoftheTop100FirmsandCapitalistIncome,1940–2014 17

Figure8:CapitalistIncomeandAmericanIncomeInequality,1913–2014 19

Figure9:StockPriceInflationandAmericanIncomeInequality,1960–2013 21

Figure10:OrganizedAmericanLaborStrengthandtheNationalWageBill,1900–2013 29

Figure11:AmericanLaborDisobedienceandLaborRemuneration,1870–2013 31

Figure12:LaborDisobedienceandClassIncomeRedistribution,1880–2013 32

Figure13:AmalgamationandMarketPower,1940–2013 35

Figure14:MarketPowerandCapital-LaborRedistribution,1940–2013 36

Figure15:AmericanLaborRemunerationandInflation,1850–2013 37

Figure16:TheEarningsofLargeFirmsandInflation,1940–2013 39

Figure17:AmericanInflationandLabor-CapitalRedistribution,1930–2013 41

Figure18:“ClassStruggle”andIncomeEquality,1900–2013 42

Figure19:The“NaturalRightofInvestment”andIncomeInequality,1937–49 45

6

PartI:CorporateConcentration,SecularStagnation,andtheGrowingIncomeGap

Themoderncorporationhaswroughtsuchachangeinthefreemarketsystemthatnew

conceptsmustbeforgedandanewpictureofeconomicrelationshipscreated.

—GardinerC.Means(1983)

Commentaryoncontemporaryeconomicaffairs,notablythedeepstagnationandsoaringinequalitythat

plagues many advanced industrial societies, leaves the impression that there are important causal

elementsthatremainoutoffocus.High-resolutionexplanationsofAmericanincomeinequalitysuchas

technological change,globalization,andexecutive compensationpracticesarenecessaryandvaluable,

butwhatseemstobemissingfromthedebatearethelow-resolution,butpanoramicallywider,factors

behindeconomicgrowthandthedistributionofitsbenefits.1Conspicuouslyabsentfromthediscussion

isrecognitionthatgrowthandinequalityareshaped,inpart,byinstitutionalpower.

Conventional economic thinking treats power as something that resides outside the normal

functioningofmarketsandbusiness.Powerisapoliticalcategory,saysthemainstreameconomist,not

aneconomiccategory.2Thecommitmentbymainstreameconomiststoignorepowerwasn’talwaysthe

case,especiallyintheUnitedStates.InTheModernCorporationandPrivateProperty,forexample,Adolf

BerleandGardinerMeans(1968[1932])arguedthattheemergenceofthemoderncorporationushered

inarevolutionsufficientlytransformativethatitsimultaneouslyalteredthesystemofprivateownership

and rendered neoclassical economic doctrine obsolete. For Berle and Means, the concentration of

corporateassets—power,inotherwords—alteredthetheoryandpracticeofAmericancapitalism.

InwhatfollowsIexplorethedeephistoryofAmericanGDPgrowthandincomeinequality,and

find that both are shaped to a remarkable extent by the institutional power of large firms. For the

purposes of this study, “corporate power” will refer to large firms operating in oligopolistic market

structures.Largefirmshavetwobroadgrowthpathwaysopentothem:buildnewindustrialcapacityor

1Thesethreetheseshavetendedtodominatethedebate.Thefirstexplanatoryschemefocusesontechnologicalchangeandtheresultingalterationindemandforcertaintypesoflabor.Theskill-biasedtechnologicalchange(SBTC)thesis,asitisknown,suggests that technological change increases thedemand for high skilledworkers anddepresses thedemand for low skilledworkers,thusexplainingheightenedwageinequality(apositionendorsedbyMooreandRanjan[2005]andcontestedbyCardand DiNardo [2002]). Others identify globalization as the culprit behind American inequality (Krugman 2008). Heightenedinternationaltrade,particularlytheimportofmanufacturedgoodsbydevelopedsocietiesfromdevelopingsocieties,altersthewage distribution in rich societies. As developed countries importmore labor-intensivemanufactured goods from low-wagecountries,downwardpressureisexertedonthewagesoflesseducatedand/orlowerskilledworkers,thuselevatinginequality.A third candidate explanation centers on executive compensation practices. Gabaix and Landier (2008) argue that surgingexecutive compensation closely tracks firm size, and because pay is linked to performance, the growth of the former isexplainedbythegrowthofthelatter.2ThisproblematizationofpowerforthestudyofeconomicswasbroughttomyattentionbyNitzanandBichler(2009).

7

purchase it in themarket for corporate control (NitzanandBichler2009). Fixedasset investment (the

former process) accelerates GDP growth and disperses corporate ownership, while mergers and

acquisitions (the latter process) depresses growth and concentrates corporate ownership. The

concentrationof corporate assets leads to the centralizationof national income in thehandsof large

firms, some of which gets paid to capitalists in the form of dividends and some of which is used to

repurchase stock, thus inflating executive compensation. Dividend payments and executive

compensationaretwokeydeterminantsofAmericanincomeinequality.IntheUnitedStates,then,the

processes that fuel the growth of corporate power put downward pressure on GDP growth and

exacerbateincomeinequality.

Thisargumentispresentedinfivesections.ThesecondsectionexploresthehistoryofAmerican

mergers and acquisitions (M&A) to determine if there is any relationship between corporate

amalgamation and power. The key finding is that the quarter century since 1990 has seen an

unprecedentedsurgeinM&Aactivity,andbecausecorporateconcentrationisdrivenbyamalgamation,

asset concentration has soared to a postwar high. The third section disaggregates American business

investmentandfindsthatheightenedM&Aactivityisassociatedwithfixedassetunderinvestmentand,

consequently, slower GDP growth. The fourth section explores the relationship between corporate

concentrationand income inequalityand findsa strong, albeit indirect, relationshipbetween the two.

Thefifthsectionconcludesbysummarizingthekeyfindings,namelythatarelativeincreaseinthesizeof

largefirmshasdepressedgrowthandintensifiedinequality.Adetailedexplanationofthedatasources

andestimationtechniquesistobefoundintheAppendix.

ABriefHistoryofMergersandAcquisitions

Theacquisitionofcorporateorganizations throughM&A isa formofmarketexchange,but it isunlike

othermarketsinafewcrucialrespects.First,wenormallythinkofacommodityassomethingproduced

for sale on a market (Polanyi 2001 [1944]), but corporate organizations are not produced in the

conventionalsenseoftheterm,norarethereestablishedmarketplaces forthemtobeexchanged—at

leastnot in theordinarysenseof the term“marketplace.”Second,commoditiesare typicallyacquired

foroneoftwopurposes:eitherasinputsinaproductionprocessorfordirectconsumption.Acorporate

organizationdoesnot fiteitherpurpose.Third,and finally, theacquisitionofa corporateorganization

hasanunusualproperty insofaras ithasthepotentialtoeliminatemarketsasabasisforexchange. In

otherwords,corporateamalgamationisaformofmarket-destroyingmarketexchange.Theseaspectsof

corporate amalgamation create puzzling questions. Why do business owners engage in this type of

8

market exchange?Andwhat are someof the long-termconsequencesof corporate amalgamation for

theAmericanpoliticaleconomy?

ThenarrativearoundthedevelopmentofM&Afromthe late19th to theearly21stcentury is

one of a series of “waves,” each leading to different organizational forms and market structures

(McCarthy2013;JoandHenry2015).ThefirstUSmergerwavebeganafterthedepressionof1883and

lasteduntil1904.Themajor formthatM&Atookwas“horizontal,”meaningthat firmscombinedwith

competitorsintheirownindustriestoformmonopolisticmarketstructures.USSteel,forexample,was

formedwhenJ.P.MorganconjoinedCarnegieSteelwithhisFederalSteel.Bytheendofthefirstmerger

wave,USSteelcontrollednearlyone-halfoftheUSsteelindustry,havingcombined785separatesteel-

makingunits.Morganwantedtodislodge“aggressivecompetitormanagers”andreplacethemwithan

“orderlymarket” (Gaughan 2007). In practice, this meant restraining price competition, which would

produceamoreproprietor-friendlydistributionofincome.

ThesecondUSmergerwavelastedfrom1916to1929andwaschristenedthe“oligopolywave”

by Nobel laureate George Stigler (1950) because vertical mergers—combinations in the same sector

amongst firms that stand in a buyer-seller relationship—predominated. It is thought that the US

Congress’spassageoftheClaytonAntitrustActof1914,whichmade itmoredifficulttomergeforthe

purpose of creating amonopoly, was one reasonwhy firms chose to expand outside their industries

(Gaughan 2007). The third US merger wave lasted from 1965 to 1969 and was baptized the

“conglomerate wave” because large firms diversified their holdings by acquiring firms in unrelated

sectors. A fourth merger wave lasted from 1984 to 1989, the twin attributes of which were the

prevalenceofmegamergersandtheroleofhostile takeovers. Intheconglomeratewaveof the1960s,

largefirmsswallowedsmall-andmedium-sizefirmsinunrelatedsectors.Themergerwaveofthe1980s

saw large firms absorb other large firms, such that the number of $100 million dollar US mergers

increased23timesfrom1974to1986.

Afifthmergerwavebeganinthe1990sthatwasinternational inscope.Whereasmostmerger

activityinpriorwaveshadbeenconcentratedintheUS,thefifthwavesawintensivetakeoveractivityin

Britain, Germany, France, Asia, and Central and South America. In addition to being international in

scope, themergerwaveof the1990swas fueled, inpart,byaglobalprivatizationpushdrivenby the

widespreadadoptionofneoliberaldoctrine following theColdWar.Another featureof the fifthwave

wastheemergenceofadevelopingcountry–domiciledacquirer,whosesizewasusuallyaconsequence

oftheprivatizationofstateassets(Gaughan2007).

9

At a minimum, explanations for M&A activity usually try to account for two things: merger

motives (causes)andpostmergeroutcomes(effects).Growthandefficiency (the latteroftendescribed

asoperatingor financial “synergies”)are twoof themostcommonmotivationscited forM&Aactivity

(Gaughan2007).But fromaheterodoxperspective, larger relative firmsizeand theattendantmarket

power that greater size bestows is the real amalgamation prize. One way of measuring M&A is to

contrastitwithinvestmentinfixedassets.A“buy-to-build”indicatorcapturesthebasiccalculusopento

proprietors:purchaseexisting industrialcapacityonthemarket forcorporatecontrolorpaytohave it

builtanew.Fixedasset investment leads to thecreationofnew industrial structuresand ishistorically

associatedwith net new employment andmore rapid GDP growth.Mergers and acquisitions, on the

other hand, merely shuffles ownership claims between proprietors while leaving the industrial base

unchanged.Corporateamalgamationmayleadtoanetreductioninemployment,asthenewlyminted

organizationshedssomeduplicatedfunctionsandleavesproductivecapacityidle.

The evolution of American M&A activity is captured in the buy-to-build indicator plotted in

Figure 1. Three things command our attention. First, the series clearly demonstrates the wave-like

patternofM&Aoverthepastcentury.Thepeaksofmajormergerwavescorrespondwithbusinesscycle

peaksandareobservableonthechart.ThesecondfeaturetonoteistheincreasingimportanceofM&A

relative to investment in fixedassets,especially in recentdecades.Between1895and1990, forevery

dollarspentonbuildingnewindustrialcapacity,Americanbusinessesspentanaverageofjust18cents

onM&A.Inthequartercenturysincetheso-called“freetrade”erabegan,averageM&Aincreasedtoan

averageof68percentoffixedassetinvestment—anearfourfoldincreaseoverthepreviouscentury.The

thirdthingtonotice isthesustainednatureofM&Aactivity inrecentdecades.Ofthesixpeakmerger

wave values since 1895, three have occurred in the past 15 years. Even though 1999 represents the

historicalhigh,theperiodsince1990hasbeenunprecedentedinAmericancorporatehistory.3

Some questions follow: How have the merger waves of recent decades altered American

businessinvestment?WhatbearingdoesamalgamationhaveonthestructureoftheAmericancorporate

sector? And, if amalgamation is a form of corporate restructuring, are there consequences for GDP

growthandthedistributionofincome?

3SimilarresultsaretobefoundforCanadainBrennan(2014;2015).

10

InvestmentandGrowth

Since thepublicationofAdamSmith’sWealthofNations in1776,mainstreameconomistshave tolda

storyofdevelopmentthatputsthecapitalistatthecenterofeconomicprogress.Byconvertingsavings

intoinvestmentandbysubmittingtothedisciplineofintensepriceandproductcompetition,capitalists

helpsettheeconomicwheelsinmotionandensuretheefficientuseofsocioeconomicresources.Partof

thisstoryisinvestmentinindustrialcapacity,whichisakeydeterminantofGDPgrowth.Theconceptual

and statistical relationship between fixed asset investment and GDP iswell established (De Long and

Summers 1992; Dougherty and Jorgenson 1997; Jorgenson 2007). Figure 2maps that relationship by

plottingtherateofgrowthofbusinessinvestmentinnonresidentialstructuresandequipmentandGDP,

bothadjustedforinflationandsmoothedas10-yearmovingaveragestocapturethecyclicallyadjusted

(“secular”)trend.Therelationshipbetweenthesetwovariablesoverthepastcenturyisverystrong—a

correlationof0.70between1925and2013, rising to0.80over thepasthalf century—which suggests

thatbusinessinvestmentinfixedassetsisinfactakeydeterminantofgrowth.4

4AllcorrelationcoefficientsarePearsonianandaresignificantat1percent(two-tailed).

11

In the early decades of the postwar era, the US (and other Organisation for Economic Co-

operation and Development [OECD] countries) experienced heavy fixed asset investment and rapid

growth. In the decades since 1980 there has been a shift to under-investment and comparatively

sluggishgrowth.InsetwithinFigure2arethedecadeaverageratesofGDPgrowthadjustedforinflation

andpopulation.Usingthismetric,AmericanGDPgrowthaveraged2.8percentbetween1940and1980

andwasmorethanhalvedbetween1980and2013,fallingtojust1.3percent.Thus,thelowerlevelsof

growthinrecentdecadesappeartobeledbyunder-investmentinfixedassets.

Figure3crystallizestheevolutionofAmericaninvestmentinindustrialcapacityoverthepostwar

periodbyplotting theproportionofbusiness spendingonnonresidential structuresandequipment in

GDP.The chart clearly shows that in theearlypostwardecades (1946–80) fixedasset investmentwas

comparatively high and trended upward. In the decades since 1980, the American business sector

investedcomparativelylessinfixedassetsandthetrendwassharplydownward.Fixedassetinvestment

peaked at 18 percent of GDP in 1979 and by 2010 it reached a postwar low of just 10 percent. The

difference between the average inflation-adjusted rate of growth of spending on fixed assets is even

12

starker: between1945 and 1980 the growth rate averaged 7.3 percent, and it fell to just 1.7 percent

between1980and2013.

Given that investment fuels growth, how has investment by the largest American-listed firms

evolved inrecentdecades?Figure4decomposes investmentamongthetop100American-listedfirms

overthepostwarerabyplottinginvestment infixedassetsandstockrepurchase(bothasapercentof

revenue), a buy-to-build indicator for the top 100 firms, and a metric capturing “notional” total

investment,thelastmeasuredasfixedasset investmentplusacquisitionsandstockrepurchaseallasa

percentof revenue. Thenotional investment series indicates thedifferentways that firms candeploy

availableassetsforthesakeofgrowth.Whatdothefactstellus?

Inthetwodecadesbetween1950and1970,fixedassetinvestmenttrendedupward,risingfrom

6 percent to 13 percent of revenue, only to trend downward in the decades after 1970, falling to a

postwar low of 5 percent over the past decade. This suggests that large firms may be leading the

stagnation tendencies of recent times through fixed asset under-investment. Stock repurchase was

nearlynonexistent in the1970s,buthasgrown in significance ineachsubsequentdecade, rising from

13

lessthan1percentofrevenueinthe1970sto7percent in2007.5Tobeclear,thismeansthatforthe

first time in American corporate history, large firms spentmoremoney repurchasing their own stock

thanontheexpansionof their industrialbase.Large firmshavealsobeenonabuyingspree in recent

decades,plowingenormousresourcesintoacquisitions.

The notional total investment series clearly shows that if large firms had spent all their

acquisitionandstockrepurchaseresourcesonfixedassetinvestment,thedownwardtrendinfixedasset

investmentwouldhaveactuallybeenaninvestmentboom.Actualfixedassetinvestmentpeakedin1970

at13percentofrevenue,butnotionaltotalinvestmentpeakedin2007at16percentofrevenue,which

wasconsiderablyhigherthanthebusinesssectorasawhole(whichwas14percentofGDPinthatyear).

Given that investment in machinery and equipment is a key driver of GDP growth, the enormous

resourceredirectionwithin large firmsfromfixedasset investmenttowardM&Aandstockrepurchase

5AsJoandHenry(2015)explain,itwasonlywiththeadoptionofRule10b-18in1982thattheSECformallypermittedsharebuybacks(upto25percentofthestock’saveragedailytradingvolume).

14

put tremendous downward pressure on American growth.Why would large firms redirect resources

awayfromindustrialexpansioninfavorofacquisitions?

CorporateAmalgamation,CorporateConcentration,andtheDistributionofIncome

Bycapturing theoverallpositionof large firms in thecorporateuniverse,manyheterodoxeconomists

have utilized aggregate concentration as a broad proxy for corporate power. In the language of

neoclassical economics, “perfect competition” is a condition in which a large number of buyers and

sellers, perfect information, freeentry andexit, andhomogenousproductsprevail.Under thismarket

structure, sellers do not have the ability to influence price. But as firms combine and the market

structuremovesfromthecompetitiveendofthespectrumtotheoligopolisticend,largefirmsgofrom

being price takers to price shapers and price makers. John Blair (1972), for example, argues that as

aggregateconcentration increases,marketbehaviourchanges.“Communitiesof interest” formaround

powerful familiesandfinancialgroupsandthisenablesthemtocoordinatetheiractivitiestoagreater

extent thanwould otherwise be possible. Independent (read: competitive) behavior is lessened, Blair

continues, as dominantproprietors andexecutivesopenlyor tacitly agree that firms should avoid the

disruptionsassociatedwith“pricecompetition”andaim,instead,atahealthy“targetprofitrate.”

MancurOlson (1982) explains some of themechanics (and impediments) that individuals and

institutionsfaceinorganizingforcollectiveactionthroughthebuildingofcoalitions:

Thelargerthenumberofindividualsorfirmsthatwouldbenefitfromacollectivegood,

thesmaller theshareof thegains fromaction in thegroup interest thatwillaccrue to

the individual or firm that undertakes the action.… The incentive for group action

diminishes as group size increases, so that large groups are less able to act in their

commoninterestthansmallones.(Olson1982,p.31)

Thenumericscaleofgroupslike(nonunionized)workers,taxpayers,andconsumersmakesitdifficultto

organizeforcollectiveaction.Theincentives,Olsonexplains,arenotstrongenoughtopullsuchgroups

together.One reason for this is that the services provided by such coalitions are often distributed to

everymemberof the coalitionequally, andamongbroad coalitions, thismakes the “perunit”benefit

small.What’s more, the cost of organizing such coalitions may be large, which acts as an additional

impedimenttocollectiveaction.

15

Theoppositelogicisatplaywithsmallgroupslikelargefirmsoperatinginoligopolisticmarkets.

In the contextof the coordinating activitiesbetween large firms—activities thatwould include setting

marketprices,containingtherateofindustrialexpansion,orlobbying—becausetheyarefewinnumber,

theorganizationalburdenismuchsmaller.What’smore,suchfirmsstandtodisproportionatelybenefit

fromcollectiveaction.Smallergroups,Olsonasserts,possessdisproportionateorganizationalpoweror

“cartelisticpowerpercapita.” It is imperative tonote that theactivitiesof suchcoalitionswill tend to

benefitcoalitionmembers,eventhoughsaidactivitiesmayreducetotalsocietalefficiencyorhamperthe

growthofaggregate income.Thesenegativeeffectsarising fromcoalitionalbehaviorwillbe feltmore

stronglybynon-coalitionmembers(i.e.,bysocietyatlarge).

Figure5contrasts the (national)buy-to-build indicatorwithaggregateassetconcentration, the

lattermeasuredas the totalassetsof the top100American-listed firmsasapercentof thecorporate

assets. The two series are tightly and positively intertwined over six decades, which signals that

amalgamationwavestendtoconcentrateassets.Intandemwiththeconglomeratemergerwave,asset

concentration increased by one-half between 1950 and 1970, rising from 8 percent to 12 percent of

16

corporateassets.Withthesubsidingofmergeractivitybetween1970and1990,assetconcentrationfell

by one-quarter. Then, with the onset of the most sustained period of merger activity in American

corporatehistory,assetconcentrationmorethandoubled,risingfrom9percentin1990to21percentin

2006. There are roughly 5.7million registered corporations in theUnited States (according to theUS

CensusBureau), but the100 largest account for roughlyone-fifthof total assets,which is a veryhigh

degreeofconcentration.

If, as ThorsteinVeblen (1908a, 1908b)posited, capital is a claimonearnings—legal title to an

incomestream—it should followthat theconcentrationof corporateassets shouldbeassociatedwith

theredistributionofincome.Figure6contrastsaggregateassetconcentrationwiththeincomeshareof

thetop100American-listedfirms,thelattermeasuredasthepercentofnetprofitinGDP(withoutlying

valuesremoved in1992and2002).Thetwoseriesaretightlycorrelatedoversixdecades.The income

shareofthetop100firmsisstableovertheearlypostwardecades,havingaveraged1.9percentofGDP

between 1950 and 1990. The elevated merger activity of recent decades and the associated

concentration of assets coincided with a doubling of the income share of the largest firms, which

reachedapostwarhighof3.9percentofGDPin2013.

17

Forinstitutionalpowertobeameaningfulcategoryinpoliticaleconomy,itmustincludecontrol

over—redistributionof—income.The facts inFigures5and6aresignificantbecause theysuggest that

thestructureof thecorporatesector,which is fueledbycorporateamalgamation,concentratesassets

andcentralizesincome.Insofarasthedistributionofincomereflectstheorganizationalstructureofthe

political economy, power becomes a meaningful heuristic. Is there a relationship between the

centralizationofnationalincomeinthehandsofthelargestfirmsandincomeinequality?

Yes,buttherelationshipisindirect.Figure7contraststheprofitshareofthetop100American-

listed firmswith the net dividends paid by theAmerican corporate universe (as a percent of national

income).Thecorrelationoversixdecadesis0.76,orhigh.Despitethecyclicality,thelevelofeachseries

remainedrelativelystable in theearlydecadesof thepostwarera.Bothmetricsbegantoclimb in the

1980s and soared in the past two decades. Of the 5.7 million registered American corporations, 98

percenthavelessthan100employeesand90percenthavefewerthan20employees.Thisimpliesthat

theoverwhelmingmajorityofUSfirmsaresmallandmediumsize,andareunlikelytopaydividends.Itis

predominantlylargefirmswithcomplexownershipstructuresthatpaydividends.

18

Basedoncalculations fromtheauthor’sdataarchives, in the late1980s fullyhalfofAmerican

corporate dividendswere paid by the largest 100 firms. Inmore recent years, roughly one-quarter of

total dividends were accounted for by the largest 100 firms. This likely why the dividend share of

national incomemoves in tandemwith theprofit shareof largestAmerican-based firms,because it is

primarily large firms that pay dividends. The relationship in Figure 7 suggests that as the corporate

universe concentrates and large firms claim a greater share of national income through enhanced

market power, the potential to distribute that profit to owners of (large) corporations, via dividends,

increases.Logicallyandempirically,then,alargerprofitshareofnationalincome,whichisdrivenbythe

amalgamation-fueledconcentrationofcorporateassets,leadstoanenlargedcapitalistincomeshare.

Capitalist income is often defined as the sum of interest and profit payments. Inmainstream

economic thinking, the abstinence endured by the owners of corporate debt is rewarded with fixed

interestpaymentswhiletherisktakenbytheownersofcorporateequityisrewardedwithprofit.Froma

personal income standpoint, however, capitalist income is dividends. Defined in classical terms, a

capitalistisanemployerofwagedlabor.Theprofitearnedbyafirmhasfourpossibleuses:thefinancing

of expansion through some form of investment, retained earnings (often in the form of cash), stock

repurchase, anddividendpayments. In the final analysis, even thoughprofit and interest are the sum

total of “capitalist income,” from a personal income perspective capitalist income is solely dividends.

Giventhis,istherearelationshipbetweencapitalistincomeandAmericanincomeinequality?

Figure8plotsthelong-termrelationshipbetweentheincomeshareoftherichest1percentof

Americans—aproxyforoverallincomeinequality—andthedividendshareofnationalincome.Thetwo

series are nearmirror images on each other, registering a correlation of 0.89 over the past century.

Dividendpaymentsplummetedduring theGreatDepressionandagainduring the SecondWorldWar,

havingreachedahistoriclowof2.3percentofnationalincomein1945.Dividendshoveredaroundthat

leveltillthelate1980s,afterwhichtheybegantosoar,pullingtherichestAmericans’incomeshareup

withthem.Notably,thedividendshareofnationalincomereachedahistorichighof6.6percentin1930

andin2007,whichisroughlythesamemomentsintimethatAmericanincomeinequalitypeaked.

The implication is that American income inequality is driven, in part, by dividend payments.

However, because dividend payments are at least partially restricted by corporate profitability, the

relationship between corporate power and income inequality remains unclear. Elsewhere (Brennan

2014)IhavemappedtheAmericancorporateprofitshareofnationalincomefromthe1920sonward.I

foundthat itcollapsedduring theGreatDepression (1929–32)beforesoaringto13.6percent in1942.

19

Forthenexthalfcenturythecorporateprofitsharetrendeddownward,havingreachedapostwarlowof

8.3 percent in 1990. Over the past quarter century the profit share trended sharply upward, having

reachedanall-timehighof14.5percentin2013.Whatdoesallofthismean?

Source:Fix(2015)

Inbroadstrokes,theprofitshareofnationalincomesetstheboundarieswithinwhichcapitalist

income(understoodasdividendpayments) isdetermined,anditappearsthatcapitalist incomeshapes

the overall level of American income inequality. The combined effect of the Great Depression and

SecondWorldWar was a halving of American income inequality (as registered in the top percentile

incomeshare).TheGreatDepressionwitnessedthedualcollapseofcorporateprofitabilityandthetop

incomeshare.However,profitsrecoveredfasterthanwages inthe1930s,whichmeantthattheprofit

shareofnationalincomeandthetopincomesharereboundedinthedecadeafter1933.Thetopincome

share collapsed again during the Second World War, but not because of a collapse in profitability;

instead,thedividendshareofprofitwasreducedbythree-quarters—thusreducingcapitalistincome—in

favoroftheheavyfixedassetinvestmentassociatedwithwartimemobilization.

20

Intheearlypostwardecades,incomeinequalitycontinuedtodecline.Thiswasnotbecauseofa

shift in dividend payments, however, which held steady at one-fifth of corporate profit. Instead, the

profitshareofnationalincomeitselfwasnearlyhalvedinthefourdecadesbetween1940and1980.So

incomeinequalityintheearlypostwardecadeswasastoryofdecliningrelativecorporateprofitability,at

least in part. Income inequality began to increase in the early 1980s, even though theprofit shareof

nationalincomedidnotsignificantlyrise.Theexplanationseemstobeadoublingofthedividendshare

ofprofit,whichrosefromone-fifthinthelate1970stotwo-fifthsby1990.After1990,theprofitshareof

nationalincomeandthedividendshareofprofitbothrosesharply.AtthetimeoftheGreatRecessionin

2008–09, the profit share of national income had reached a postwar high of 13.7 percent and the

dividendshareofprofithadreachedapostwarhighofthree-fifths.

So it is the combined effect of surging corporate profit and a higher dividend share of profit

distributedtocapitaliststhathasbeendrivingAmericanincomeinequalityinrecentdecades.Thehigher

dividend share of corporate profit implies that there are fewer resources available for fixed asset

investment.SotheprocessesthatrestrictinvestmentandGDPgrowtharealsotheprocessesthatinflate

thetopincomeshareandexacerbatepersonalincomeinequality.

Eventhoughdividendpaymentsappeartotheprimecandidate inexplainingAmerican income

inequality, it has been long understood that executive compensation practices also play a role in

Americanincomeinequality.Whilemanyexecutiveshavecompensationpackagesthatincludesharesin

thefirmstheymanage,weoftendistinguishthecapitalistclass(i.e.,theownersofcorporateequitywho

employworkers)fromthemanagerialclass(i.e.,thosewhoareemployedbythefirmtorunday-to-day

operations).HowdoesexecutivecompensationfitintotheAmericanincomeinequalitypicture?

Central to Berle and Means’s 1932 “separation thesis” was the positing of a three-pronged

process:anincreasingconcentrationofcorporateassets,coupledwithanincreasingdispersionofstock

ownership, resulting in a separationofownership fromcontrol. Putting aside the validityof the claim

thatcontrolhadactuallydelinkedfromownership,theideaexertedconsiderableinfluenceoneconomic

theorists and policymakers. If the large corporation was no longer under proprietary control (having

fallen under managerial control), the incentive structure would no longer compel those exercising

corporateauthoritytosteerthefirminaprofit-maximizingdirection,thusthreateningtheequilibrium-

seeking nature of laissez-faire capitalism (or so mainstream economists reasoned). Managers might

instead steer the firm in a direction that enriched themselves while sacrificing the interests of

stockholders,whoweretoonumerousanddispersedtochallengemanagerialauthority.

21

Theriseofstockoptionsinthe1980sandtheirexplosioninthe1990smaybethoughtofasone

institutional response to thealleged separationofownership fromcontrol (Frydmanand Jenter2010;

Murphy 2012). By compensating managers with stock, their interests and attendant behavior would

presumably realign with those of stockholders, thus transcending the separation thesis and ensuring

firms behave in a profit-maximizing manner (again, according to mainstream assumptions). Murphy

(2012)tellsusthatinthe10yearsafter1992,medianCEOcompensationforfirmslistedontheS&P500

morethantripled,fueledinlargepartbystockcompensation.Eventhoughthisisjustoneaspectofthe

so-called “shareholder revolution,” it has clearly led to important changes in the governance of large

firms.

Whatistherelationshipbetweenstockcompensation,whichhasbeendrivingtrendsinCEOpay

inrecentdecades,andpersonalincomeinequality?Figure9plotsthevalueofstockrepurchase(relative

to revenue) among the top 100 American-listed firms with the income share of the top percentile

incomegroup.Thetwoseriesaretightlysynchronizedandhaveacorrelationcoefficientof0.93between

1971and2013.Thissuggeststhat,inconjunctionwithheighteneddividendpayments,theredirectionof

corporate income away from fixed asset investment toward stock repurchase has not only slowed

22

growthbutalsoexacerbatedinequality.ThereisclearlymoretoAmericanincomeinequalitythanstock

options,but insofaras top incomeearnersdrive inequality trends,and insofarascorporateexecutives

make up a substantial proportion of the top income group, the evolution of executive compensation

playsakeyroleindeterminingtheoveralllevelofAmericaninequality.

Summation:CorporateConcentrationIsaMissingElementintheInequalityPuzzle

Recent decades have seen themost sustainedM&A activity in American corporate history. For every

dollar spent on expanding America’s industrial base, 68 cents was spent redistributing corporate

ownershipclaimsbetweenproprietors.Asaparallelphenomenon,theaveragerateofGDPgrowthinthe

threeplusdecadesafter1980washalvedincomparisonwiththefourpriordecades.Andbecausefixed

asset investment is an expansionary activity that is associated with job creation, the restructuring of

AmericaninvestmentinfavorofM&AhaslikelybeenasignificantfactorinthedeceleratingrateofGDP

growth.

Whenwedisaggregatethecorporatesector,focusingonthelargest100firms,wefindthatthe

level of fixed asset investment increased in each successive decade between 1950 and 1980 and

decreasedineachdecadesince1980.Asaproportionofrevenue,fixedasset investmentintheperiod

since2000was less thanhalfofwhat itwas in the1960sand1970s.Andwhile the largestAmerican-

listed firms have spent comparatively less on the expansion of industrial capacity, they have plowed

enormousresources intotheacquisitionofotherfirms(viaM&A)andon inflatingtheirsharepricevia

stock repurchase. For the first time in American corporate history, it appears that large firms are

spending more resources purchasing their own stock than on the expansion of their industrial base.

Significantly, when we combine fixed asset investment with acquisitions and stock repurchase (thus

arrivingat “notional’ total investment),wediscover that insteadof fixedassetunderinvestment there

hasbeenaninvestmentboom,albeitonecloakedinM&Aandsharebuybacks.

ThemotivationtodivertresourcesawayfromfixedassetinvestmentinfavorofM&Aappearsto

be the concentrationof corporateassetsand resulting increase inmarketpower. Elsewhere (Brennan

2012)Ihavedocumentedastrongandpersistentrelationshipbetweencorporateconcentrationandthe

earningsmargins,profit,andcashflowoflargeCanadian-basedfirms.Bymerging,largefirmsnotonlyto

absorbtheirrival’s income,theyreducecompetitivepressure,whichelevatesearningsmargins.Sothe

causal pathway runs from amalgamation through concentration toward enhancedmarket power and

profitability.

23

This set of relationships is present in the United States. Over the long haul, mergers and

acquisitions centralize corporate ownership and concentrate corporate assets. Increased corporate

concentration is associated with an enlarged income share for large firms. These firms have plowed

historicallyunprecedentedresourcesintostockrepurchase.Andbecausecorporateexecutivesareoften

compensatedwithstock,thishasbeenonefactor,viasharepriceinflation,ofheightedAmericanincome

inequality.AsecondfactorbehindAmericanincomeinequalityistheincreasingproportionofcorporate

profitdistributed to shareholders in the formofdividends.Thisappears tobe the keydeterminantof

Americanincomeinequalityoverthelongterm.

ThefactsdemonstratethatcorporateAmericadoesnotsufferfroma“shortageofinvestment”

in the general sense; rather, resource redirection within large firms, with comparatively less going

toward growth-enhancing industrial expansion and comparatively more going toward asset-

concentrating amalgamation and share price-inflating stock repurchase, helps explain the stagnant

growththatplaguestheUnitedStates.TherehasbeenaninvestmentboomintheUnitedStates,albeit

an invisible one, because it has been hidden in amalgamation and stock option-related activities. The

formerredistributescorporateownershipclaimsbetweenproprietorsandconcentratesassets,whilethe

latterinflatesshareprices.

Ultimately, themerger boom since 1990 has concentrated corporate power and redistributed

national income toward the largest American corporations. From the standpoint of the average

American worker, the casualties arising from this massive resource redirection are shrinking job

opportunitiesandsoaringincomeequality.

24

PartII:LaborUnions,Inflation,andtheMakingofanInclusiveProsperity

The produce of the earth … is divided among three classes of the community.… To

determine the lawswhich regulate this distribution is theprincipal problem in Political

Economy.

—DavidRicardo(1817)

The preservation of past experience in cultural memory can be lost, often with serious social

consequences.Without an adequate understanding of the past, we are bound tomisunderstand the

present.Andmisunderstandingthepresenthampersourabilitytorealizefuturegoals.Butcanweforget

somethingwedidnotunderstandinthefirstplace?Americanlaborunionshavelongbeenattackedby

factions of the business-government alliance. The success that this alliance has had in undermining

unions is fueled, inpart,bycollectiveculturalamnesia. It iseasiertobeapathetic,evencynical,about

assaultsonasocialinstitutionifonedoesnotunderstandwhythatinstitutionemergedorwhatrolethat

institution plays in enhancing the quality of human life. Many Americans do not have the luxury of

forgetting about the socially beneficial aspects of unions for the simple reason that they never

understoodthatroletobeginwith.

Astheepigraph indicates,DavidRicardo—thegreatclassicalpoliticaleconomist—believedthat

theprimarytaskofeconomicscienceistolaybaretheunderlyingpatternsandregularitiesthatgovern

thedistributionofincomeandwealth.Giventhecentralityofincomeinconditioninghumanpossibilities

on both an individual and a social scale, it’s no wonder he thought it imperative to come to a

satisfactoryaccountofdistribution. It is typically left toeconomics to sortouthow thedistributionof

income works. Indeed, orthodox (“neoclassical”) economics would confidently assert that is has firm

knowledge about how incomes are formed and, by implication, how they are distributed. However,

therearegoodreasonsfordoubtingthevalidityofneoclassicaldogmas.

ThestartingpointofthisinvestigationisthefactthatintheKeynesianwelfareregime(roughly

1935–80)GDPgrewatarapidclipandincomeinequalitywasmorethanhalved.Withtheemergenceof

neoliberalglobalization(1980–present),growthsharplydeceleratedandinequalitysoared.Theanalytical

andpolicychallengesassociatedwiththisdouble-sidedphenomenonhavebeenexploredbyresearchers

usingtools fromthestandardeconomictoolbox.Partofthe inadequacyofexistingexplanations is the

absenceof institutionalpoweranddistributiveconflictasexplanatoryvariables.Thefollowingseeksto

fill this gap using tools from early American institutionalism and Post Keynesianism.While there are

25

manymovingparts to thestoryofgrowthand inequality, theamassmentandexerciseof institutional

power in conjunctionwithdistributive conflict between competing incomegroups are two important,

thoughunderexplored,parts.

Specifically, I explore the points of contact between American unions, inflation, and income

inequality.Insteadofprobingtheconventionalcausesofinflation,adistinctlyheterodoxsetofquestions

is explored.Does inflation tend to appear amidst distributive conflict? Is inflation associatedwith the

redistribution of income between different income groups? Does the amassment and exercise of

institutional power have a bearing on changes in the price level? Putting the questions together: can

inflationbeviewedasapowerprocessthatisnourishedonsocialconflictandsystematicallyboundup

withtheredistributionofincomebetweenincomegroups?Thefollowingbuildsontheconflicttheoryof

inflationbydrawing togetherandassessinga complex rangeofempiricaldata to show thatAmerican

inflationcanbeviewedasapowerprocess.

Supplyingasatisfactory,albeittentative,answertotheforegoingquestionsentailsrespondingto

the following conditional statements: if it can be shown that inflation is closely associated with

outbreaks of distributive conflict,measured as strikes and lockouts, such that higher levels of conflict

entail more rapid inflation and lower levels of conflict entail disinflation or deflation; and if the

distributive conflict is associated with the redistribution of income between different categories of

owners, namely the owners of labor power and the owners of corporations, such that accelerating

inflationtendstoappearwiththeredistributionofincomefromcapitaltolabor(andviceversa);and if

distributive conflict is positively associated with the redistribution of personal income, such that

intensified conflict is progressively redistributive and diminished conflict is regressively redistributive;

thenwecanbereasonablyconfidentinthenestedassertionthat(1)inflationmayvalidlybethoughtof

asapowerprocess(2) insofaras its level isshapedbydistributiveconflictbetweencompeting income

groupsand(3)thewinnersofsaidconflicthaveincomeredistributedintheirfavor.

Theremainderofthisanalysisispresentedinsixsections.Thefirstsectionreviewsthe“conflict

inflation”approach,payingparticularattentiontothewayheterodoxscholarsimaginedpriceformation

and inflation in power-laden terms. It also discusses the interplay between institutional structure and

marketpower,arguingthat largefirmsandlaborunionspossessmeasurabledegreesofmarketpower

andthatthecommoditypricestheyshapereflecttherelativepowerofeachgroup.Thesecondsection

explores thepointsof contactbetween the“countervailingpower”of tradeunionsand labor income.

Over the longhaul,average laborcompensationandtheshareofnational incomegoingtoworkers in

the formofwagesandsalariesare shapedbyuniondensity (institutional structure)andstrikeactivity

26

(distributive conflict). The third section explores the commodified power of large firms and finds that

corporateamalgamationnotonlyfuelstheconcentrationofassets,afindingdisclosedinpartI,butalso

increases themarketpowerof large firms.Andelevatedmarketpower among large firms shapes the

distributionofincomebetweenworkersandcorporations.

The fourth section contrasts inflation with the income gains made by labor and capital. The

historical facts suggest that thehourlyearningsof laborandof large firmsarebothcloselyassociated

withinflation.Becausethenominalincomegainsoflaborandcapitalaresynchronizedwithinflation,the

fifthsectionshiftstheanalysistodifferentialterms(seeNitzanandBichler2009)inordertoassessthe

distributiveaspectsofAmerican inflation.Over thepast century, inflationhaspartiallymanifested the

conflict-fueledredistributionofincomefromcapitaltolaborandfromtheuppertothelowerstrataof

the personal income hierarchy. The sixth section summarizes the findings and makes some

recommendationsforhowneoliberalcapitalismcanbereformedinamannerthatbolstersgrowthand

reducesinequality.

Overall,Iarguethatlaborunionsareprogressivelyredistributive—theyredistributeincomefrom

capitaltolaborandfromtheuppertothelowerincomebrackets—atthecostofsignificantinflationary

pressure.TheimplicationisthatAmericaninflationmustbeunderstoodintermsofinstitutionalpower

anddistributiveconflict,bothbecausethesefactorsassistintheproductionofinflationarypressureand

becauseinflationhasbeensystematicallyassociatedwiththeredistributionofincome.

DistributiveConflictandInflation

Outsidemainstreameconomics,aschoolofheterodoxeconomistsemergedthatviewedinflationasthe

productof theexcessiveclaimsmadebydifferent incomegroupsovernational income—theso-called

“conflictinflation”approach(seeRosenbergandWeisskopf1981).Thewagebargainsecuredbyworkers

andthepricingpolicyofbusinesshasthepotential,RobertRowthorn(1977)argued,ofexceedingwhat

is available for each group from national income. The excess of income claims over available income

produces inflation, which Rowthorn asserts will always transfer “real income from workers to

capitalists,” implying that any inflationary redistributionwill always at the expenseofworkers. In this

perspective class conflict over national income fuels inflationary spirals. Richard Burdekin and Paul

Burkett (1996) tellus that the“winners” from inflationwillbe the“claimantsenjoying relativelygreat

economic and political power” (p. 24). Jonathan Nitzan and Shimshon Bichler (2009) make a similar

argumentwhen they tellus that inflation is the“surfaceconsequence”ofa“redistributional struggle”

foughtbetweendifferentgroups.Oneimplicationofthisclaimisthatthosewhoraisetheirpricefaster

27

thanothers simultaneously redistribute income in their favor, thuscreatingdistributive“winners”and

“losers.”

Ratherthanattributingcausalitytotheearningsmarginsoflargefirms,someputtheemphasis

onworkerwagedemands. SidneyWeintraub’s (1978, 1978–79) “wage costmarkup” theoryproclaims

that inflationisaconsequenceofwagesrisingfasterthanproductivity,andbecausetheflowofwages

and salaries are what determine societal purchasing power, there is no effective difference between

“demandpull”and“wagepush”(1978–79).WallacePeterson(1982)wouldhaveusbelievethatinsofar

as wages rise above productivity, power is what explains the gap. For Peterson (1980), “power” is

“control over income,” and is derived fromeither organizations like labor unions and corporations or

throughthepressureputongovernmentalbodiestoshapepoliciestoone’sadvantage.Finally,thework

ofPaulDalziel(1999–2000),MarcLavoie(1992;2014),andotherPostKeynesianscholarspositthatthe

wage bill determines firm claims on the bank system and ultimately influences the supply of money

(technically referred to as the “horizontality” and “endogeneity” of money). The implication is that

inflationisproducedatthelevelofwagebargaining.Dalzielconcludesthatinflationaryepisodescanbe

tracedtoproblemsofsocialprotestandcompetition,particularlythedistributiveconflictoverincome.

The writers surveyed here often speak of “market power” or “organizational power” as an

importantaspectoftheinflationaryprocess.Thisfliesinthefaceoforthodoxscholarship,whichfocuses

solelyontheneutralityofmoneyinadvancingamonetaryviewofinflation.Itisthislackofattentionto

thepowerdynamicsassociatedwithinflationthatformsacriticallimitationintheorthodoxscholarship.

While the conflict approach to inflationadvances the intellectual needle, this lineof thinking requires

concretesubstantiation.Whatisneeded,then,isanexaminationoftheinstitutionalpowerofdifferent

groupsasapointofentry into the interplaybetweendistributive conflictandvariation in the levelof

inflation.

TradeUnionPowerandLaborCompensation

Havingacceptedtheviewthat largefirmsno longeracceptpricesthatareset inperfectlycompetitive

markets, and so could no longer be thought of as being socially optimizing, John Kenneth Galbraith

(1952), the lateHarvardeconomist,wenton toposit that alternative institutional arrangementswere

neededtomakemoderncapitalismmorefunctionalandfair.Becausebusinessescombinewithaviewto

administeringprofit-friendlyprices,wageearnersought tocombine inunionswithaviewtoelevating

laborcompensation.Galbraithutilizedtheterm“countervailingpower”todenoteaninstitutionalsetting

inwhich thepowerof large firms isoffsetby thepowerof laborunionsand thewelfare state. In the

28

generalevolutionofpolicy,politicsandculture,laborunionsactasa“check”onthecommodifiedpower

oflargefirms,andthiscountervailenceisfeltinwaysasdiverseassocialpolicy,politics,andculture,not

justwages.

Thisviewstandsinoppositiontoneoclassicaleconomics,whichviewsthemarketpriceoflabor

powerlikeothercommodities—intheshortrunitisdeterminedbysupplyanddemand,andinthelong

run the “absolute”wage rateand thenationalwagebill reflectproportionalproductive contributions.

Forneoclassicists,organizedlabormaybeabletoelevatelaborcompensationto“artificially”highlevels,

butitdoessoattheexpenseofnonunionizedlaborand/oremployment.Inotherwords,unionscanonly

redistributeincomewithinagivennationalwagebill—theyareunabletoredistributenationalincomeas

such(SamuelsonandNordhaus2010). Iftheneoclassicalviewistrue,thereneednotbearelationship

betweentradeunionpowerandlaborcompensation.

Unionsrepresentworkersatthebargainingtablewithemployersand,becausetheyareableto

negotiateasacollectiveunit,theirbargainingpositionisenhancedrelativetowhatitwouldbeifeach

individual bargained in isolation. An enhanced bargaining position (often) enables unions to increase

their compensation and benefits to a greater extent than would otherwise occur. Furthermore, by

increasingtheremunerationoforganizedworkers,laborunionsservetoraisesocialexpectationsaround

thelevelofcompensationforworkmorebroadly.Thishasspillovereffectsinnonunionizedworkplaces.

What arewe tomakeof themainstreamargument thatunionized labor canonly redistribute income

within a given national wage bill, but not raise it? Oneway of assessing this claim is to contrast the

institutionalpowerof laborunionswith theshareofnational incomegoing toworkers. If the twoare

positivelycorrelatedoverthe longterm,thiswouldsuggestthattradeunionpowerdoes infactshape

thedistributionofincome.

Figure 10 plots union density, measured as the proportion of union membership in total

employment,andanadjustednationalwagebillfrom1900to2013—thelattermeasuredastotalwages

andsalariesdividedbyGDPlessthewagesandsalariessharepaidtothetoppercentileincomegroup.By

adjustingthenationalwagebill inthismannerweapproximatetheclass-baseddistributionof income.

And becausemost people in the top-percentile income group are not in a union, wewill be able to

determine if there is a relationshipbetween the levelof tradeunionpowerand the shareofnational

incomegoingtowhatusedtobecalledthe“workingclass.”Thecorrelationbetweenthetwovariablesis

0.86overthepastcentury,orveryhigh(andstatisticallysignificant).

29

In1900,unionizationwasbelow3percent,thoughittrendedupwarduntil1920.Thecombined

effect of demobilization after the First WorldWar and a deep recession in the early 1920s led to a

decline in union density. It was not until 1933, after four years of the Great Depression, that the

American unionization began to surge. Facilitated by the National Labor Relations Act of 1935 (the

WagnerAct)andotherNewDeal legislation,uniondensitysoaredfrom7percent in1933toahistoric

high of 29 percent in 1954. Antiunion legislation such as the Taft–Hartley Act6 in 1947 and the

reactionary political forces associated with McCarthyism in the early 1950s contributed to declining

unionization in the subsequent period. In the quarter century between 1954 and 1979, unionization

modestlydeclined,fallingfrom29percentto24percent.ThedeclineofAmericantradeunionsspedup

with the Reagan administration’s antiunion efforts as well as other economic and technological

developments,suchasoffshoringandautomation.By2013overallAmericanuniondensitystoodatjust

11percent.

6LaborManagementRelationsActof1947.

30

The adjusted national wage bill followed a similar pattern. The wage bill for the bottom 99

percentwas43percent in1929and rose inagradual fashion toahistorichighof54percent in1970

before falling to 42 percent in 2013—a historic low. It isn’t a coincidence that the national wage bill

declined from the 1970s onward. The main driver of a shared prosperity—unionization—markedly

declinedoverthatperiod.

In order to meaningfully analyze institutional power we must be able to distinguish the

amassment of power from the exercise of power. The facts in Figure 10 speak to the organizational

capacityofunionizedworkers (amassedpower),but theyare silenton theextentofworkplaceaction

thatAmerican laboractuallyundertook (theexerciseofpower).Theabilityofworkers to intentionally

act in concert represents a type of power. In addition to the level of density, the collective ability of

workerstorefusetoworkwithoutasatisfactorycontract—aworkstoppageorstrike—imposesapenalty

on employers who fail to meet demands around compensation, benefits and working conditions.

Arguably,thisisthemaininstitutional“weapon”thatlaborpossesses.Andastrikeisamongtheclearest

manifestations of the distributive conflict between different income groups, namely profit-seeking

proprietorsandwage-earningworkers.

Workers strike for reasons other thanwages, of course, but does the extent of strike activity

helpexplainchangesinAmericanlaborcompensation?Figure11contraststherateofchangeofaverage

nominalhourlyearningswiththeextentofstrikeactivity,bothsmoothedasthree-yearmovingaverages

toeasethevisualassessment.Thelatterisasuperindexcomposedoftheaverageoffoursub-indices:

the number of work stoppages, the number of workers involved, the number of days idle, and the

percent of work time lost.7 In the resulting super index, a value of 0 means strike activity is at its

historicalaverage,withpositiveandnegativevaluessignalingaboveaverageandbelowaveragestrike

activity, respectively. The extent of strike activity is tightly and positively correlated with the rate of

growthofhourlyearningsandthestrengthofthestatisticalrelationshipsteadilyincreasesovertime.

Over the past 130 years there have been three major strike waves. The first wave occurred

during the FirstWorldWar and peaked in 1919. Over the next 15 years the extent of strike activity

declined(intandemwithdeunionization),butasecondstrikewavebeganintheearly1930sthatpeaked

in1947.Strikeactivitytrendeddownwardtill1963whenathirdwavebegan,whichpeakedintheearly

1970s.Americanlabordisobediencedeclinedprecipitouslyinthelate1970sand,asof2013,stoodatan

7AmoredetailedexplanationofthisstrikeindexistobefoundintheAppendix.

31

all-timelow.Thepatternofwagegrowthalsowentthroughthreewavesofsimilarduration.Notethat

majorstrikewavesandepisodesofrapidwagegrowthcoincidewithsocialcrises—twoworldwars,the

GreatDepression,adomesticculturalrevolutionintheUnitedStatesduringthe1960sand1970s,andan

energycrisis–linkedwar in theMiddleEast.Thetimingof theseepisodeswillbecomemoresignificant

onceweexaminetherelationshipbetweenlaborcompensationandinflation.

It is not entirely clear why domestic labor developments are linked with outbreaks of

internationallyorganizedviolenceand/orcrisis.Inthecaseofthefirsttwostrikewaves,thebargaining

positionoflabormayhavebeenstrengthenedintheshiftfrommoderateorheavystagnationto(near)

fullemployment.Theexperienceof the1960sand1970swasdifferent,of course.Unemploymentdid

notdroptohistoriclows;itroseto(whatatthetimewere)postwarhighs.Itfollowsthatif(successful)

workerwagestrugglesleadtohigherlaborcompensation,thiswillbetransmittedtothenationalwage

bill,whichstrengthenstheclaimthattheinterplaybetweeninstitutionalpoweranddistributiveconflict

leadstotheredistributionofincome.

32

Figure 10 documents the relationship between the institutional power of organized American

labor and the national wage bill. Figure 11 shows that changes in average labor compensation are

shapedbydistributiveconflict,namelyworkerrevoltsagainstproprietors.Iftherelationshipspresented

in these two figuresare tobebelieved, thenstrikeactivity shouldhavea redistributiveproperty to it.

Does American strike activity redistribute income between capitalists and workers? The validity of

Galbraith’scountervailingpowerargumentisreaffirmedinFigure12,whichcontrastsstrikeactivitywith

ametric thatcaptures thedistributivestrugglebetween laborandcapitaloverwagesandprofits.The

latterismeasuredasaratioofaveragehourlyearningstotheS&P500priceindex.Whenthisindexrises,

workersredistributeincomefromcapitalists;whenitfalls,capitalistsredistributeincomefromworkers.8

Both series are smoothedas three-yearmoving averages andare tightly correlatedover thepast 130

years,withthestrengthofthestatisticalrelationshipincreasingovertime.

8ThismethodofcapturingthedistributivestrugglebetweenworkersandcapitalistswasfirstbroughttomyattentioninJonathanNitzan’spoliticaleconomygraduatecourseatYorkUniversity.

33

Therehavebeen threemajor strikewavesover thisperiod, twoofwhichunfoldedduring the

1930s–40s and the 1960s–70s. These outbreaks of labor disobediencewere hugely and progressively

redistributiveandappeartohaveplayedacrucialrole inthecreationofan inclusiveprosperity. Inthe

decades since1980, labormilitancyhas fallen toahistoric lowand,perhapsunsurprisingly, the labor-

capital redistribution index has reached a 130-year low. The pacification of the labor force, driven

partiallybyantiunionlaborlaws,hascontributedtowhatmightbethemostregressiveredistributionof

incomeinAmericanhistory.Thesharpdeclineinthelabor-capitalredistributionindexsincethe1980sis

consistentwithotherfindings,namelythatcapitalhastendedtoclaimalargershareofnationalincome

attheexpenseof labor.SonotonlyareAmericanworkersstrikinglessthaneverbefore,butthegains

fromgrowtharealsomoreheavilytiltedinfavorofcapitalthantheyeverhavebeen.

The orthodox economic argument that labor unions cannot enlarge the national wage bill or

increase laborcompensation isnotsupportedbythehistorical facts.Unionswere(andare) integral to

raisingtheaveragestandardoflivingintheUnitedStatesandindetermininghowthegainsfromgrowth

areshared(withinthefirm)betweenthetwomainincomeclasses.Ifuniondensityandstrikeactivityare

power processes that progressively redistribute income, why has American income been regressively

redistributed in recentdecades?Thedeclineofunionsand thedescentof strikeactivity appear tobe

causalvariables,buthowcanwemakesenseoftheamassmentandexerciseofpowerwithrespectto

business?

CorporatePowerandCapitalistIncome

Part I of this analysis presented new estimates of American mergers and acquisitions (M&A) and

aggregate asset concentration. The latter is understood by heterodox economists as a proxy for the

overall power of large firms. The deep historical facts suggest that corporate amalgamation fuels the

concentration of corporate assets. The analysis went on to show that as the corporate sector

concentrateslargefirmsclaimagreatershareofnationalincome.Thisisconsistentwiththeheterodox

viewthatmarketpowercomeswith firmsize. In the languageofclassicalandneoclassicaleconomics,

“perfectcompetition”isaconditioninwhichalargenumberofbuyersandsellers,perfectinformation,

freeentryandexit,andhomogenousproductsprevail.Underthismarketstructure,sellersdonothave

theabilitytoinfluenceprice.Butasfirmscombineandthemarketstructuremovesfromthecompetitive

endofthespectrumtotheoligopolisticandmonopolisticend,largefirmsgofrombeingpricetakersto

priceshapersandpricemakers.

34

GardinerMeans(1935)andMichałKalecki(1971[1938];1971[1943])arguedthatlargefirmsin

concentrated market structures behave differently than small- and medium-size firms operating in

competitivemarkets.Kaleckidevisedhisconceptofthe“degreeofmonopoly”tocapturethisdifference

in price behavior. Large firms have greater pricing discretion than smaller firms and their monopoly

powercanbeapproximated,Kaleckiargued,inthespreadbetweentheircostsandtheirsellingprice.For

thepurposesofthisinvestigation,theearningsmarginsoflargefirmscanbeusedasanapproximation

for market power. By merging, large firms not only absorb their rival’s income but also reduce

competitivepressure,whichelevatesearningsmargins.Sothecausalpathwayrunsfromamalgamation

throughconcentrationtowardmarketpower.

Thoughintuitivelycorrect,thislineoftheoreticalreasoningmustbetestedtoassessitsempirical

validity. If there is a relationship between M&A and corporate concentration, does this imply that

mergeractivityleadstoelevatedmarketpoweramongstlargefirms?Figure13replotsthebuy-to-build

indicator (measured as the percent of M&A in fixed asset investment) alongside the markup of the

largest100American-listedfirms,thelatteraproxyforKalecki’sdegreeofmonopoly(measuredasthe

percentofnetprofit in revenue,withoutlying values removed in1992and2002). The two series are

positivelycorrelatedoversixdecadesandthestrengthofthestatisticalrelationshipincreasesovertime.

Inthedecadeswhenmergeractivity increased,themarketpowerof largefirmstendedto increase. In

the decades when merger activity subsided, market power also declined. So the claim that

amalgamationandconcentrationleadstoelevatedmarketpowerhasconsiderableempiricalsupportin

theUnitedStates.9

Large corporations in “semi-monopolistic” settings not only tend to have greater pricing

discretion, as Means (1935) showed, but they tend to have deeper earnings margins. Kalecki (1971

[1943]) also argued that the degree of monopoly is of “decisive importance for the distribution of

incomebetweenworkersandcapitalists”(p.51).Thepowertoinflateearningsmargins,Kaleckiargued,

shapesthedistributionofincome:

The long-run changes in the relative share of wages … [are] determined by long-run

trendsinthedegreeofmonopoly.…Thedegreeofmonopolyhasageneraltendencyto

increase in the long run and thus to depress the relative share ofwages in income…

[although]thistendencyismuchstrongerinsomeperiodsthaninothers.(Kalecki1971

[1938],p.65)

9TheserelationshipsarealsopresentinCanadaoverthepostwarperiod.SeeBrennan(2012;2014;2015).

35

Forthepurposesofthepresentdiscussion,Kalecki’sbasicassertioncanberestatedasaquestion: is it

truethatthedegreeofmonopolyamonglargefirms(measuredusingthemarkup)hasabearingonthe

relative share of wages, and therefore shapes the distribution of income between workers and

capitalists?

Broadlyspeaking,incomewithinthefirmissharedbetweenworkers(intheformofwagesand

salaries) and capitalists (in the formof profit). At a national level, the totalwagebill andprofit share

capturetheseproportions.Whenwedivideaggregatecorporateprofitbythenationalwagebillforthe

bottom99percentoftheworkforce,excludingtherichest1percentincomegrouptoapproximatethe

class-based distribution of income, we arrive at a metric that approximates the distributive struggle

betweencorporationsandworkersoverprofitandwages.Whenthismetricincreases,corporationsare

redistributing national income fromworkers andwhen it declines workers are redistributing national

income fromcorporations. It is important tonote thatwhereasFigure12was comprisedofaverages,

these are aggregate figures, which means this metric captures how two of the main components of

nationalincome—wagesandprofits—arebeingsharedbetweenlaborandcapital.

36

Figure14plotsthemarkupofthetop100firms(withoutlyingvaluesremovedin1992and2002)

and the aggregate capital-labor redistribution metric. The two series are tightly and positively

intertwined over six decades. Between 1940 and 1980—the embedded liberal era—the capital-labor

redistribution metric trended downward, which signals that workers tended to win the distributive

struggleinthatperiod.The1980sservesasaninflectionpointandthen,intheperiodofafter1990—the

so-called “free trade” era—the redistributionmetric trended upward,which signals that corporations

tendedtowinthedistributivestruggleinthatperiod.Ifthemarkupisaproxyforthemarketpowerof

largefirms,andifthecapital-laborredistributionmetriccapturesthestrugglebetweenproprietorsand

workers,itishighlysignificantthattheformermovesintandemwiththelatterbecauseitsuggeststhat

corporatepowerisonedeterminantofthedistributionofAmericanincome.

Theprecedingtwosectionssubstantiatedtheassertionthattheinterplayofinstitutionalpower

and distributive conflict determines how national income is shared between various income groups.

Tradeunionpower,manifested inuniondensity and strikeactivity,progressively redistributes income

37

while corporate power, manifested in aggregate concentration and the markup, regressively

redistributesincome.Howaretheseprocessesrelatedtoinflation?

LaborIncome,CapitalistIncome,andInflation

Recall that some scholars viewed worker’s wage demands as a source of inflationary pressure while

othersarguedthat themarketpowerofoligopolisticcorporations is thedriving forcebehind inflation.

These claims were often advanced without the requisite evidence. This section will unpack these

hypotheses for the United States. We have already discussed the linkages between the institutional

powerof labor, distributive conflict, andmarket prices, noting that themarket priceof labor and the

nationalwagebillareshapedbytherelativesizeoflabororganizationsandextentofstrikeactivity.Are

there linkagesbetween laborcompensationand inflation?Andwhatrole, ifany,doesthe institutional

powerofbusinessplayingeneratinginflation?

Figure15contrasts inflationandlaborcompensationbyplottingthenominalrateofchangeof

averagehourlyearningsandtheannualrateofconsumerpriceinflationfrom1850through2013(both

series are smoothed as three-yearmoving averages). The statistical associationbetween inflation and

38

worker compensation is visually unmistakable. The correlation coefficient is very high (0.83) and

statistically significant. Note that in the embedded liberal era the annual inflation rate tended to be

relativelyhighand/orrising;intheneoliberaleraithasbeenrelativelylowand/orfalling.Averagelabor

compensationfollowedasimilarpattern.

Takentogether,Figures10–12and15indicatethatuniondensityandtheextentofstrikeactivity

shape labor compensation, and labor compensation is closely associated with inflation. When the

bargaining position of workers improved either as the result of a social crisis like a world war—a

situation in which society moves toward full employment—or when workers are able to successfully

utilizethestrikeweapontoincreasethevalueofwagesettlements,theresultisahigherproportionof

firmrevenueaccruingtoworkersintheformofwages.Thismaycompelproprietors/managerstoinflate

earnings margins to accommodate heightened wage demands. The resultant increase in worker

purchasingpowermaythenbetransmittedthroughtoahighersocietalpricelevel.

Thisfindingdoesnotimplythatworkersarethesoleorevenprimarycauseofinflation.Thefacts

portrayedinpartIofthisanalysisandinFigures4and5indicatethattherelativesizeofthetop100US-

basedfirmsispositivelyrelatedtotheirearningsmargins,whichisunderstoodtomeanthatthereisa

linear relationship between institutional size and market power. Given the evidence on the linkages

between laborpowerand inflationaswell as theearlier conclusionsof theconflict inflation theorists,

this suggests that theremayalsobea relationshipbetween theearningsof large firmsandAmerican

inflation.

Figure14 tests thevalidityof thisclaimbycontrastingAmerican inflation (measuredusing the

consumerprice index)with theaverageearningspershare (EPS)of thetop100American-listed firms.

Whereas Figure 15 contrasted inflationwith hourly earnings—aunit of ownership of labor power—in

utilizingcorporateEPS,Figure16disclosestheinflationarypressuregeneratedbylargefirmsperunitof

ownership.Thetwoseriesarepositivelycorrelatedoversixdecades.Whenweperiodizethedata,the

strength of the correlation increases. This implies that inflation and the earnings of large firms have

become increasingly synchronized over the past six decades. This also suggests that the accelerating

increase in prices between the 1960s and 1980s—the “crisis inflation” or “stagflation”—was at least

partiallydrivenbythemarketpoweroflargefirms.

Tosummarize,theevidencesuggeststhat inflation isat leastpartiallyaproductofthe income

gainsmadebyworkers andby capitalists. Labor income is shapedby strikeactivity anduniondensity

whereas the earnings of the 100 largest firms are shaped by their market power (manifested in

aggregateconcentrationandthemarkup).Therefore,thetwostreamsoftheconflictinflationliterature

39

fit ratherneatlywith theAmericanexperience. This doesnot indicate thatboth typesof inflation are

similarlypresent inallperiodsor that theyhaveanequal impacton theprice level. In theabsenceof

other intervening variables, American inflation reflects generalized forms of labor disobedience,

monopolypricingpower,or,mostlikely,acombinationofboth.Whenwecontrolforthefactthatwages

account for a much higher proportion of GDP than does corporate profits—the national wage bill is

anywherefromfourtoeighttimeslarger—twoconclusionsensue:(1)itisunlikelythatariseintheprice

level equally benefits both capital and labor; and (2) higher levels of inflation are more likely to be

reflectiveofintensifiedlabordisobedience.

WhoWinsfromInflation?

The research resultspresented thus far indicate thatAmerican inflationhasmultiplepointsof contact

withinstitutionalstructureanddistributiveconflict,whichimpliesthatthepowerdifferentialsbetween

differentgroups—notablylaborandcapital—maybesignificantifwearetounderstandthecauses(and

consequences)ofinflation.Thisbringsustoanotherimportantelementofthisargument,namely,that

40

changes in theoverallprice level are shapedby the complex interplayofpowerdifferentialsbetween

incomegroups.Onewayoftestingthisconclusionistoseeif,overthelongterm,thereisaredistributive

dimensiontoAmericaninflation.Thisentailsshiftingthemethodologicalemphasisfromabsoluteincome

gains todifferentialanddistributive incomegains.Thenecessityofdifferentialanalysis foruncovering

thepowerdimensionsofinflationisdescribedbyNitzanandBichler(2009).Theirargumentisvitaltothe

analysisthatfollowsandwequotethematlength:

…the conventional definition [of inflation] focuses wholly and only on averages and

totals. This fact is crucial, since to define inflation in this way is to miss the point

altogether. The crux of inflation is not that prices rise in general, but that they rise

differentially.…Althoughmostpricestendtoriseduringinflation,theyneverriseatthe

same rate. There is always a spread, with some prices rising faster and others more

slowly.Fromthisviewpoint, theengineof inflation isa redistributionalstruggle fought

throughrisingprices.Theoveralllevelofinflationismerelythesurfaceconsequenceof

thatstruggle.Sointheend,MiltonFriedmanisright—butonlyinpart.Inflationisalways

and everywhere a monetary phenomenon; but it is also always and everywhere a

redistributionalphenomenon.(NitzanandBichler2009,p.369)

ForNitzanandBichler,then,itispreciselybecausetheconsequencesofinflationdependontherelative

powerofcapitaland laborthattherelationshipbetweeninflationandpowercannotbeunderstoodin

absolute terms. They support their assertions by mapping postwar American inflation onto the

differential income gains of different income groups to find that inflation has tended to redistribute

incomefromworkerstocapitalistsandfromsmallfirmstolargefirms.

Shiftingouranalysisfromabsolutetodifferentialtermsmayassist indeterminingthepointsof

contact between inflation and distribution. Figure 17 plots inflation against the labor-capital

redistributionmetric, the latter computed as a ratio of average hourly earnings to the S&P 500 price

index.Bothseriesaresmoothedasthree-yearmovingaverages.Thisdifferentialindicatorcrystallizesthe

distributive strugglebetween theownersof laborpowerand theownersof corporateequity:when it

rises,laborisredistributingincomefromcapital,andwhenitfalls,capitalisredistributingincomefrom

labor.BasedontheevidencepresentedinFigures10–12andadjustingforthefactthatwagesrepresent

a considerably larger share of GDP, we would expect the labor-capital redistribution metric to rise

alongsidepositivechangesintheannualinflationrate.Thetwoseriesarepositivelycorrelatedandthe

41

strength of the statistical relationship increases over time. This supports the assertion that American

inflationisapartialmanifestationofthedistributivestrugglebetweenlaborandcapitaloverwagesand

profits.Thedistributivegainsmadebylaborbetweentheearly1960sandthelate1970sappeartohave

beenapartialdriveroftheacceleratinginflationofthatperiod,whereasthecapital-favoringdistribution

after1980appearstohavebeenapartialdriverofthedisinflationinthatperiod.

Another way of assessing whether inflation has a redistributive property is by examining the

distribution of personal income. If conflict-fueled inflation appears with the redistribution of income

fromcapitaltolabor,wewouldexpecttheincomeshareoftherichestAmericanstodecreaseintandem

with episodes of high and/or rapid inflation. The top income share-eroding aspects of inflation are

rooted in two crucial assumptions: first, the American corporate sector has controlling owners; and

second,thosewhoownthecorporatesectorpopulatetheupperechelonsofthetopincomegroup. In

otherwords,dominantcapitalistsoccupythetopofthepersonalincomehierarchyintheUnitedStates.

Figures11and12demonstratedthattheextentofstrikeactivityshapestherateofwagegrowth

and, by implication, how firm revenue is shared betweenworkers and proprietors. Figure 15 showed

42

thatworkerwagegainsareassociatedwith inflation.Figure18contrastsadifferentmeasureof strike

activity—the absolute number of strikes—with the Pareto–Lorenz coefficient between 1900 and 2013

(withbothseriessmoothedasthree-yearmovingaverages).Thequestionis:doesstrikeactivity(“labor

disobedience”) increase incomeequality?ThePareto–Lorenz coefficient captures the concentrationof

incomeamongtherich:thehigherthecoefficient,thelowertheconcentration.Overthepastcentury,

the extent of strike activity maps on very tightly to the level of income equality. This implies that

working-classstrikeactivity—whichhasademonstratively inflationaryaspect to it—contributed to the

redistributionofAmerican income from theupper to the lower incomebrackets. The two series rose

togetherfromthe1920s,peakedinthe1970sanddeclinedthereafter,reachingapostwarlowin2013.

The term is used pejoratively, but insofar as strike action involves the owners of labor power

revolting against the owners of corporate equity—workers and capitalists—this activity is a

manifestation of what the classical political economists referred to as “class struggle.” This type of

struggle iscloselyassociatedwithaninflationaryredistributionof income,suchthatthepacificationof

Americanlaborforceinrecentdecadeshasmeantstagnantwagegrowth,ashrinkingnationalwagebill

43

andasurgingtopincomeshare.ThefactsinFigure18supportthenotionthatAmericaninflationis(1)

closely associated with distributive conflict and (2) it systematically manifests the redistribution of

income.Anadditional implicationseemstobethatwhatwetodaycallthe“middleclass”wasapartial

outcomeofclassconflict.

There are a number of implications to these findings. The first implication is that heightened

distributive conflict andworkplace revolt contributed to an inflationary scramble between large firms

andlabor.Overthelonghaul,thisconflicttendedtoredistributeincomefromcapitaltolaborandfrom

theupper to the lowerechelonsof thepersonal incomehierarchy.Second, theAmericanmiddleclass

was largely built between 1940 and 1980. This was a period of relatively high and/or accelerating

inflation, driven partially by union-backed worker wage struggles. Beginning in the late 1970s, the

American state and the Federal Reserve embraced an anti-inflationarymonetary policy. If inflation is

beneficial for the working and middle classes, and if inflation is harmful to larger firms and the top

incomegroup,thenoneinterpretationoftheshifttowardanti-inflationarymonetarypolicyisastheuse

ofstatepowertoredistributeincomefromlabortocapitalandfromthelowertotheupperechelonsof

thepersonalincomehierarchy.Farfromneoliberalglobalizationimplyingthe“retreat”ofstatepower,in

thisinstancethemeaningofneoliberalismistheutilizationofstatepowertorestrainthewagedemands

oftheworkingclassandtostrengthenthesocialpositionofbusiness,especiallylargefirms.Underthis

interpretation, what is sometimes referred to as “sound monetary policy” is, in effect, working-and-

middle-class-restraining,business-class-promotingstatepoliciesthatupwardlyredistributeincome.

ConclusionsandtheReformofNeoliberalCapitalism

Let’ssummarizethefindingsofthistwo-partanalysis.Thecentripetalforcesofcorporateamalgamation

lead to the centralization of corporate ownership manifested in asset concentration. Increased

concentrationreducescompetitivepressure,thusthickeningearningsmargins,andenlargestheincome

share of large firms. The diversion of corporate resources away from growth-expanding industrial

projects puts downwardpressureon growth and leaves evenmore corporate income in thehandsof

large firms.With the rise of stock-based forms of compensation, corporate executiveswere given an

institutional incentive to divert corporate income into share price-inflating stock repurchases, which

increases the income of executives and exacerbates inequality. Andwith higher earnings, large firms

distributed comparatively more to stockholders in the form of dividends. It is in this way that the

creationofatop-heavycorporatedistributionsimultaneouslyputsdownwardpressureongrowthwhile

elevatinginequality.

44

Ontheothersideoftheledger,thetradeunionpowermanifestinbothuniondensityandstrike

activity has systematically and progressively redistributed factor and personal income, such that the

weakenedpowerofAmericanunionssincethe1970shascontributedtoheightenedincomeinequality.

We have already seen that insofar as the American corporate sector goes, the institutional-

organizational structure (aggregate concentration) is closely shadowed by themarket power of large

firms (registered in themarkup). A similar, though countervailing, set of relationships are present for

organized labor. Union density, a proxy for the institutional power of organized labor, is positively

correlatedwith incomeequality.Density increasedmodestlyand inanonlinear fashionbetween1880

and 1930. Then, with the advent of New Deal legislation such as theWagner Act, density increased

fourfold,risingfrom7percentin1933toahistorichighof29percentin1954,beforefallingtoapostwar

lowof11percentin2013.Theconcentrationofincomeamongtherichwassignificantlyerodedinthe

decadesthatAmerican laborbuilt theirunions,and inthedecadesthattradeunionpowerdiminished

incomere-concentratedamongtherich.

So what does this mean for the reform of neoliberal capitalism? In comparison with the

Keynesian state-led model of capitalism that predominated between 1935 and 1980, the neoliberal

corporate-ledmodel in theUnitedStates (andelsewhere)has, since1980,beennotable for its slower

growthandinequality.Itappearsthatatop-heavycorporatedistributionamidthedeclineoforganized

laborhaveplayedacausal role in thisdouble-sidedphenomenon.Sowhatcanbedoneto reformthe

American economy in away that unleashes growth and reduces inequality? Three baskets of policies

standout:(1)fullemployment;(2)anamplifiedvoiceforlabor,includinglawsthatnurtureunions;and

(3)capitalcontrols.

Chronic unemployment (and underemployment) has been one of themost socially damaging

featuresofneoliberalcapitalism.Thestatecommitmenttofullemploymentwoulddramaticallyalterthe

labor market in a way that empowers workers, unleashes growth, and reduces income inequality.

ConsiderthefactsinFigure19,whichcapturethedistribution-compressingeffectsoffullemploymentby

contrasting income inequality (registered inthetop0.1percent incomeshare)andtheunemployment

rate,withtheSecondWorldWarshaded ingray.Prior tothewar,bothunemploymentand inequality

were comparatively high, but in the war years both were dramatically reduced. Why would both

unemploymentandinequalityfallinatightlysynchronizedfashionoverthisshortperiod?

RecallVeblen’sclaimaboutthe“naturalrightofinvestment”:namely,thatprivateownershipof

industrial equipment grants proprietors the legal right to enforce unemployment—an act of

institutionalized exclusion that restricts production below full socioeconomic potential and alters the

45

distributionofincomeinaproprietary(businessowner)favoringmanner(seeVeblen2004[1923]).Now

considerthereconfiguredroleoftheAmericanstateduringtheSecondWorldWar,theprimehistorical

exampleofthecommitmenttofullemployment.Until1939employment(industry)wasfirmlyunderthe

control of private proprietors (business). The SecondWorldWar partially changed that insofar as the

federal government oversaw themove toward a centrally planned economy. Business considerations,

althoughnottotallyeliminated,weregreatlydiminishedvis-à-visindustrialconsiderationsandcapitalists

lost(some)controloveremploymentandpricing.

The consequences for the American power elite were devastating, as evidenced by a near

halvingof the income share going to the richest1percent. In termsofprice formation,wage ceilings

wereimposed,producerandconsumerpriceswerefrozen,exchangerateswerecontrolled,andtherate

of profit was capped. As a result, unemployment shrank from 12 percent to less than 2 percent and

income inequality fell by more than a third. Gross domestic product (adjusted for inflation and

population)grewby10percentannually,onaverage, rising from$94billion in1939to$228billion in

46

1945.Withoutexception, the1940s represents themost rapidgrowthperiod inAmericanhistoryand

theclosestitevercametofullemployment.

Being nomere reform, the capitalist powermanifest in unemploymentwas severely curtailed

through the use of state power and the consequences included rapid GDP growth and a radically

redistribution of income. In the contemporarymilieu, the pathways to full employment could include

policy tools as conventional as infrastructure investment and growth-friendlymonetary policy (Stiglitz

2015) or as unconventional as the creation of a national development bank and public ownership in

strategicindustrieslikeenergyoradvancedmanufacturing.

If the state commitment to full employment would help solve the problem of stagnation, a

stronger voice for labor would help ensure the gains from growth are widely distributed. But labor

unions need a nurturing policy environment to grow. TheWagner Act enshrined the right to bargain

collectively and compelled private sector employers to recognize the representatives of unionized

workers. The National War Labor Board’s “maintenance of membership” rule (1942) automatically

included new hires in the union so long as the union had been recognized by the employer. It was

partially because of these (and other) policies that unionizationmore than tripled between 1935 and

1945,risingfrom8percentto25percent.Laborunionsintandemwiththeexerciseofthestrikeweapon

have a demonstratively and progressively redistributive property. Antiunion legislation like the Taft–

HartleyAct in1947andmanyother contemporarymanifestationsof that legislation, has aided in the

demobilizationoforganizedlaborinthepostwarera,withregressivelyredistributiveconsequences.

Anotherpathwaytowardtherebalancingofpowerbetweenlaborandcapitalistohaveworker

representationoncorporateboards.Thismodelofcorporategovernance—thestakeholderasopposed

toshareholdermodel—iscommoninGermany(ClarkeandBostock1997).Besidestendingtoflattenthe

compensationschemewithinfirms,thismodelreducestheproclivityofmanagementtoengageinhigh-

riskactivitythatmaybelucrativeforshareholdersintheshortrun,butwhichthreatensemploymentand

the long-term viability of the organization. Volkswagen has worker representation on its board of

directors and is 20 percent owned by the state government of Lower Saxony (the “VW law”)—an

additionallayerofstakeholderaccountability.AndwhileNorthAmericanautomakershavebeenclosing

assemblyplantsfordecades,becauseVWhaslaborandgovernmentrepresentationonitsboardithas

managedtoblockhostiletakeoversandhasnotallowedasingleVWassemblyplanttocloseinGermany

since1945—andthisdespitethefactthatGermanmanufacturingworkersarepaidroughlyone-quarter

morethantheirNorthAmericancounterparts.

47

Another democratizing policy pertaining to labor would see the establishment of sector

developmentcouncilsinstrategicareas,comprisingrepresentativesfromunions,employerassociations,

government, and colleges, universities, and trades schools. These multi-stakeholder councils could

undertake initiativesas simpleas long-termskills trainingprograms (tobettermanage innovationand

emerging labor market needs) to initiatives as transformative as the setting of sectoral standards,

including minimum standards pertaining to the conditions and compensation of work. Historically,

unionsprovidedworkerswithsomedegreeofcontrolover the laborprocess in theirownworkplaces.

Sectordevelopmentcouncilswouldprovideworkerswithsomedegreeofcontrolovertheevolutionof

entireindustries.

A third transformation would constrain the size and market power of large firms through

aggressiveantitrustlegislation.Inthe40yearsbetween1940and1980,M&Aactivityconstitutedjust20

percentofinvestmentinfixedassets.Intandemwithcapitalcontrols,thisperiodwasremarkableforits

elevated levels ofGDP growth and reduced income inequality. In the decades since 1980, large firms

have plowed enormous resources into acquiring competitors. The resulting semi-monopolisticmarket

structures and attendant market power concentrates income among large firms. By restricting the

oligopolistic drive of M&A, the intention would be to unleash the centrifugal forces of fixed asset

investment, which are associated with higher levels of GDP growth, job creation, and an inclusive

prosperity.

EarlyAmericanmergerwavestendedtobefollowedbyantitrustlegislation.Americanlegislators

werefearful thatcentralizedcorporateauthoritywouldharmthepublicgood.Thedriveformonopoly

associatedwiththefirstmergerwave,forexample,ledtotheShermanAntitrustActof1890,whichwas

institutedtocombat thepowerof large firms.Ageneration later, theUSCongresspassedtheClayton

Actof1914,whichalsomadeitmoredifficulttomergeformonopoly(Gaughan2007).Anti-monopolistic

policies designed for the 21st century could help diffuse the increasingly centralized corporate power

manifestinlargefirms,whichwouldhelpaccelerategrowthanddiminishinequality.

The statecommitment to fullemployment in conjunctionwitha rebuildingof the tradeunion

movement and restrictions on business consolidation are three democratizing policies that would

weakenthepoweroflargefirmswithouteliminatingthecapitalistcharacteroftheAmericaneconomy.

Therewouldstillbeprivateenterprise,investmentforprofit,andwagelabor,butthisbasketofpolicies

would make the accumulation of capital compatible with the national goals of widening economic

opportunityandaninclusiveprosperity.

48

Appendix

DataSources,Calculations,andEstimationTechniques

Buy-to-BuildIndicator(Figures1,5,13)

To the best of the author’s knowledge, internally consistent long-term data are not available for

AmericanM&A.ThedollarvalueofallreportedM&AcomesfromMergerStatthroughWilmerHaleLLP

(2014,Slide2) for2008–13andfromThomsonSecuritiesFinancialDatathroughthesoftwarepackage

associatedwith Gaughan (2007, Figure 2.7) for 2000–7. The estimated dollar value ofM&A for prior

yearscameinaseriesofsteps.Thefirststepwastocreateaunifiedandrebasedseriesofannounced

M&Atransactions.M&Aannouncements come fromGaughan (2007,Tables2.3and2.5) for1963–99.

Lamoreaux (2006a,chapter422;2006b,chapter416)suppliesM&Atransactions inmanufacturingand

mining for 1919–62 and 1895–1918, respectively. The second step involved multiplying the S&P 500

CompositePriceIndex(GlobalFinancialData,Code:GFD_SPXD)bythetotalnumberofannouncedM&A.

Inthethirdstep,arebasingnumberwascreatedsothatthetotalnumberofM&Acouldbemultipliedby

theS&P500proxyvalue. Instepfour,theresultingnumberwasmultipliedbytherebasingnumberto

produceanestimateofthedollarvalueofM&Afrom1895through1999.Stepfiveinvolvedcreatinga

unifiedseriesforthedollarvalueofgrossprivatedomesticinvestmentinnonresidentialstructuresand

equipment by splicing the Bureau of Economic Analysis, Table 1.1.5, from 1929–2013with a rebased

valueofSimonKuznets’s(n.d.)estimateofgrossfixedcapitalformation(1895–1928),thelatterbrought

totheauthor’sattentionbyFrancis(2013).

NationalAccounts(Figures2–3,6-8,10,14)

Total American corporate assets are the sum of nonfinancial and financial corporate business assets,

retrieved from theFederalReserve (Z1/Z1/FL102000005.AandZ1/Z1/FL792000095.A).Grossdomestic

productandbusinessexpenditureonnonresidentialstructuresandequipmentare fromtheBureauof

EconomicAnalysis(BEA),Table1.1.5.Nationalincomeand(national)netdividendsfromBEA,Table1.12.

Relevant seriesweredeflatedusing theconsumerprice indexandan industrial commoditiesproducer

price index, retrievedfromGlobalFinancialData (CodesCPUSAMandWPUSAICM).Civilian labor force

andtotalemploymentfromCarter(2006).

49

Top100Firms(Figures4–7,9,13-14,16)

Dataforthetop100American-listedfirms,rankedannuallybyequitymarketcapitalization,comesfrom

Compustat through Wharton Research Data Services. Variables include common shares outstanding,

closingshareprice,revenue,totalassets,acquisitions,pretaxprofit,aftertaxprofit,capitalexpenditures

andpurchaseofcommonandpreferredstock.

TopIncomeShare(Figures8–10,14,18-19)

Top1 percent income share (including capital gains),wages and salaries portionof the top 1 percent

incomeshare,andthePareto–LorenzCoefficientfromPikettyandSaez(2007),retrievedfromWorldTop

IncomesDatabase:topincomes.g-mond.parisschoolofeconomics.eu/,accessedDecember23,2014.

UnionDensity(Figure10)

To the best of the author’s knowledge, a continuous data series is not available for American

unionization.Membershipdatafrom1973–2013arefromHirschandMacpherson(2003),retrievedfrom

theironlinedatabase(www.unionstats.com/).Mayer(2004,TableA1,pp.22–23)suppliesmembership

datafrom1930–72.Estimatesfor1880–1929cameinaseriesofsteps.Thefirststepwastospliceand

rebaseabsolute figures forunionmembership from theBureauof Labor StatisticsHandbookof Labor

Statistics(1950)andG.Friedman,“NewEstimatesofUnionMembership:TheUnitedStates,1880–1914”

Historical Methods 32, no. 2 (1999), both retrieved from Rosenbloom (2006a, chapters Ba4783 and

Ba4789). The second step was to estimate the American workforce. Sobek (2006, chapter Ba340)

provides an absolute figure for eachdecadebeginning in 1850.Adecadeaverage rateof growthwas

imputedtocreateacompleteestimateoftheAmericanworkforce.Inthethirdstep,unionmembership

wasdividedbytheestimatedlaborforcetoarriveatafigureforuniondensity.

StrikeActivity(Figures11-12,18)

Workstoppagesincludestrikesandlockouts.Tothebestoftheauthor’sknowledge,internallyconsistent

long-termdataarenotavailableforAmericanworkstoppages.TheBureauofLaborStatistics(BLS)tracks

labor disputes from 1947 onward using fourmetrics: number ofwork stoppages, number ofworkers

involved,daysidle,andpercentofestimatedworkingtimelost.Thedifficultywiththesefiguresisthat

thenumberofworkstoppagesisrestrictedtoestablishmentswith1,000ormoreemployees.Between

automation, lean production, and the resulting decline in large establishments, the data will tend to

artificiallydepress strike intensityover recentdecades. To remedy thisdefect, anewstrike indexwas

50

built.Continuousestimatesfortheabsolutenumberof(1)workstoppages(from1881), (2)numberof

workersinvolved(from1881),(3)daysidle(from1927),and(4)percentofestimatedworkingtimelost

(from 1939), with proper rebasing, was created by fusing the BLS data with data from the Historical

StatisticsoftheUnitedStates.Withaviewtocombiningthefourmeasuresofstrikeactivityintoasingle

supermetric,eachserieswasindexed(theactualvalueineachyearlesstheseriesaveragevaluedivided

bythestandarddeviation).Intheresultingfourseries,avalueof0meansstrikeactivityisatitshistoric

average,withpositiveandnegativevaluessignallingaboveaverageandbelowaveragestrikeactivity.A

simpleaverageofthefour indexedserieswascomputedtocapturetheoverallstriketrendinAmerica

from1881to2013(withthedatagapfrom1906to1913interpolatedwithastraightline).SeeBureauof

Labor Statistics, Table 1, for data from 1947 to 2013 (www.bls.gov/news.release/wkstp.t01.htm),

accessedMay5,2015.Dataonnumberofworkstoppagesandworkersinvolved(1881–1905,1914–81),

person-days idle (1927–81), and days idle as a percent of estimated working time (1939–81) from

Rosenbloom(2006b,chaptersBa4954-4958).

ClassIncomeandItsRedistribution(Figures12,14,17)

Aggregate Capital-Labor Redistribution Index. National corporate profits (with inventory valuation

adjustmentandcapitalconsumptionallowance)andtotalcompensationofemployeesfromtheBureau

ofEconomicAnalysis,Tables1.12and2.1,respectively.AverageLabor-CapitalRedistributionIndex.S&P

500CompositePrice Index fromGlobal FinancialData (Code:GFD_SPXD).Averagehourlyearnings for

the years 1850–1940 are for unskilled labor, retrieved fromMargo (2006, chapter Ba4218); the years

1940–2013areformanufacturingworkers,retrievedfromGlobalFinancialData(Code:USAHEMANM).

InflationandUnemployment(Figures15-17,19)

Consumer price index fromGlobal Financial Data (Code: CPUSAM). Number of unemployed and total

civilianlaborforcefromCarter(2006).

51

References

Berle,A.A.,andG.C.Means.1968(1932).TheModernCorporationandPrivateProperty.Rev.ed.New

York:Harcourt,Brace&World.

Blair, J.M. 1972. Economic Concentration: Structure, Behavior, and Public Policy. New York: Harcourt

BraceJovanovich.

Brennan,J.2012.AShrinkingUniverse:HowConcentratedCorporatePowerIsShapingIncomeInequality

inCanada.Ottawa:CanadianCentreforPolicyAlternatives.

———. 2014. “The Business of Power: CanadianMultinationals in the Postwar Era.” Ph.D. diss., York

University.

———. 2015. Ascent of Giants: NAFTA, Corporate Power, and the Growing Income Gap. Ottawa:

CanadianCentreforPolicyAlternatives.

Burdekin, R. C., and P. Burkett. 1996. Distributional Conflict and Inflation: Theoretical and Historical

Perspectives.Houndmills,Basingstoke:PalgraveMacmillan.

Card,D.,andJ.E.DiNardo.2002.“SkillBiasedTechnologicalChangeandRisingWageInequality:Some

ProblemsandPuzzles.”JournalofLaborEconomics20,no.4(October):733–83.

Carter, S. B. 2006. “Labor Force, Employment, and Unemployment: 1890–1990.” In S. B. Carter, S. S.

Garter,M. R. Haines, A. L. Olmstead, R. Sutch, and G.Wright, eds.Historical Statistics of the

UnitedStates,EarliestTimestothePresent:MillennialEdition.NewYork:CambridgeUniversity

Press.TableBa470-477.

Clarke,T.,andR.Bostock.1997.“GovernanceinGermany:TheFoundationsofCorporateStructure?”In

K. Keasey, S. Thompson, andM.Wright, eds. Corporate Governance: Economic,Management

andFinancialIssues.NewYork:OxfordUniversityPress.

Dalziel, P. 1999–2000. “A Post Keynesian Theory of Asset Price Inflation with Endogenous Money.”

JournalofPostKeynesianEconomics22(2):227–45.

DeLong,J.B.,andL.Summers.1992.“EquipmentInvestmentandEconomicGrowth:HowStrongIsthe

Nexus?”BrookingsPapersonEconomicActivity23(2):157–211.

Dougherty,C.,andD.W.Jorgenson.1997.“ThereIsNoSilverBullet:InvestmentandGrowthintheG-7.”

NationalInstituteEconomicReview,no.162(October):57–74.

Fix,B.2015.“InequalityandCapitalistIncome.”Blogpost,April18.Availableat:

http://www.capitalaspower.com/forum/viewtopic.php?t=427&p=1147#p1147

Francis,J.A.2013.“TheBuy-to-BuildIndicator:NewEstimatesforBritainandtheUnitedStates.”Review

ofCapitalasPower1(1):63–72.

52

Frydman, C., and D. Jenter. 2010. “CEO Compensation.” NBERWorking Paper No. 16585. Cambridge:

NationalBureauofEconomicResearch.AccessedMay12,2015.www.nber.org/papers/w16585.

Gabaix,X.,andA.Landier.2008.“WhyHasCEOPayIncreasedSoMuch?”QuarterlyJournalofEconomics

123(1):49–100.

Galbraith, J. K. 1952. American Capitalism: The Concept of Countervailing Power. Boston: Houghton

Mifflin.

Gaughan,P.A.2007.Mergers,Acquisitions,andCorporateRestructuring.4thed.Hoboken:JohnWiley&

Sons.

Hirsch,B.T.,andD.A.Macpherson.2003.“UnionMembershipandCoverageDatabasefromtheCurrent

PopulationSurvey:Note.”IndustrialandLaborRelationsReview56(2):349–54.Seriesupdated

bysameauthors.Dataavailableat:www.unionstats.com/.AccessedMay5,2015.

Jo, T.-H., and J. F. Henry. 2015. “The Business Enterprise in the Age ofMoney-Manager Capitalism.”

JournalofEconomicIssues49(1):23–46.

Jorgenson,D.W.2007.“InformationTechnologyandtheG7Economies.”InE.R.BerndtandC.R.Hulten,

eds.Hard-to-MeasureGoodsandServices:EssaysinHonorofZviGriliches.Chicago:Universityof

ChicagoPress.

Kalecki,M. 1971 (1938). “DistributionofNational Income.” InSelected Essays on theDynamics of the

CapitalistEconomy,1933–1970.Cambridge:CambridgeUniversityPress.

———.1971(1943).“CostsandPrices.” InSelectedEssaysontheDynamicsof theCapitalistEconomy,

1933–1970.Cambridge:CambridgeUniversityPress.

Krugman,P.R.2008.“TradeandWages,Reconsidered.”BrookingsPaperonEconomicActivity(Spring):

103–37.

Kuznets, S. n.d. “Annual Estimates, 1869–1955.” Technical tables underlying series in Capital in the

AmericanEconomy:ItsFormationandFinancing.Baltimore:JohnsHopkinsUniversity.Accessed

May6,2015.www.nber.org/data-appendix/c1454/appendix.pdf.

Lamoreaux,N.R.2006a.“Mergers,Acquisitions,andJointVentures—NumberandAssets:1919–1979.”

InS.B.Carter,S.S.Garter,M.R.Haines,A.L.Olmstead,R.Sutch,andG.Wright,eds.Historical

Statistics of the United States, Earliest Times to the Present: Millennial Edition. New York:

CambridgeUniversityPress,TableCh422-429.

53

———.2006b.“Mergers inManufacturingandMining—Entities,Capitalization,andType:1895–1930.”

InS.B.Carter,S.S.Garter,M.R.Haines,A.L.Olmstead,R.Sutch,andG.Wright,eds.Historical

Statistics of the United States, Earliest Times to the Present: Millennial Edition. New York:

CambridgeUniversityPress,TableCh416-421.

Lavoie,M.1992.FoundationsofPostKeynesianEconomicAnalysis.Aldershot:EdwardElgar.

———.2014.Post-KeynesianEconomics:NewFoundations.Northampton:EdwardElgar.

Margo,R.A.2006.“IndexofMoneyWagesforUnskilledLabor:1774–1974.”InS.B.Carter,S.S.Garter,

M. R. Haines, A. L. Olmstead, R. Sutch, and G.Wright, eds.Historical Statistics of the United

States,EarliestTimes to thePresent:MillennialEdition.NewYork:CambridgeUniversityPress,

TableBa4218.

Mayer, G. 2004. Union Membership Trends in the United States. Washington, D.C.: Congressional

ResearchService.AccessedMay5,2015.digitalcommons.ilr.cornell.edu/key_workplace/174.

McCarthy,K.J.2013.“TheBusinessEnvironment—MergersandMergerWaves:ACenturyofCauseand

Effect.” InK. J.McCarthyandW.Dolfsma,eds.UnderstandingMergersandAcquisitions in the

21stCentury.London:PalgraveMacmillan.

Means,G.C.1935.“PriceInflexibilityandtheRequirementsofaStabilizingMonetaryPolicy.”Journalof

theAmericanStatisticalAssociation30(190):401–13.

———.1983.“CorporatePowerintheMarketplace.”JournalofLawandEconomics26(2):467–85.

Moore,M.P.,andP.Ranjan.2005.“GlobalisationvsSkill-biasedTechnologicalChange:Implicationsfor

UnemploymentandWageInequality.”TheEconomicJournal115(503):391–422.

Murphy, K. J. 2012. “Executive Compensation: Where We Are, and How We Got There.” In G.

Constantinides, M. Harris, and R. Stulz, eds. Handbook of the Economics of Finance. Elsevier

ScienceNorthHolland(forthcoming);MarshallSchoolofBusinessWorkingPaperNo.FBE07.12.

AccessedMay12,2015.dx.doi.org/10.2139/ssrn.2041679.

NitzanJ.,andS.Bichler.2009.CapitalasPower:AStudyofOrderandCreorder.London:Routledge.

Olson,M. 1982. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities.

NewHaven:YaleUniversityPress.

Piketty,T.,andE.Saez.2007.“IncomeandWage Inequality in theUnitedStates,1913–2002.” InA.B.

Atkinson and T. Piketty, eds. Top Incomes Over the Twentieth Century. A Contrast Between

Continental European and English-speaking Countries, chap. 5. New York: Oxford University

Press.Seriesupdatedbythesameauthors.

Peterson,W.C.1980.“StagflationandtheCrisisofCapitalism.”ReviewofSocialEconomy38(3):277–87.

54

———.1982.“WorkerDiscontent,Wages,andStagflation.”JournalofEconomicIssues16(4):977–84.

Polanyi,K.2001(1944).TheGreatTransformation:ThePoliticalandEconomicOriginsofOurTime.2nd

ed.Boston:BeaconPress.

Pollin,R.2012.BacktoFullEmployment.Boston:TheMITPress.

Smith,A.1994(1776).AnInquiryintotheNatureandCausesoftheWealthofNations.EditedbyEdwin

Cannan.NewYork:RandomHouse.

Ricardo,D.1973(1817).ThePrinciplesofPoliticalEconomyandTaxation.IntroductionbyDonaldWinch.

Toronto:DentDutton.

Rosenberg,S., andT.E.Weisskopf.1981. “AConflictTheoryApproach to Inflation in thePostwarU.S.

Economy.”AmericanEconomicReview71(2):42–47.

Rosenbloom,J.L.2006a.“UnionMembership:1880–1999.”InS.B.Carter,S.S.Garter,M.R.Haines,A.

L.Olmstead,R.Sutch,andG.Wright,eds.HistoricalStatisticsoftheUnitedStates,EarliestTimes

tothePresent:MillennialEdition.NewYork:CambridgeUniversityPress,TableBa4783-4791.

———. 2006b. “Work Stoppages, Workers Involved, Average Duration, and Person-days Idle: 1881–

1998.” In S. B. Carter, S. S.Garter,M. R.Haines,A. L.Olmstead, R. Sutch andG.Wright, eds.

Historical Statistics of theUnited States, Earliest Times to the Present:Millennial Edition. New

York:CambridgeUniversityPress,TableBa4954-4964.

Rowthorn,R.E.1977.“Conflict,InflationandMoney.”CambridgeJournalofEconomics1(3):215–39.

Samuelson,P.A.,andW.D.Nordhaus.2010.Economics.19thed.NewYork:McGraw-Hill.

Sobek, M. 2006. “Labor Force Participation, Employment, and Unemployment, by Sex: 1850–1990

[Censusestimates].” InS.B.Carter,S.S.Garter,M.R.Haines,A.L.Olmstead,R.Sutch,andG.

Wright, eds.Historical Statistics of the United States, Earliest Times to the Present:Millennial

Edition.NewYork:CambridgeUniversityPress,TableBa340-354.

Stigler,G.1950.“MonopolyandOligopolybyMerger.”AmericanEconomicReview40(2):23–34.

Stiglitz, J. 2015. Re-writing the Rules of the American Economy: An Agenda for Growth and Shared

Prosperity.NewYork:W.W.Norton.

Veblen,T.1908a.“OntheNatureofCapital.”TheQuarterlyJournalofEconomics22(4):517–42.

———.1908b. “On theNatureofCapital: Investment, IntangibleAssets,and thePecuniaryMagnate.”

TheQuarterlyJournalofEconomics23(1):104–36.

———.2004(1923).AbsenteeOwnershipandBusinessEnterpriseinRecentTimes:TheCaseofAmerica.

NewBrunswickandLondon:TransactionPublishers.

55

Weintraub, S. 1978. Capitalism’s Inflation and Unemployment Crisis: Beyond Monetarism and

Keynesianism.London:Addison-Wesley.

———.1978–79.“TheMissingTheoryofWages.”JournalofPostKeynesianEconomics1(2):59–78.

Wilmer Cutler Pickering Hale and Dorr LLP. 2014. 2014 M&A Report. Accessed May 5, 2015.

www.wilmerhale.com/2014MAreport.