ReichEconomics of Racism

8/3/2019 ReichEconomics of Racism

1/6

-1-

The Economics of Racism

by Michael Reich

Michael Reich is Professor of Political Economy at U. C. Berkeley. This article was written in1974, hence used only data available at that time. His research was repeated on later census data inhis book Racial Inequality: A Political-Economic Analysis, (Princeton, N.J.: Princeton UniversityPress), 1981, and by other researchers, with similar results. His main thesis is that most workers areharmed by racism, regardless of their race.Reich believes this analysis to be based on ideas fromMarx, but whether Marx would agree with his assumptions about how wages are determined is debat-able.

In the early 1960s it seemed to manythat the elimination of racism in the U.S. wasproceeding without requiring a radical restruc-turing of the entire society. There was a grow-

ing civil rights movement, and hundreds ofthousands of blacks were moving to Northerncities where discrimination was supposedlyless severe than in the South. Government re-ports pointed to the rapid improvement in thelevels of black schooling as blacks moved outof the South: in 1966 the gap between the me-dian years of schooling of black males aged 25to 29 and white males in the same age grouphad shrunk to one-quarter the size of the gapthat had existed in 1960.1 By 1970, how-ever, the optimism of earlier decades had van-

ished. Despite new civil rights laws, elaborateWhite House conferences, special ghettomanpower programs, the War on Poverty, andstepped-up tokenist hiring, racism and theeconomic exploitation of blacks has not less-ened. During the past twenty-five years therehas been virtually no permanent improvementin the relative economic position of blacks inAmerica. Median black incomes have beenfluctuating at a level between 47 percent and63 percent of median white incomes, the ratiorising during economic expansions and falling

to previous low levels during recessions.2

Seg-regation in schools and neighborhoods hasbeen steadily increasing in almost all cities,

1U.S. Department of Labor, Bureau of Labor Statis-

tics, Report No. 375, "The Social and EconomicStatus of Negroes in the United States, 1969," p.50.2

The data refer to male incomes: see Table 10-A,p. 360.

and the atmosphere of distrust between blacksand whites has been intensifying. Racism, in-stead of disappearing, seems to be on the in-crease.

Besides systematically subjugatingblacks so that their median income is 55 per-cent that of whites, racism is of profound im-portance for the distribution of income amongwhite landowners, capitalists, and workers. Forexample, racism clearly benefits owners ofhousing in the ghetto where blacks have nochoice but to pay higher rents there than ischarged to whites for comparable housingelsewhere in the city. But more importantly,racism is a key mechanism for the stabilizationof capitalism and the legitimization of inequal-

ity. We shall return to the question of whobenefits from racism later, but first we shall re-view some of the economic means used tosubjugate blacks.

THE PERVASIVENESS OF RACISMBeginning in the first grade, blacks go

to schools of inferior quality and obtain little ofthe basic training and skills needed in the labormarket. Finding schools of little relevance,more in need of immediate income, and lessable anyway to finance their way through

school, the average black student still dropsout at a lower grade than his white counterpart.In 1974 only 8.1 percent of blacks aged 25 to34 were college graduates, compared to 21.0percent of whites in the same age bracket.3

Exploitation really begins in earnestwhen the black youth enters the labor market.

3U.S. Bureau of the Census, Series P-60, "Educa-

tional Attainment."


2/6

-2-

A black worker with the same number of yearsof schooling and the same scores on achieve-ment tests as a white worker receives muchless income. The black worker cannot get asgood a job because the better-paying jobs arelocated too far from the ghetto or because heor she was turned down by racist personnel

agencies and employers or because a uniondenied admittance or maybe because of anarrest record. Going to school after a certainpoint doesn't seem to increase a black per-son's income possibilities very much. The moreeducated a black person is, the greater is thedisparity between his income and that of awhite with the same schooling. The result: in1966 black college graduates earned less thanwhite high school dropouts.4 And the higher theaverage wage or salary of an occupation, thelower the percentage of workers in that occu-

pation who are black.The rate of unemployment among

blacks is generally twice as high as amongwhites.5 Layoffs and recessions hit blacks withtwice the impact they hit whites, since blacksare the "last hired, first fired." The ratio of aver-age black to white incomes follows the busi-ness cycle closely, buffering white workersfrom some of the impact of the recession.

Blacks pay higher rents for inferiorhousing, higher prices in ghetto stores, higherinsurance premiums, higher interest rates in

banks and lending companies, travel longerdistances at greater expense to their jobs, suf-fer from inferior garbage collection and lessaccess to public recreational facilities, and areassessed at higher property tax rates whenthey own housing. Beyond this, blacks are fur-ther harassed by police, the courts, and theprisons.

When conventional economists attemptto analyze racism they usually begin by tryingto separate various forms of racial discrimina-tion. For example, they define "pure wage dis-

crimination" as the racial differential in wagespaid to equivalent workersthat is, those withsimilar years and quality of schooling, skilltraining, previous employment experience and

4U.S. Bureau of the Census, Series P-60, "Income

in 1966 of Families and Persons in the UnitedStates."5

See, for example, U.S. Department of Labor,Manpower Report of the President, various years.

seniority, age, health, job attitudes, and a hostof other factors. They presume that they cananalyze the sources of "pure wage discrimina-tion" without simultaneously analyzing the ex-tent to which discrimination also affects thefactors they hold constant.

But such a technique distorts reality.

The various forms of discrimination are notseparable in real life. Employers' hiring andpromotion practices; resource allocation in cityschools; the structure of transportation sys-tems; residential segregation and housingquality; availability of decent health care; be-havior of policemen and judges; foremen'sprejudices; images of blacks presented in themedia and the schools; price gouging in ghettostoresthese and the other forms of socialand economic discrimination interact stronglywith each other in determining the occupational

status and annual income, and welfare, ofblack people. The processes are not simplyadditive but are mutually reinforcing. Often, adecrease in one narrow form of discriminationis accompanied by an increase in anotherform. Since all aspects of racism interact, ananalysis of racism should incorporate all its as-pects in a unified manner.

No single quantitative index could ade-quately measure racism in all its social, cul-tural, psychological, and economic dimensions.But while racism is far more than a narrow

economic phenomenon, it does have verydefinite economic consequences: blacks havefar lower incomes than whites. The ratio ofmedian black to median white incomes thusprovides a rough, but useful, quantitative indexof the economic consequences of racism forblacks. We shall use this index statistically toanalyze the causes of racism's persistence inthe United States. While this approach over-emphasizes the economic aspects of racism, itis nevertheless an improvement over the nar-rower approach taken by conventional econo-

mists.

COMPETING EXPLANATIONS OF RACISMHow is the historical persistence of ra-

cism in the United States to be explained? Themost prominent analysis of discriminationamong economists was formulated in 1957 byGary Becker in his book, The Economics of


3/6


4/6

-4-

cisely the opposite. In the latter approach ra-cism has an equalizing effect on the white in-come distribution, whereas in the former ra-cism has a disequalizing effect. The statisticalrelationship between the extent of racism andthe degree of inequality among whites providesa simple yet clear test of the two approaches.

This section describes that test and its results.First, we need a measure of racism.

The index we use, for reasons already men-tioned, is the ratio of black median family in-come to white median family income (abbrevi-ated as B/W). A low numerical value for thisratio indicates a high degree of racism. Wecalculated values of this racism index usingdata from the 1960 Census, for each of thelargest forty-eight metropolitan areas (bounda-ries are defined by the U.S. Census Bureau,who use the term standard metropolitan statis-

tical areasSMSA's). There is a great deal ofvariation from SMSA to SMSA in the B/W in-dex of racism, even within the North; SouthernSMSA's generally demonstrated a greater de-gree of racism. The statistical techniques usedare based on this variation.

We also need measures of inequalityamong whites. Two convenient measures are:(1) the percentage share of all white incomethat is received by the top 1percent of whitefamilies; and (2) the Gini coefficient of whiteincomes, a measure which captures inequality

within as well as between social classes.8

Both of these inequality measures vary

considerably among the SMSA's; there is alsoa substantial amount of variation in thesewithin the subsample of Northern SMSA's.Therefore, it is very interesting to examinewhether the pattern of variation of the inequal-ity and racism variables can be explained bycausal hypotheses. This is our first source ofempirical evidence.

A systematic relationship acrossSMSA's between our measure of racism and

either measure of white inequality does existand is highly significant: where racism isgreater, income inequality among whites is

8The Gini coefficient varies between 0 and 1, with 0

indicating perfect equality and 1 indicating perfectinequality. For a more complete exposition, see H.Miller, Income Distribution in the United States(Washington, D.C.: U.S. Government Printing Of-fice, 1966).

also greater.9 This result is consistent with ourmodel and is inconsistent with the predictionsof Becker's model.

This evidence, however, should not beaccepted too quickly. The correlations reportedmay not reflect actual causality since other in-dependent forces may be simultaneously influ-

encing both variables in the same way. As isthe case with many other statistical analyses,the model must be expanded to control forsuch other factors. We know from previous in-ter-SMSA income distribution studies that themost important additional factors that shouldbe introduced into our model are: (1) the indus-trial and occupational structure of the SMSA's;(2) the region in which the SMSA's are located;(3) the average income of the SMSA's; and (4)the proportion of the SMSA population that isblack. These factors were introduced into the

model by the technique of multiple regressionanalysis. Separate equations were estimatedwith the Gini index and the top 1 percent shareas measures of white inequality.

All the equations showed strikingly uni-form statistical results: racism as we havemeasured it was a significantly disequalizingforce on the white income distribution, evenwhen other factors were held constant. A I per-cent increase in the ratio of black to white me-dian incomes (that is, a 1 percent decrease inracism) was associated with a .2 percent de-

crease in white inequality, as measured by theGini coefficient. The corresponding effect ontop 1 percent share of white income was twoand a half times as large, indicating that mostof the inequality among whites generated byracism was associated with increased incomefor the richest 1 percent of white families. Fur-ther statistical investigation reveals that in-creases in the racism variable had an insignifi-cant effect on the. share received by the poor-est whites and resulted in a decrease in theincome share of the whites in the middle in-

come brackets.10

This is true even when the

9For example, the correlation coefficient between

the B/W measure of racism and the Gini coefficientof white incomes is r = -.47. A similar calculation byS. Bowles, across states instead of SMSA's, re-sulted in an r = -.58.10

A more rigorous presentation of these and othervariables and the statistical results in available inMichael Reich, "Racial Discrimination and the White


5/6

-5-

Southern SMSA's are excluded.Within our model, we can specify a

number of mechanisms that further explain thestatistical finding that racism increases inequal-ity among whites. We shall consider twomechanisms here: (1) total wages of white la-bor are reduced by racial antagonisms, in part

because union growth and labor militancy areinhibited; (2) the supply of public services, es-pecially in education, available to low- andmiddle-income whites is reduced as a result ofracial antagonisms.

Wages of white labor are lessened byracism because the fear of a cheaper and un-deremployed black labor supply in the area isinvoked by employers when labor presents itswage demands. Racial antagonisms on theshop floor deflect attention from labor griev-ances related to working conditions, permitting

employers to cut costs. Racial divisions amonglabor prevent the development of united workerorganizations both within the workplace and inthe labor movement as a whole. As a result,union strength and union militancy will be lessthe greater the extent of racism. A historicalexample of this process is the already men-tioned use of racial and ethnic divisions to de-stroy the. solidarity of the 1919 steel strikers.By contrast, during the 1890s, black-whiteclass solidarity greatly aided mineworkers inbuilding militant unions among workers in Ala-

bama, West Virginia, and other coalfield ar-eas.11

The above argument and examplescontradict the common belief that an exclu-sionary racial policy will strengthen rather thanweaken the bargaining power of unions. Racialexclusion increases bargaining power onlywhen entry into an occupation or industry canbe effectively limited. Industrial-type unions aremuch less able to restrict entry than craft un-ions or organizations such as the AmericanMedical Association. This is not to deny that

much of organized labor is egregiously racistor that some skilled craft workers benefit fromracism.12 But it is important to distinguish ac-

Income Distribution" (Unpublished Ph.D. diss., Har-vard University, 1973).11

See footnote 44 above.12

See, for example, H. Hill, "The Racial Practices ofOrganized Labor: the Contemporary Record," inThe Negro and the American Labor Movement, ed.

tual discriminatory practice from the objectiveeconomic self-interest of most union members.

The second mechanism we shall con-sider concerns the allocation of expendituresfor public services. The most important ofthese services is education. Racial antago-nisms dilute both the desire and the ability of

poor white parents to improve educational op-portunities for their children. Antagonisms be-tween blacks and poor whites drive wedgesbetween the two groups and reduce their abil-ity to join in a united political movement press-ing for improved and more equal education.Moreover, many poor whites recognize thathowever inferior their own schools, blackschools are even worse. This provides somedegree of satisfaction and identification withthe status quo, reducing the desire of poorwhites to press politically for better schools in

their neighborhoods. Ghettos tend to be lo-cated near poor white neighborhoods moreoften than near rich white neighborhoods; ra-cism thus reduces the potential tax base ofschool districts containing poor whites. Also,pressure by teachers' groups to improve allpoor schools is reduced by racial antagonismsbetween predominantly white teaching staffsand black children and parents.13

The statistical validity of the abovemechanisms can be tested in a causal model.The effect of racism on unionism is tested by

estimating an equation in which the percentageof the SMSA labor force that is unionized is thedependent variable, with racism and the struc-tural variables (such as the SMSA industrialstructure) as the independent variables. Theschooling mechanism is tested by estimating asimilar equation in which the dependent vari-able is inequality in years of schooling com-pleted among white males aged 25 to 29.

Once again, the results of this statisticaltest strongly confirm the hypothesis of ourmodel. The racism variable is statistically sig-

nificant in all the equations and has the pre-

J. Jacobson (New York: Anchor, 1968).13

In a similar fashion, racial antagonisms reducethe political pressure on governmental agencies toprovide other public services that would have a pro-poor distributional impact. The two principal items inthis category are public health services and welfarepayments in the Aid to Families with DependentChildren program.


6/6

-6-

dicted sign: a greater degree of racism resultsin lower unionization rates and greater degreeof schooling inequality among whites. Thisempirical evidence again suggests that racismis in the economic interests of capitalists andother rich whites and against the economic in-terests of poor whites and white workers.

However, a full assessment of the im-portance of racism for capitalism would proba-bly conclude that the primary significance ofracism is not strictly economic. The simpleeconomics of racism does not explain whymany workers seem to be so vehemently rac-ist, when racism is not in their economic self-interest. In non-economic ways, racism helpsto legitimize inequality, alienation, and power-lessnesslegitimization that is necessary forthe stability of the capitalist system as a whole.For example, many whites believe that welfare

payments to blacks are a far more importantfactor in their taxes than is military spending.Through racism, poor whites come to believethat their poverty is caused by blacks who arewilling to take away their jobs, and at lowerwages, thus concealing the fact that a substan-tial amount of income inequality is inevitable ina capitalist society. Racism thus transfers thelocus of whites' resentment towards blacks and

away from capitalism.Racism also provides some psycho-

logical benefits to poor and working-classwhites. For example, the opportunity to partici-pate in another's oppression compensates forone's own misery. There is a parallel here tothe subjugation of women in the family: after a

day of alienating labor, the tired husband cancompensate by oppressing his wife. Further-more, not being at the bottom of the heap issome solace for an unsatisfying life; this argu-ment was successfully used by the Southernoligarchy against poor whites allied with blacksin the interracial Populist movement of the latenineteenth century.

Thus, racism is likely to take firm root ina society that breeds an individualistic andcompetitive ethos. In general, blacks provide aconvenient and visible scapegoat for problems

that actually derive from the institutions of capi-talism. As long as building a real alternative tocapitalism does not seem feasible to mostwhites, we can expect that identifiable and vul-nerable scapegoats will prove functional to thestatus quo. These non-economic factors thusneatly dovetail with the economic aspects ofracism discussed earlier in their mutual serviceto the perpetuation of capitalism.

Reichs Later Research

Using later Census data, Reich was able to make a statistical estimate of the effect of tworace-related variables to the rate of profit for employers. BW = the ratio of black to white median fam-ily income in a given industry. %NW = the percent of non-white workers in an industry. We then havethe following equation for the rate of profit in various industries:

Rate of Profit = 20.5% - 14% x BW - 21.7% x %NW

This equation implies that more racial inequality is associated with higher profits, and also that a lowerpercentage of non-white workers in an industry is associated with higher profits.

See M. Reich, Racism: A Political Economic Analysis, Princeton, 1981.

ReichEconomics of Racism

Documents

Transcript of ReichEconomics of Racism