Category Archives: Social Class


Video: Race and Crime


Video: Cultural Criminology


Keeping it Radical: Exploring Income, Income Inequality and Poverty Data for the US.

Keeping it Radical: Exploring Income, Income Inequality and Poverty Data for the US

Michael J. Lynch (Department of Criminology, University of South Florida)

This study reports on data reported in:
“Income, Poverty, and Health Insurance Coverage in the United States, 2007”
US Bureau of the Census.
By Carmen DeNavas-Watt
Bernadette D. Proctor
Jessica C.. Smith
Current Population Reports. (P-60-235).

August, 2008.

The following discussion presents a summary along with some implications from a portion of the data present in the 2008 US Bureau of the Census report “Income, Poverty and Health Insurance Coverage in the United States, 2007” (August; Current Population Reports, P-60-235; see  These summaries and interpretations are offered to aid other criminologists in obtaining, using and understanding US income, poverty and inequality data.

The use and interpretation of income, income inequality and poverty data are important aspects of doing radical criminology. Radical criminology, in order to strip away ideological images and reveal the material conditions in a society that impact the life patterns, life-styles, life courses, and the life chances and choice the ordinary people who live in and make up the majority of society, must continually refer to and use data that addresses dimensions of class relationships and power upon which society is built. Doing so requires employing data to address the structural relationships and forms of equality that support the basis of inequality in a class-based society.

In recent years, much of critical criminology has attempted to shift the analysis of crime, justice, law, social control and life issues to the non-material conditions of life. In those views, for example, the making of culture and the construction and limits of language as a structure has replace the focus on class relations, power relations, and other structural dimensions of inequality as the most important theoretical and practical concerns. This has lead to an extra-ordinarily wide-spread neglect of more traditional materialist theoretical and empirical work, though such work continues to remain a prominent concern in other disciplines.

Criminological perspectives based in post-modern and cultural explanations have contributed to undermining class analysis within criminology. In addition, these approaches offer less of a challenge to orthodox criminology, and one rarely noted refer to this work within the orthodox criminological literature. Perhaps those engaged in these forms of critical criminology are uninterested in this observation or in having their work effect orthodox theorizing.  In such a case, critical criminology has abandoned one of its central concerns, or at least a central concern of its radical variants, which is not only to get to the root of social problems, but to understand those problems in ways that force a change in how they are understood, and upon which active social policies may be built (See Lynch and Michalowski, 2006, Primer in Radical Criminology, Monsey, NY: Criminal Justice Press).  This is based on Marx’s notion that “the point isn’t to interpret the world. The point is to change it.” It is toward this end that there continues to be a need to keep abreast of, examine, interpret and describe the uses of data on class relationships such as income, income inequality and poverty.

Median Household income over time.

From 1967 to 2007, or over 40 years, real median household income in the US rose $ 11,462 or from $38,771 in 1967 to $50,233 by 2007. This fact, in itself, seems quite significant, and on some level appears to indicate the economic advancement of the average family in the US over the past four decades. Consistent with, it seems, capitalist ideology, in the long run, people in a capitalist society are better off. As capitalism advances, as more value is produced, these values become more widely distributed, appearing, for example, as a “trickle-down” economic effect which enhances the standing of the average person in society.

In reality, when we interpret and reassess the meaning of the rise in median household income over the past forty years, it becomes clear why the above interpretation is not only misleading, but entirely inadequate as an explanation of what has happened to median family income over the past forty years.

To understand what has happened to median family income over the past forty years, let us first transform the rise in median family income into a percent gain.  Over this 40 year period, real median household income increased by 29.6 %, which appears to be a substantial gain.  However, we must realize that this gain, this additional $ 11,462, occurred over a very long time period, and in order to get a better picture of what this means, we need to divide the 20.6% gain (or the $ 11,462 gain) by the forty years over which this gain occurred. Doing so, we find that the nearly 30% gain in median income experienced by families was less than a 1 percent (0.73%) annual gain. Moreover, on an annual basis, the expansion of median family income amounts to only $286.55 per year.

Thus, once we consider the average annual gain, the income gained by the average US family over the past forty years turns out to be quite small. In fact, it is even smaller than these statistic, which treat the gain annually, reveals.

Why are the gains made by the average family even smaller than they appear? In short, because in calculating the average annual gains over the past forty years we have yet to take account of inflation.  Indeed, if we examine the rate of inflation we find that the gain in income falls well below the rate of inflation for this time period. Consequently, this slow rate of income growth that occurred over the past forty years indicates that the average US family lost ground to inflation over this time period, or that the average American family in 2007 was worse off economically than the average American family in 1967.

How much worse off was the average family in 2007 compared to 1967? Much worse off than most Americans would imagine. This can be illustrated in two ways: by examining actual changes in the median family-inflation ratio, and by examining a hypothetical situation where the rate of inflation is set to a predetermined rate.

Let us begin with the predetermined hypothetical rate scenario since here we will set the rate of inflation quite low in order to illustrate the effect of the real rate of inflation on mean family income. In the predetermined example, let us set inflation to only 1% per year or just slightly greater than the annual rate of increase in real median family income for this period which was 0.73%.  Here, we are interested in discovering the effect of a rate of inflation that is only fractionally higher than the rate of growth in income – a rate which is just 0.27% higher per annum.

At a 1% annual inflation rate, the average family is only losing 0.27% a year to inflation, which seems like a rather small amount. Even with this small difference, a family earning $38,771 in 1967 would need to earn $57,208.17 in 2007 to keep pace with inflation where inflation is a mere 1%.  Yet, in 2007 the median American family earned only $50,233, or found themselves $6,975 behind the rate of inflation IF the rate of inflation were only 1 percent per year. At only 1%, then, the median inflated household income of $57,208 was nearly 14 percent higher than actual median family income for this time period ($50,233). And, unfortunately for the average American family, the rate of inflation was been much higher on an annual basis throughout this time period than the hypothetical 1 percent rate this example employed..

What was the real difference between the 40 year rate of inflation in the US and the 40 year gain in median family income? For the period 1967-2007, the average real annual inflation rate was 4.67%, substantially higher than the low inflations rate of 1% set in our predetermined example.  According to the Bureau of Labor Statistics inflation calculator (, the buying power of $38,771 in 1967 – the median family income in 1967 – –  was equivalent to the buying power of $240,684 in 2007 !!!! In other words, once inflation is considered, median family income would need to increase by 520% between 1967 and 2007 to keep pace with 1967 purchasing power. As noted, between 1967 and 2007, median family income rose by only about 30% to $50,223. This figure is more than $190,000 less than the median family income needed just to keep pace with the real rate of inflation during this time period.

Over time, then, the average American family lost substantial ground in terms of income.

Today’s family is far worse off than the typical family in 1967.  Thus rather than hard work producing substantial gains for most Americans, hard work produces substantial declines in family income; rather than the vast wealth produced in the US over this time period being shared, it appears that this is not the case; rather than getting a head, the average family fell far behind. On top of these unfortunate circumstance, these income losses were distributed differently across households with distinct characteristics, and some households suffered more so than others.

In short, the average US family was worst off in 2007 than in 1967, so much worse off that this may be hard to contemplate.  To some extent, knowing that this is true helps explain other significant issues in the US economy which are not the focus of the present discussion such as the extraordinary expansion of credit during this time period (Note: the rapid and widespread expansion of credit is needed to continually fuel economic expansion in a capitalist economy where the share of income accumulated by the average family is shrinking; that is, as economic class inequality expands and relative income for the mass of families decline, this contradiction can only be solved by the expansion of credit, which at some points created its own problem since the amount of credit extended to the mass of families will need to surpass the income of those families in order to maintain expanding consumption, which will clearly cause an economic collapse, especially in an economy where the transition from a manufacturing base is already well underway, and where the only option is to become an exporting rather than an importing nation, or, in contrast to the Obama economic stimulus package, to constrain domestic consumption while promoting foreign consumption of US products. But, this is the material for an entirely separate analysis).

Household Characteristics and Median Family Income, 2007. 

Related and Unrelated Households. For 2007, median family income varied significantly with household characteristics. For example, the median household income for married couples was $ 72,785, a figure that is nearly 45% higher than overall or non-differentiated median family income for 2007. Compared to married couples, median household income was significantly lower for other household relationship categories such as male-headed households without a wife ($49,839; less than 1 percent lower than median family income, but 31.5% lower than married couple median household income); female-headed, no husband households ($33,330;  33% lower than male-headed households, and 54% lower than married couple median household income). In comparison, non-family member household had the lowest median income at $30,176.

Race and Ethnicity. Significant variation also exists between median household income by race/ethnicity of household. For example, the median income of Asian households was $66,103; for White-non-Hispanic households median income was $54,920 (about 17% lower); for Black households $33,916 (approximately 49% lower than for Asian households and 38% lower than White-non-Hispanic households). In addition, median family income for Hispanic households of any race ($33,679) were similar to median household incomes for Blacks, though among all racial and ethnic groups, median Hispanic household income was the lowest.

Racial and ethnic differences in household income appear fairly substantial, and actually exaggerate racial and ethnic difference in income.  Indeed, racial and ethnic distinctions in income are less evident in per capita median income comparisons made across racial and ethnic groups. It is important to note that the difference between per capita and median household income for race and ethnicity comparisons are affected by the number of wage earners each method reflects. For example, there is no standard number of wage earners in a household, and a higher median household income across racial or ethnic groups may, for example, reflect the number of wage earners in an average household for each racial and ethnic group rather than the actual difference in income for individuals in each racial and ethnic group. In order to account for variations in household size across racial and ethnic groups, median per capita income levels should, therefore, also be consider since these remove the effects of multiple wage earners included within median household calculations.

The Census indicates that on a per capita basis, White, non-Hispanics earned the highest mean income in 2007 — $31,051, compared to Asians ($29,901), Blacks ($18,428) and Hispanics of any race ($15,603), which provides a slightly different rank order of racial/ethnic groups when measuring mean income.


Despite the difference between per capita incomes by race/ethnicity and median household incomes by race/ethnicity, the relative rank order of the relationship between race/ethnicity and income is fairly consistent across these methods, although these methods would lead to a fairly significant difference with respect to indentifying the top earning racial/ethnic group. The following table shows the differences in income ranks for median household income compared to per capita income by race and ethnicity for 2007:

Rank by Race/Ethnicity by Different Income Measures

Per Capita                   Median Household

White, Non Hispanic                      1                                        2

Asian                                             2                                        1

Black, Non-Hispanic                      3                                        3

Hispanic, Any Race                        4                                        4

One reason that race/ethnicity estimates of income by family and individual varies is related to the effect of the number of wage earners within families of different racial and ethnic origins. The data on median household and per capita incomes can be used to estimate the number of wage earners in a household. This is accomplished by taking a ratio of the median household income to per capita income by racial and ethnic groups. For example, for white, non-Hispanics, the median household/per capita income ratio is 1.77, meaning that, on average, White-non-Hispanic families are expected to contain 1.77 income earners.  For Asians, the ratio is 2.21, meaning that Asian households on average have 2.21 income earners. For Blacks, the ratio is 1.84, indicating an average of 1.84 income earners in the average Black household.  Finally, for Hispanic families of any race, the family/per capita ratio is 2.16 indicating that for Hispanic families there are an average of 2.16 income earners per household.

Income Inequality

An important concern is the extent to which income is distributed unequally. Generally, it is assumed that the distribution of income reflects effort or the idea that economic reward is commensurate with the amount of work one does, or in what we will call the work-reward paradigm. Above, this idea was challenged by, for example, evidence that differentials in income were seen across both families and individuals depending upon race and ethnicity. In the work-reward paradigm, the existence of these differentials must mean that (in the household income example) Asians work harder than Whites, who work harder than Blacks who work harder than Hispanics.  We could, of course, address this issue directly if data on hours worked by race were available, but even hours worked data may not indicate how hard someone works or how hard they prepared for that work, or even the degree of difficulty associated with a particular form of work, such as its requirement for hard, physical labor on a sustained, repeated basis.

In any event, here we are less concerned with the reasons for inequality and instead focus on two primary issues. First, establishing that income inequality exists and second that contrary to what many believe, income inequality in the US is fairly extensive.

Since our focus here is on establishing the nature of income inequality in general and in demonstrating the class issues involved, we will leave aside issues that blend income inequality and race/ethnicity. Although this is obviously an important concern it is beyond the scope of the present discussion.

In order to describe income inequality, it is necessary to have an indicator or measure of the extent of inequality. Such a measure is readily available in the GINI coefficient of income inequality, or for short, the GINI coefficient.  While the GINI coefficient provides one widely employed measure of income inequality, it is not an intuitive measure and requires some explanation. Consequently, here let us instead examine the distribution of incomes by comparing the percentage of income found in each population quintile.

A population quintile is made up of 20% of the population. In a society where incomes are fairly evenly distributed, one would expect that the percentage of income associated with each quintile would be nearly identical to the percentage of the population within a quintile. In other word, if we took the lowest 20% of wage earners in a society where income was distributed evenly, we would expect that this would also represent about 20% of the population more generally. That is, twenty percent of the population would receive twenty percent of the income in a society with an equity based income distribution.

Unfortunately, in the modern word income is not distributed so evenly. Indeed, income is often distributed quite unevenly. Sometimes, the inequality in the distribution of income is so lopsided that we are shocked to discover its true extent.

As noted, in order to examine the distribution of income across the population, we will compare income across population quintiles. In the Census data examined here (Series P-60-235), the income data is family based meaning that the inequality and income measures represent income inequity across families rather than individuals.

If we divide the US population of families into quintiles or fifths, with each fifth representing 20% of families, we might expect to see some evidence of unequally distributed incomes given that in the US there is an assumption that incomes are based on “work effort” rather than “need” or “equity” divisions (here, we will not address the assumption of distribution of income reflecting work effort, and instead simply note that much work has critiques the basis of this assumption such as Robert Frank And Phillip Cooks, The Winner Take all Society: Why the Few at the Top Get So Much More than the Rest of US). Indeed, when we examine the relationship between income quintiles and family quintiles (see Table 1), there is only one quintile – the fourth quintile – in which the distribution of the (family based) population and income is fairly equivalent. In the fourth quintile, which is the second highest income group, consists of 20% of families and 23.4% of the income for the US. (*One of the issues that cannot be address here is the relationship between individuals and income since these data are based on families).

Interestingly, the income for the 4th quintile is only slightly skewed or unequal. If, for example, we defined equity (e) as the ratio of income percentage (p i) to the family quintile percentage (p f-q) or as e = p i / p f-q, then e would be 1 in all cases where there is equality because p i  = p f-q.  In contrast, when the income distribution is skewed so that a quintile is receiving income greater than its family distribution, it will have an equity skewness index (esi) greater than 1.  For the 4th quintile, the ESI is 23.4/20 = 1.17. In other words, the ESI of 1.17 indicates that each 1 percentile in the quintile (recall that a quintile is made up of 20 percentiles) earns 0.17 percent above the state of equity or equal income distribution.

If we reexamine Table 1 paying attention to the ESI,  we see that the first three quintiles or the lowest family income quintiles have ESIs below 1, indicating that these quintiles receive less income than expected if income were evenly distributed. How uneven is the income distribution? Fairly uneven —  the first quintile or 20 percent of families receives only 3.4 percent of all income; the second quintile, 8.7 percent of income, and the third quintile, 14.8 percent of income. All together, this 60 percent of families receives only 26.9 percent of all income.

The problem of inequity at the bottom end where income is under distributed necessarily involves income being over-distributed at the top end. In other words, the problem for the bottom one-fifth and by extension the bottom three-fifths of the population is that the top quintile keeps so much income. The problem, in other words, is that the fifth or highest income quintile earns an extraordinarily unequal share of income. But, more importantly, within this quintile is the influence of the top 5 percent of wage earners. While the top 20 percent of families take home almost half – 49.7 percent – of all income, the top 5 percent of families alone take home 21.5 percent of all income.

The problem this data displays is that the poor are very poor and that their condition is a consequence of having a small group of families that are very rich. We will return to this point below.

Another way to illustrate the extent of this poverty and inequity produced by the current distribution of incomes across families is to create in Table (which shows each quintile, the percent of income for each quintile, population percentage grouping, and the ESI) a measure of dollar per family distribution of income (DPFDI) and the DPFDI ratio.  To derive the DPFDI and its ratio, we simply impose an assumption over these data which is a mathematical representation of these data: namely that each population percent is equivalent to one family. All this assumption accomplishes is that it translates a percentage into a family so that we create a population of 100 families that directly reflect the income distribution found in society. In other words, in the first group (the old lowest quartile) there is 20 percent of the population, or in a population of 100 families, 20 families, and so on across each quartile with the exception of the 5th quartile which was split into two groups to allow an examination of income inequality for the highest 5 percent of families (or in this case 5 persons) in the population of 100 families.

The DPPDI represents the dollar amount earned by each family.  It is calculated as a mean relative to each category, and reflects the amount earned for each $100,000 of gross income across all 100 families.  The DPPDI ratio measures income inequality relative to the mean income of a member of the top 5 percent.  The DPPDI shows, for example, that for every dollar earned by a family in the lowest quintile (bottom 20 percent; Q 1), a person in the top 5 percent of society (Q 5-2) earns $101.2.

The DPFDI indicates that even among the wealthiest 20 percent of families (Q 5-1 and Q 5-2) there is a large disparity in the distribution of income. Compared to the top 5 percent of wage earners (Q 5-2), the remaining portion of the top quintile of income earners (Q 5-1) makes, on average, $ 6.9 dollars less for every dollar earned by a Q 5-2 family.  In the event this sum looks paltry, consider that income earners – the top one percent of individual earners – earned an average income of $1.1 million in 2005. That same year, the top 1/10 th of one percent of top individual income earners – about 300,000 people — took home an average of $ 5.6 million each.  At an even more extreme level, the top 1/100th percent of income earners – the top 30,000 income earners in the US – made, on average, $25.7 million in 2005 (  In other words, if you make $6.9 dollars for every dollar I make, and you made $ 1.1 million dollars, then I would have made only $159,430, or $ 940,540 less. That $940,540 dollars is the difference between being in the top 1 percent of wage earners and being in the 80-95 percentile range.

Table 1: Income Distributions by Quintiles* (% Population), Equity Rations (ESI), and Per Family (DPPDI), 2007

% Population               %Income         ESI      Mean $            DPFDI                        DPFDI Ratio

___________              ________        ____    _______          ______            __________

20 (Q 1)                       3.4                   0.17     $  3,400           $ 170                    101.2

20 (Q 2)                       8.7                   0.435   $  8,700           $ 435                      39.5

20 (Q 3)                       14.8                 0.74     $14,800           $ 740                      23.2

20 (Q 4)                       23.4                 1.17     $23,400           $ 1,170                   14.7

15 (Q 5-1)                   28.2                 1.88     $37,600           $ 2,507                     6.9

5 (Q 5-2)                   21.5                 4.30     $86,000           $ 17,200               ——–

* The quintile measure (Q 1, Q 2, etc.,) is divided for the 5th quintile into two parts: (1) Q 5-1, or the bottom 15 percent of the 5th quintile, and (2) Q 5-2, the top 5 percent of the 5th quintile or the richest 5% of the entire population of families.

Income Inequality Reconsidered

There is, in the US, much concern that any effort to redistribute income would adversely impact the majority of society, producing, for example, a disincentive for innovation, productive activity, and economic investment.  Furthermore, the assumption that investment by the rich is needed to stimulate continued economic growth has become a more common and widespread assumptions since the 1980s, when the super-rich among the population were granted “tax incentives” (tax cuts) to accelerate economic growth through “trickle down” spending. However, as the recent recessionary period of the 21st century has indicated, this is indeed not the case; the rich cannot and do not save the economy from crisis. Indeed, tax cuts to the rich appear to restrict capital turnover by causing a higher proportion of net income to be absorbed into saving and investment among those with elevated incomes, or which only stimulate narrow portions of the economy tied to luxury consumption (in the way, for example, suggested by Thorsten Veblen in his 1899 book, Theory of the Leisure Class,

Regardless of these assumptions, income in the US was, earlier in the twentieth century, much more evenly distributed in the US.  One of the factors that contributed to greater income equality was the highly progressive tax structure that characterized early 20th century America (see, US Federal Individual Income Tax Rates History, 1913-2009, Moreover, despite this highly progressive tax rate, economic investment was not discouraged, individuals amassed great wealth, and the US economy expanded and grew at a steady, and if not at time a rapid pace.

With respect to the progressive tax structure in early 20th century America, consider that in 1917 the income tax rate for persons earning over $500,000 was 54% to 67%. A tax rate in excess of 50% or more on incomes of more than $500,000 was in effect until 1924 before it was lowered to 25% for income earners over $100,000. The 25% tax rate remained in effect until 1932, when the tax rates for upper income groups (greater than $100,000) reverted to 56% to 63%.

The tax break for the rich which occurred from 1924 through 1931 was an historical aberration during the early 20th century. In 1936, the tax rate for upper income levels increased. Income rates for all income levels rose during World War II. By 1950, these persistent tax increases meant that a married couple filing jointly earning in excess of $150,000 was taxed at a rate of 90 percent.

In the 1950s, as economic growth accelerated, tax rates declined – which was somewhat a paradoxical policy since extensive economic growth after WWII occurred despite high tax rates. By the mid-1950s, tax reform efforts pushed the income levels for those taxed at the highest level (still 90 percent or more) to $300,000. Granted, $300,000 was a fairly high income, and in 2009 dollars was equivalent to approximately $2.3 million dollars. Nevertheless, through the mid-1950s, income taxes helped redistribute incomes and were extraordinarily progressive, especially at the highest income levels.

The 1964 tax revisions lowered the tax rate on those earning $200,00 or more to 76.5%, effective excluding those with the highest incomes from paying the 90% tax rate that was historically established. Further tax reductions for upper income individuals occurred in 1965. Tax rates for the highest income groups remained relative stable for the next 15 years.

Significant tax reductions for upper income groups and a shift to accelerated tax rates for lower income groups occurred in the early 1980s. In the early 1980s both the rate of taxation and the level of taxable income affected by tax rates were radically changed. For example, the tax rate for families in the highest tax bracket was lowered to $85,600 – which effectively expanded a portion of the tax base at the lowest end of the upper income levels for families —  while the tax rate for this newly defined high income group declined to 50 percent.  By 1984 the income level for the highest income tax rate nearly doubled to $162,400, and increased through 1987. Further tax reductions for the highest income groups were achieved in the major federal tax reforms of 1988 which reduced the tax rate on families earning more than $150,000 to only 28 percent. This reduction was significant for several reasons. First, it marked the lowest tax rate on the highest income earners in the 20th century. Second, this kind of tax structure clearly indicated that taxation would no longer be used to help maintain an equitable distribution of incomes in the US. Third, the regression of the tax rate on the upper income group also indicated that  there was no longer an assumption that it was part of the wealthy’s responsibilities to employ their good fortune for the benefit of society or for the greatest good, or that the government should help reinforce that value system through a progressive tax structure. Finally, the declining tax rate on  the upper income group also made it evident that the state would now function more openly in defense of the power of the upper income and propertied classes.

It should be noted that the drastic tax cuts on upper income groups that occurred in 1988 were marked by a second tax strategy that revealed the basis and class bias in the new tax structure. That bias was evident in the tax rate paid by the second highest tax bracket – families earning from $71,900-$149,999 – which paid a higher income tax rate of 33 percent or 5% more than families in the highest income bracket. In effect, in the late 1980s, income earners got a tax break for being richer than other income earners, a practice that was revoked by 1992 federal tax reforms.  Currently, the highest effective tax rate is 35 percent on income earners (married filing jointly/families) receiving more than $372,950.

As indicated above, tax rates in the first half of the 20th century were much higher for the highest income levels – in a number of years exceeding 90% for those at the top of the income scale. This level of taxation did not derail America’s economic growth and, perhaps, contributed to greater income equity by lowering the tax rates for lower income groups and by redistributing incomes. In recent years, however, there has been an increased tendency to believe that high tax rates retard economic growth even though evidence prior to the mid-1960s suggests that this assumption is incorrect. Perhaps there is also an assumption that progressive tax rates on upper income groups appeared “socialistic,” even though this kind of progressive rate of taxation was evident in American capitalism well before the emergence of socialist/communist nations.

Income Inequality, Taxation and Poverty

One of the reasons to remain concerned with income inequality and regressive tax rates on the wealthiest income earners is the effect of these outcomes on poverty in the US.  For a nation with an overall high level of wealth and which ranks among the nations with the highest mean and per capita income levels, there remains extensive poverty in the US. In 2007, 12.5% of the US population lived below the officially identified poverty level (in 2008-2009, poverty thresholds were increased to $10,830 for a single individual;  $14,570 for a family of two; $18,310 for a family of three; $22,050 for a family of four).  For children under 18, the poverty rate increases to 17.6%, and for those under 6 to 20.8%.

Poverty and race/ethnicity are highly interrelated in the US. While the general rate of poverty in the US was 12.5%, the poverty rate for non-Hispanic Whites was significantly lower — 8.2%. For Blacks (24.5%) and Hispanics of any race (21.5%), poverty rates are significantly higher than the mean US poverty rate.

Finally, there is also a gender component to poverty. For example, 13.6% of male head single parent households lived under the poverty level. For female headed single parent households the rate of poverty was more than twice the rate fro men, 28.3%.


Household income, economic inequality, taxation and poverty data are central to understanding the ways in which US society remains divided into distinct classes with varied access to economic resources. Economic inequality and class inequality accompany one another.

It is, at this point, beyond my intention to make much of the data reviewed above, especially within the specific context of criminological research. Nevertheless, I would be remiss if I completely dismissed making any observations relevant to the use of these data for furthering observations pertinent to radical criminology. Thus, in conclusion let me offer these brief observations which I hope to return to at a future date.

First, in recent years, economic inequality and class relationships have played an increasingly larger role in producing class-linked forms of social control. Second, these class-linked social controls have become increasingly necessary in a society that is more widely divided by economic and class inequities. The massive prison expansion of recent decades, for example, has much to do with controlling the marginalized populations produced during the restructuring of American capitalism from an industrial to a service basis. Indeed, it is necessary during this restructuring which expands the general marginalization of the work force, that formal mechanisms of social control that provide for the physical constraint of the marginal classes expand as well (for further discussion see, Lynch, Michael J., Big Prisons, Big Dreams, Rutgers University Press, 2007).  Third, based on these observations and keeping in mind the assumptions of radical economic and political analysis, we can link these observations together noting that, indeed, it is no accident that in the US prison expansion coincides with the decline of manufacturing.  The year 1973, for instance, marks the beginning of the intensive decline of the manufacturing sector, the beginning of the end of oil period in the US, and the first year of the current stretch of 36 years of unbroken increases in the rate of punishment in the US.

In closing, let me also be clear about one of the assumptions that has greatly impacted radical analysis within criminology. That assumption is that the US, and the world more generally, has entered a new phase of developed, one that philosophers and other social analysts have identified as the “post-modern” world. Unfortunately, there is really nothing post-modern about the current world, which is merely a restructured version of capitalism rather than any form of post-modern world. To some extent nations may have become post-industrial – and these nations would include the US. But, a post-industrial nation and a post-modern nation are two different beings.  Furthermore, it may be a mistake to refer to nations such as the US as post-industrial since we have no idea, for example, how long this state may last, or how changing world relationship impact the transfer of industrial production within the world, which at some point may cause the redevelopment of industrial capitalism in the US. To be sure, it is likely that the “reindustrialization” of the US may occur or has already begun to occur in ways that look unfamiliar to us since, as economists who analyze such issues imply, part of this industrial restructuring involves the emergence of post-Fordist techniques of production. While post-Fordist production has had some impact on the organization of US industry, it has not played as central a role here as in other nations, and the future path of the US economy remains quite unsettled at this point. What these future developments means for crime and social control is, at this point, anyone’s guess.


Class, Race, and Gender in Criminology and Criminal Justice: Ways of Seeing Difference

Class, Race, and Gender in Criminology and Criminal Justice: Ways of Seeing Difference

Gregg Barak, Eastern Michigan University

The following is a Symposium Speech delivered at the Second Annual Conference on RACE, GENDER and CLASS Project in New Orleans on October 20, 2000.


In the post-modern and multicultural worlds of criminology and criminal justice characterized by post-structuralism, post-Marxism, post-affirmative action, and post-feminism, the variables of class, race, and gender remain fundamental to both theory and practice. After all, the disciplines of criminology and the fields of criminal justice have always been about the real and imagined differences between “criminals” and “non-criminals.” Theoretically, explanations of crime and crime control, regardless of perspective or school of thought, have sought to make sense out of these differences. In the process of trying to sort out these differences, virtually every theoretical framework has addressed class and race overtly, and gender at least covertly. Up until recently, the problem with this line of inquiry was not only that there had been very little, if any, agreement on the effects of these three critical variables, but worse yet, folks were still debating whether or not these variables matter.

By the turn of the 21st century, however, a growing number of criminologists from several orientations, including but not limited to critical, feminist, Marxist, positivist, and integrative, had come to appreciate, in different yet related ways, that class, race, and gender matter. Today, many inquiries are interested in finding out just how exactly class, race, and gender matter in the production of crime and criminal justice. Some inquiries focus on class, race, and gender as autonomous variables. Some inquiries focus on these three variables as inter-related. Of course, key questions on the complexities of these relations and on the means of exploring them still remain. And even though the ways of seeing difference or of approaching class, race, and gender vary, there is certainly an emerging consensus on the importance of these three variables, and increasingly, on the intersections between them or on their interactive or reciprocal relationships.

In fact, it is my contention that in present-day criminology and criminal justice, there are at least four approaches to the study of class, race, and gender: (1) quantitative studies; (2) time and place studies; (3) ethnographic studies, and (4) social construction studies. These approaches are part and parcel of older and newer traditions in the study of crime and criminal justice. At the same time, they are also reflective or representative, over the last twenty years or so, of larger movements in academia to distance itself from both essentialism and determinism. Finally, each of these approaches is capable, more or less, of studying class and crime, race and crime, or gender and crime as separate or related phenomena. Whether or not class, race, and gender are studied in isolation or in relation to each other, depends to a very large extent on the kinds of questions that are asked by each of these approaches. When class, race, and gender are studied together, the way they are linked or connected will also depend on the questions asked.

Quantitative Studies:

Quantitative studies in crime and criminal justice are concerned with empirically measuring, capturing, or nullifying the “casual relationships” or “differencing effects” of class, race, or gender on crime, delinquency, violence, law enforcement, adjudication, sentencing, and punishment. In the tradition of “functionalism,” crime and crime control are viewed as indexes of misconduct. Typically, class, race, and/or gender become the “independent” variables and crime and crime control become the “dependent” variables. In the tradition of “positivist” social science, these studies normally engage large data sets involving samples that number into the three or four digits. Data, in the case of crime, is generally gathered through responses to “self-reported” questionnaires, that strive to represent as best as possible the real world breakdowns or population demographics. In the case of the administration of criminal justice, data stems mostly from the FBI’s Uniform Crime Reports, the U.S. Department of Justice, and other governmental documents produced locally or at the state level. Inevitably, there are usually some limitations that restrict generalization such as under or over representation of particular socio-economic group/s.

For the purposes of this presentation, I turned to the May (2000) issue of Criminology, the official quarterly publication of the American Society of Criminology (ASC), and discovered that five out of the ten articles dealt with class, race, and/or gender. Four of those were quantitative studies: “Minority Threat and Police Brutality: Determinants of Civil Rights Complaints in U.S. Municipalities”; “Gender, Structural Disadvantage, and Urban Crime: Do Macrosocial Variables also Explain Female Offending Rates?”; “Perceived Sanction Threats, Gender, and Crime: A Test and Elaboration of Power-Control Theory”; and “The Myth of Social Class and Crime Revisited: An Examination of Class and Adult Criminality.” It is interesting to note, though each of these studies was primarily concerned with race, or gender, or class, that three of them addressed all three variables, and the fourth addressed two out of the three.

In order to convey the complexity and sophistication of these and other quantitative studies, permit me to elaborate a bit more from one of these articles. In this way, I hope to illustrate how thematically there is an emerging consensus around the importance of class, race, and gender, even among quantitative researchers whose studies have tended to nullify the importance of these variables. For example, in “The Myth of Social Class and Crime Revisited,” Dunaway et al. (2000: 589) concluded from a sample of 555 adults living in a large, midwestern city that “regardless of how class or crime were measured, social class exerted little direct influence on adult criminality in the general population.” Nevertheless, in their analysis, the authors revealed a number of caveats about their findings, such as: “The lack of both significant class effects and any race effects in our general crime scale may suggest a possible interaction effect between social class and race” (Ibid: 607).

Regarding the contemporary sophistication of quantitative studies, the authors in this study employed 15 individual indicators of social class, broken down into four gradational class measures (e.g., personal income, family income, SES, education), four underclass measures (e.g., unemployed, welfare, foodstamps, public housing), and seven Marxian class measures (i.e., bourgeoisie, petty bourgeoisie, worker, self-employed). True to quantitative form, they conducted multivariate analyses with respect to general crime measures for the year prior to their survey and since their respondents turned 18. In addition, they conducted multiple regressions on the impact of their social class measures on the prevalence of violent crime as a subscale in their analysis. Finally, as would be expected, they controlled for race, sex, and age. They also controlled for being a parent and for being married.

As for the authors’ caveats or qualifications, their study revealed a number of interesting findings that expose the complexities of class, race, and gender. For example, even though the results left a relatively weak overall impression of direct class impact on general crime, the outcomes were able to show which of the three ways of conceptualizing class, fared the best. The study also found that the “respondents’ gender and age were the most important predictors of crime,” and that “family income was the only class measure observed to significantly affect the incidence of crime in the past year,” (Ibid: 600, 602).

In addition, the study found some support for specifying the class-crime relationships by gender and by race. In the case of men and women, personal income negatively and significantly affected crime for males. By contrast, family income, significantly influenced crime for females. In the case of whites and nonwhites, social class was related to criminal involvement for nonwhites. Finally, the study also points to a significant limitation with respect to its general representativeness: the sample undercounted nonwhites. The percent of nonwhites surveyed was 14.1% compared with the community’s percent nonwhites of 35.3%.

In sum, sophisticated studies like that of Dunaway et al., intimate that there is not only a need for further quantitative research to consider the conditions under which social class is criminogenic, but for the importance of doing qualitative studies as well. As the authors, echoing the insights of John Hagan’s 1991 Presidential address to the ASC on “The Poverty of a Classless Criminology,” underscore in their discussion portion of the article, the direct effect of class on crime is also mediated by cultural and contextual factors and, therefore, it is inevitably bound to be weak! Recent work, such as Wright et al. (1999), suggests “that the actual direction of class effects may be dependent on an array of social psychological factors. Thus, the class/crime relationship may be masked by interactive effects” related to a host of other unaccounted for variables (Ibid: 624).

Time and Place Studies:

Unlike quantitative studies in crime and criminal justice, time and place studies are not engaged in the perennial pursuit of a fine-tuned measure of real crime, nor are they as reluctant to reach “definitive” conclusions regarding the relationships among class, race, gender, and crime. Whether these studies are historical or comparative, incorporate long-term or short-term perspectives, they are concerned empirically with explaining the varying levels of criminal punishment. These studies want to account for why some men or women, or some socio-economic classes, or some racial or ethnic groups, have been more likely or less likely to be sanctioned by the criminal justice system for their involvement in non-conforming behavior. These accounts typically relate the differences in crime control to their structural and institutional relations of class, race, and gender, rather than to their individual or interpersonal relations.

Accordingly, focus shifts away from measurement of crime and crime control as responses to individual or group misconduct in micro society. Instead, crime and social control are viewed in relationship to the dominant political, economic, and social interests of macro society. In the tradition of “insiders” versus “outsiders,” time and place studies want to know how the changing institutionalized relations of social control in general, and in the administration of criminal justice in particular, have been used by the more powerful groups to maintain privilege and inequality in the context of “social conflict.” Thus, time and place studies of class, race, and gender change the emphasis of inquiry from “social conduct” to “social standing,” and to the ways in which institutions of social control reproduce relations of the status quo. In these inquiries, the pursuit of data revolves around exposing the real and perceived relationships between the cultural threats of the “dangerous groups” and the mechanisms of individual and group control, all played out within the context of the prevailing political and economic arrangements.

For the purposes of this presentation, allow me to make passing reference to two recently published anthologies in the areas of criminology and criminal justice. The first is Ethnicity, Race, and Crime: Perspectives Across Time and Place (1995), edited by Darnell Hawkins. The second is Race, Gender, and Class in Criminology: The Intersection (1996), edited by Martin Schwartz and Dragan Milovanovic. What both of these edited readers share in common are “sociology of knowledge” approaches to the study of crime and crime control. That is, each text takes a “critical” stance in relationship to the disciplines of criminology and criminal justice. Moreover, both books not only explore the history, but they reflect on the ways in which the disciplines of criminology and criminal justice have studied, viewed, and treated crime and crime control in relation to the politics and/or ideology of race, ethnicity, class, and gender. Time and place, also become analytical constructs for these books as each consciously set out to include a diversity of intellectual perspectives on the conceptualizations of class, race, and gender relations. Finally, both of these anthologies move back and forth in time and space as they examine concrete applications of and practices in social control.

Comparatively speaking, the objectives of these two books vary, yet they are still related. Ethnicity, Race, and Crime wanted to distinguish not only between the physical and social realities of racial difference, but to look beyond the usual white and non-white distinctions, to include comparative complexities of multiple ethnic group experiences in crime and social control over time. Toward this end, ethnic and racial, and class, and to a lesser extent, gender experiences in social control are analyzed with consideration to the changing relations of inequality and the changing conditions of the socioeconomic orders. Hence, labor market inequalities, distribution of jobs, economic disadvantage, isolation, marginality, moral panics, institutionalized racism, poverty, and more, are brought into the time and place discussions. Finally, the interrelationship among ethnicity, race, and crime is examined in the contexts of the United States, France, and Germany. A sampling of the chapter titles from the “contemporary issues and debates” section of Ethnicity, Race, and Crime reads as: “Ethnicity, Labor Markets, and Crime”; “Crime Control and Ethnic Minorities: Legitimizing Racial Oppression by Creating Moral Panics”; “The Contribution of Institutionalized Racism to Minority Crime”; and “Minority Group Threat, Crime, and the Mobilization of Law in France.”

The objectives inRace, Gender, and Class in Criminologyare explicitly to investigate the various intersections of class, race, and gender, to explore how these relations or configurations are more than the sum of their parts, and to examine how these intersections may structure criminal opportunities and shape criminal behavior. Each of the contributions whether addressing theory or practice, do so from the mutual vantage points of class, race, and gender. At the same time, a variety of theoretical perspectives from critical criminology, including neo-Marxism, feminism, left realism, postmodernism, peacemaking, and newsmaking, are heard from. Each of these contributions tries to capture the way that its particular theoretical framework has or could look at these intersections in relationship to the production of crime and crime control. Thus, all kinds of relationships are discussed including: “structured choices,” “life histories,” “unequal power,” “layers of domination,” “psychoanalytic semiotics,” “mass mediations,” “fluid social constructs,” “axes of differences,” and “interpenetrating effects.” Contributors to this volume come from the USA, Canada, the United Kingdom, and Australia. A sampling of the chapter titles from the “applications” section ofRace, Gender, and Class in Criminology reads: “White Collar Crime and the Class-Race-Gender Construct”; “Aboriginal Australia: Current Criminological Themes”; “Controlling Homeless Mothers: The Surveillance of Women in a Homeless Shelter”; and “Adolescence and the Socialization of Gendered Fear.”

Ethnographic Studies:

Ethnographic studies in crime and criminal justice, particularly those that examine the urban underclass and incorporate community ecology approaches to group related behavior and social control, are concerned with documenting the connections between and among the institutional orders of class, race, and gender and the community-level effects of economic, political, and social deprivation. Ethnographic data is typically gathered at the “grass roots” or street level; it is usually up close and personal. Ordinarily, ethnographic studies are based on in-depth interviews of relatively small samples of representative persons, ranging from the low to high double digits. These ethnographic studies want to “get inside the heads” of perpetrators, victims, police, and so on and so forth. These studies want to capture the experiences of the intersections of class, race, and gender that go beyond statistics and into the realms of the familiar and biographical. As the authors of “Voices from the Barrio: Chicano/a Gangs, Families, and Communities,” convey: “Listening to the multiple voices of community members allows for a multifaceted understanding of the complexities and contradictions of gang life, both for the youths and for the larger community” (Zatz and Portillos 2000).

For the purposes of this presentation, I refer to two noteworthy studies in the sociology of crime that have captured the various nuances in the interactions between class, race, and gender, and the ways in which these influence or socialize each other. The first is Esther Madriz’ examination of women’s fear of crime, and the second is Mark Totten’s investigation into adolescent girlfriend abuse. In both of these ethnographies, the authors are able to present the qualitative differences in the life experiences of men and women, boys and girls, majorities and minorities, in relation to socio-economic status, and to crime and crime control. By taking class, race, and gender into their accounts, both studies demonstrate that there is no particular “class” experience, or “race” experience, or “gender” experience, but rather a repertoire of class, race, and gender experiences that have emerged in the context of social groupings or various combinations of two or more of these inseparable ingredients in the formation of personal and social identity.

In Nothing Bad Happens to Good Girls, Madriz (1997) explored the relationship of the fear of crime among young and old, African American, Latina, and white upper, middle, and working class women, living in the Big Apple. In the process, she was able to demonstrate how fear of crime perpetuates gender inequalities and contributes to the differential social control of women by class and race/ethnicity. In Guys, Gangs, and Girlfriend Abuse, Totten (2000) explored the relations between early childhood abuse, ideologies of family and gender, and the construction of masculinity, on the one hand, with the marginal male socialization experiences of straight, gay, white, black, and Asian teenagers, on the other hand. In this integrative study of class, race, gender, sexuality, and abuse in Toronto, Totten was able to make sense out of the patterned differences of girlfriend abuse with respect to the physical, sexual, and emotional violence meted out by boyfriends. He was also able to explain how the reproduction of violence and social control in these young people’s lives was related to or interacted with the abuse of gays and racial minorities.

What these and other ethnographies on crime and social control reveal is an appreciation for the relations of privilege and inequality that cuts across class, race, and gender oppression. They also demonstrate an appreciation for the fact that crime and crime control cannot be separated from the totality of the ordered, structural, and cultural contexts of their productivity. In other words, the inequalities and the biases in the administration of criminal justice or in social control more generally, are part and parcel of the socialization of class, race, and gender differences, as these are experienced in relationship to differential place, order, conflict, and perception.

Social Construction Studies:

Social construction studies in criminology and criminal justice are concerned with documenting and analyzing the ways that mass institutions-political, media, and cultural-help to produce and reproduce public order and social control. Borrowing from the traditions of “symbolic interactionism,” “labeling,” and “cultural studies,” these interdisciplinary inquiries explore the notions, stereotypes, and discourses on class, race, and gender, in the belief that these help to shape and influence common images of crime, criminals, crime-fighters, and criminal justice, and that these are, in turn, inseparable from images associated with both crime and crime control policies. Data bases have consisted of case studies on the presentation and portrayals of various “crime problems” vis-’Þ¬ê’äΠ-vis institutions of mass communication. Typically, but not always, studies in social construction have involved analyses of the representations of class, race, and/or gender.

Certainly, the most prolific contributor to the literature on “social problems” and the criminalization of deviance, has been Philip Jenkins. His books on the topic include, Intimate Enemies: Moral Panics in Contemporary Britain(1992); Using Murder: The Social Construction of Serial Homicide (1994); Pedophiles and Priests (1996); and Molesters: The Cycle of Sex Offender Panics (1998). Other noteworthy books are Helen Benedict’s Virgin or Vamp: How the Press Covers Sex Crimes (1992); Joel Best’s Random Violence: How We Talk about New Crimes and New Victims (1999); and Drew Humphries’ Crack Mothers: Pregnancy, Drugs, and the Media (1999). Three significant anthologies include: Gregg Barak’s Media, Process, and the Social Construction of Crime: Studies in Newsmaking Criminology (1994); Coramae Richey Mann and Marjorie Zatz’Images of Color, Images of Crime (1998); and Gary Potter and Victor Kappeler’s Constructing Crime: Perspectives on Making News and Social Problems (1998).

Finally, reference is made to another edited collection of social construction, my Representing O.J.: Murder, Criminal Justice, and Mass Culture (1996/1999). As editor and contributor, I deliberately set out to use the Simpson case to examine the relationship between mass-mediated representations of class, race, and gender and the administration of criminal law in the United States. For the purposes of this presentation, allow me to reconstruct (and deconstruct) the class, race, and gender relations in the situation of the national preoccupation with the trial of O.J. Simpson in 1995.

One of the most celebrated courtroom dramas of all time was the nine months long televised trial of O.J. for the cold-blooded murder of his ex-wife and her male friend. For more than 18 solid months, the Simpson case was both a media circus and a public obsession, not to mention a small cottage industry of consumer goods, legal pundits, and television specials-the latter still going on at the time of this writing. I am referring to the making of the “mini-series” TV movie of O.J.’s life and trials, scheduled for network broadcasting some time next year. One can certainly psychoanalyze that the interest, appeal, attraction, disgust, or whatever, with this case had much to do with its converging issues of class, race, and gender.

One can also safely say that the O.J. trial, both inside and outside the courtroom, represented the civics lesson of the 1990s, as it socially constructed and reconstructed, over and over, the general workings of the American systems of law enforcement and criminal justice. More particularly, O.J. became a “crash course” for the masses in constitutional and criminal law, and in articulating the rights of the individual versus the rights of the state. Beyond the social realities and legal realisms of whether or not the criminal justice system was “fixed” or “broken,” were the historical experiences and perceptions that whole groups of people, based on the complexities of their class, race, and gender backgrounds, brought to their evaluations of the systems of law and justice in the United States. These real (and imagined) differences in experience of the legal systems undoubtedly shape and influence people’s views of the administration of justice. The evidence is clear that our social experiences based on class, race, and gender were more important than the actual facts of the case.

In other words, for the most part, people’s views of the criminal justice system and of Simpson’s guilt or innocence, remained the same from beginning to end. In short, beliefs and attitudes were consistent before, during, and after the trial. Some commentators have claimed that the case was an exercise in the reification of whatever people believed in the first place. Other commentators claimed that the Simpson case represented a Rorschach test of sorts. Thus, people could make anything they liked out of it. As both an analyst and a radio commentator during the criminal trial, I would say that the first of these two claims is much closer to the truth. After all, in reality there were many more “spinners of” than there were “spins on” the O.J. phenomenon. For me, however, the interesting question has, less to do, with the fact that people’s views of criminal justice and Simpson remained fairly constant throughout the debacle, and more to do, with the ways in which class, race, and gender shaped those views.

Take the question of guilty or innocent. Generally, persons from higher socio-economic groups thought that O.J. did the murders, and it appears that race and gender made no difference. Among blacks, 70 percent thought O.J. was innocent; more African American males than females thought he was guilty. Among whites, 70 percent thought that Simpson was guilty with slightly more affirmative women than men. How did the jury compare to the public at large? The jury officially voted 12-0, not guilty, on the second round of “polling” themselves. On the first round it was different as one Hispanic and eight black women and one black man had voted not guilty, and the two white women had voted guilty. So the breakdowns of the first jury reactions appear similar to those of the general public.

As meaningful as some of these differences appear, such black and white distinctions were incomplete and misleading to the extent that they failed to poll the reactions of Asians, Hispanics, and other societal groupings. More importantly, these polls in black and white, unlike the more complex and sophisticated polling of the body politic or electorate, failed to breakdown these interpretations by combining age, occupation, class, gender, sexual orientation, religion, and so on and so forth. Such data would have helped to shed light on the background similarities and differences, for instance, between the 30% of the blacks who agreed with 70% of the whites that O.J. was guilty, and conversely, with the 30% of whites who agreed with 70% of the blacks that he was not guilty. In future public discussions of crime and punishment, for example, expanded data of other ethnic and racial groups in relation to their socio-economic and gender positions, would help the body politic move beyond simple black and white distinctions and closer to the more complex relations of class, race, and gender.

What was particularly interesting to observe during the O.J. saga were the mass-mediated reconstructions to “normalize” this case within the context of the everyday practices of criminal justice in America. In other words, the Simpson case was an aberration in the administration of criminal justice as it departed from the more traditional images and stereotypes of criminal defendants, trial attorneys, expert witnesses, and juries of one’s peers. For example, criminal prosecutors and criminal defense attorneys are much more often than not white and male, the bailiffs are usually men and more often than not of color, court reporters are invariably women, and juries, as infrequent as they are, are rarely constituted by one’s peers. Typically, juries are from higher socio-economic classes than criminal defendants. Ordinarily, both the behavior of the police and the credibility of expert witnesses, are beyond reproach. That is, they are generally treated with a decorum of deference and respect.

In the circumstances of defendant O.J., the status quo was ripped apart. After all, Simpson was a wealthy African American male accused of murdering his formerly dependent–psychologically and economically–white wife and her white working class male friend in a “sexual triangle” of sorts. Of course, Simpson was also a media celebrity from television and films, and a former all Pro running-back for the Buffalo Bills, who was able to retain a million-dollar “dream team” of well-known criminal attorneys, eventually led by the indefatigable Johnnie Cochran. In fact, unlike 99.9% and higher of criminal defendants, O.J. had “deeper pockets” than the prosecution did. As for the prosecuting team, they were led by the unusual combination of a white woman and an African American man. As for the jury, they were composed of 11 women and one man; nine African Americans, one Hispanic, and two whites; all members of the working classes. Finally, presiding over this trial was an Asian rather than an Anglo or Euro American judge.

These and other differences from the normal relations of class, race, and gender that usually surround a murder trial, accounted for the differential applications of the law, or for the special privileges, that O.J. received during his jailed incarceration period, prior to and pending the outcome of his trial. For example, even before the trial began, Simpson reached an unheard of deal in the annals of American criminal justice history. He was able, through his attorneys, to successfully negotiate a deal with the prosecution that should he be convicted of the double murder, that the state would not execute him. Generally, if such deals are reached, the accused has to, in exchange, plead guilty to some crime or another, saving the state the expenses of a costly trial and eliminating the possibility of a non-conviction. O.J. traded nothing except his incredible popularity.

Similarly, because of the high powered nature of the defense team, Simpson’s attorneys were able to effectively put the motives and competencies of the Los Angeles Police Department and District Attorney’s Office on trial. In the process, they raised what appears to have been the “reasonable doubt” in the minds of the jurors; the key to his acquittal in the criminal trial. In sum, the contradictions in the management of criminal justice between the treatment of O.J. and that of the typical person accused of murder, black or white or whatever, were informed by a novel combination of class, race, and gender relations of crime control.


The four ways of seeing difference in the social relations of class, race, and gender that I have described as characteristic of criminology and criminal justice are, of course, ideal types or social constructs themselves. As a proponent of integrative criminology and as one who has advocated for integrating the different criminologies, there are no hard and fast boundaries between the four approaches (Barak 1998). As most social and behavioral scientists of crime and justice would agree, using a variety of methods to validate any phenomenon is generally better than using only one method. By way of discussion and closure, let me try to share some of the thinking behind Barak, Flavin, and Leighton’s forthcoming book, Class, Race, Gender, and Crime: Social Realities of Justice in America (2001).

To begin with, we regard the criminal justice system as a culturally powerful, label-conferring institution that has evolved in relation to the changing definitions of “crime” over time. Additionally, we view the defining of “crime” and “criminals” as a product of moral agents, social movements, political interests, and media dissemination. In other words, what becomes a “crime” and who becomes a “criminal” are politically, economically, and socially constructed phenomena, reproduced daily through various discussions in the streets, the home, the school, the church, the government, the courts, the airwaves, and the other cultural bodies. Finally, we approach “crime” and “justice” from historically developing standpoints of class, race, and gender as these undergo social construction.

When we specifically examine class, race, and gender in relationship to law, order, and crime control, on the one hand, we appreciate the unique histories of these social groupings both in isolation and in combination, and on the other hand, we appreciate the way these different social attributes and cultural constructions represent interrelated axes of privilege and inequality. At any moment, class, race, and gender may “feel more salient or meaningful in a given person’s life, but they are overlapping and cumulative in their effects on people’s experience (Andersen and Hill Collins 1998: 3). As we historically demonstrate in our book, in terms of the social realities of justice in America, the experiences of diverse groups of people in society have contributed to the shaping of the types of criminals and victims that we have had. Like Andersen and Hill Collins (1998: 4) in their discussion of what they refer to as a “matrix of domination,” we too conceive that class, race, and gender represent “multiple, interlocking levels of domination that stem from the societal configurations of these structural relationships. These patterned actions, in turn, affect [ing] individual consciousness, group interaction, and individual and group access to institutional power and privileges.”

For example, Roberts (1993) in her examination of the intersections of crime, race, and reproduction, discusses the convergence between the racial construction of crime and the use of reproduction as an instrument of punishment. She has argued that the “technology of power” that links crime, race, and reproduction epitomizes how racism and patriarchy function as mutually reinforcing systems of domination that help to determine “who the criminals are, what constitutes a crime, and which crimes society treats most seriously” (Roberts 1993: 1945). More specifically, in terms of abortion, birth control, and social control, Roberts discusses how this domination is meted out through the control of black women’s bodies that discourage procreation, subordinate groups, and regulate fertility. As part of our integrative analysis of class, race, and gender, we similarly attempt to explore how each of these hierarchies helps to sustain the others, and how these reinforce the types of crimes and justice we have in society.

More generally, we bring at least four related assumptions to our study in the social relations of class, race, gender, and crime control:

First, that each of these categories of social difference share similarities and dissimilarities of justice, especially as these relate to power resources and to the allocation and distribution of rewards and punishments in society.

Second, that the systems of privilege and inequality derived from the social statuses of class, race, and gender, share distinct as well as integrative, or overlapping and accumulating, affects on the type of crime control that various groups of people receive.

Third, that there are connections and linkages between these systems of difference, inequality, and privilege as each, separately and together, helps reproduce the social divisions of hierarchy and stratification that dynamically affect people’s life experiences, inside and outside, the criminal justice system.

Fourth, that systems of crime control socially construct selectively enforced and differentially applied norms to social groups, according to relationships of power, status, and authority.

Historically, we know that the legal differences favoring corporations over individuals, workers, and consumers, or the wealthy over the middle, working, and poor classes, have remained fairly constant over time despite efforts to regulate and control monopolies of wealth or to assist poverty’s destitute. During the 20th century, we recognize that, on the one hand, the more blatant forms of discrimination based on alleged differences of race, ethnicity, and gender, have been significantly reduced in the United States. On the other hand, we also recognize that although the legalized and institutionalized forms of bias have been reformed and abolished by law, that, in practice, differential treatment based on race and gender still persists. Hence, in terms of the operations of crime control, poor persons still have fewer resources or less power working for them in negotiating outcomes within and without the criminal justice system than the affluent or middle classes. And, when poor persons are of color or are female too, they usually hold even less power, and if they are all three-poor, of color, and female-then they typically possess lesser power still.

Our study is not an ethnographic study of victims or victimizers, but rather its an analytical investigation into the institutionalized practices and outcomes of crime control. Nevertheless, we share the insights and the desires of Madriz and Totten to unravel the complexities of class, race, and gender as these interact with the cultural production of crime, justice, and inequality. We also share their critical view that crime, justice, and crime control cannot be separated from the totality of the ordered, structural, and cultural contexts of their productivity. Each of our cultural approaches holds that the inequalities in control and justice are part and parcel of the social constructions of class, race, and gender differences, as these are experienced in relationship to place, order, conflict, and perception.

Moreover, social perceptions of what constitutes unacceptable social injuries and acceptable social controls are shaped by the underlying elements of social organization, or by the production and distribution of economic, political, and cultural services (Michalowski 1985). Following Antonio Gramsci (1971), we are not talking about conspiracies of elites and decision-makers here, but rather, we are referring to agreed upon definitions of harms and injuries, pains and sufferings, and crimes and punishments that reflect capitalist political-economic relations and interests. Hence, in the final judgment serious crime defined from above or below, from the suite to the street, and from the official reports of the Federal Bureau of Investigation to the cultural media, all become statistically mediated and socially constructed phenomena.

In culturally generated numbers, narratives, and pictures alike, a distorted view and limited perception of harmful behavior emerges. Crimes and criminals are restricted primarily to the tabulations and representations of conventional criminal code violations, such as homicide, rape, burglary, robbery, theft, and less often, assault. From a comparative perspective, whatare traditionally omitted from these images and narratives of justice are two things: from below, the grossly under-reported and/or hidden crimes, such as the trafficking and possession in stolen merchandise, illicit sex, gambling, loan sharking, internal pilferage, and less so, the smuggling of weapons; and from above, the severely ignored and invisible crimes, such as the frauds and embezzlements of white-collar and professional criminals as well as numerous corporate offenses against the environment, workplace, and consumer.

Just as these materially and culturally produced images of crime and criminals leave impressions that reinforce one-dimensional notions that criminality and harmful behavior are exclusively the responsibility of the poor and marginal members of society, the material and cultural images of crime control and the administration of justice leave impressions that reproduce limiting social realities of social control and crime prevention. As mass consumers, for example, we all share a virtual reality of mediated facsimiles of lawbreakers and crime-fighters. Common narratives or stories of crime and criminal justice appear and reappear so often in the news, in films, in television, in literature, and in popular discourse, that most Americans imagine similar renderings of crime, criminals, law enforcement, adjudication, punishment, and so forth.

Is it no wonder that when people try to picture the typical American crime, the common images that emerge are of mostly young victimizers and victims of color? In repetitive news stories, African American and Hispanic male youths in particular, have been encountered in pools of blood lying dead, victims of so-called “random” or “senseless” violence. There are also the numerous police action reenactments that can be viewed regularly on such television programs as Top Cops or America’s Most Wanted, that similarly recycle images of these young men as dangerous drug dealers whose dwellings must be invaded during the early hours of dawn by “storm trooper” police and other law enforcement personnel, in order to secure the “war on crime.” In like fashion, the images of crime control that are constructed throughout the criminal justice system as we move from law enforcement to adjudication and from sentencing to incarceration, again serve to reinforce limited and fairly biased portrayals of the realities of criminal justice in America.

When we imagine a criminal courtroom, for example, images come to mind from relatively long and involved trials, exposed either in feature length films, or from Court Television’s gavel-to-gavel coverage of such celebrated trials as the murder conviction of Sandy Murphy and Rich Tablish for killing multi-millionaire and former Las Vegas casino owner, Ted Binion, in September 1998. The actual trial of these two “sympathetic” murderers did not convene until February, 2000, and ended in May with both of them receiving the minimum sentences for murder that the state of Nevada permits. At the same time, the public is led to believe, based on very succinct and curt shots of highly charged courtroom scenes from various television series like The Practice and Law and Order, that competent attorneys for each side are present and engaged in vigorous battle, always doing their best to secure justice for all. In these fictional and non-fictional dramatizations, the images that do not come to mind are the ones where the rights of defendants have been all but eliminated. Reference is made to the overwhelming majority of criminal cases, 90 percent, that are plea-bargained everyday in courthouses throughout the nation. These negotiated deals in lieu of trials usually take less than a few minutes for judges and courts to process and uphold.

Moving from adjudication to punishment, popular images come to mind of dangerously violent offenders who need to be locked up indefinitely. Such pictures make unimaginable the possibility of ever re-aligning the offender, the victim, and the community. As part of the politics of American punishment and the political economy of incarceration, the languages and images of retribution serve to negate efforts in the “rehabilitation” of people while they reproduce the United States’ 100 billion dollar a year criminal justice-industrial complex (Shelden 1999). Overall, the representations of offenders depict feuding convicts divided into racial and religious cliques doing “scared time,” with young and inexperienced inmates guarding their derrieres from sexual predators, rather than images of residents engaged in school or a vocation, and of former offenders “fitting back” into society.

Lastly, the award winning HBO dramatic series of life in a maximum security prison, OZ, portrays a based-on-facts fictional account of the complexity of one of those “hell on earth” holes or archipelagos. On the one hand, such representation ignores the social realities of some 1500 other state and federal prisons of less severity and pain. On the other hand, OZ does not do justice to the growing apartheid like conditions of crime and punishment that disproportionately affects black and brown Americans. At the same time, commercially successful prison films, like Lock Up (1989) or The Shawshank Redemption (1994), tend to personify a plurality of ethnic and cultural diversity in prison, as they tell stories of mostly white inmate protagonists doing conflict with mostly white correctional antagonists, against a background of “out of control” systems of criminal justice. These narratives are not only dated, but they represent “white-washed” versions of life behind bars in the United States.

In the final analysis, what we try to show in our book is how the social relations of class, race, gender, and crime control as well as the ways of seeing difference, are both related to the inequalities of crime, social justice, and culture production. After reviewing in the first portion of the book, the histories of “class justice,” “race justice,” and “gender justice” in the U.S. states, we then examine class, race, and gender in the administration of criminal justice: first, in terms of the relative uniqueness and isolation of class, race, and gender; and second, in terms of the interactions between class, race, and gender. We then go on to discuss alternative discourses on crime and justice as well as to propose policies that reflect upon crime, marginality, and justice in relationship to the prevention and the reduction of harms and crimes in American society.


Andersen, Margaret L. and Patricia Hill Collins. 1998. Race, Class and Gender: An Anthology, 3rd edition. Belmont, CA: Wadsworth Publishing.

Barak, Gregg. 1994. (ed.). Media, Process, and the Social Construction of Crime: Studies in Newsmaking Criminology. New York: Garland Publishing.

Barak, Gregg. 1996/1999. (ed.). Representing O.J.: Murder, Criminal Justice, and Mass Culture. Albany, NY: Harrow and Heston. Republished with a new preface and new title, Media, Criminal Justice, and Mass Culture. New York: Criminal Justice Press.

Barak, Gregg. 1998. Integrating Criminologies. Boston: Allyn and Bacon.

Barak, Gregg, Jeanne Flavin, and Paul Leighton. 2001. Class, Race, Gender, and Crime: Social Realities of Justice in America. Los Angeles: Roxbury Publishing.

Benedict, Helen. 1992. Virgin or Vamp: How the Press Covers Sex Crimes New York: Oxford University Press.

Best, Joel. 1999. Random Violence: How We Talk About New Crimes and New Victims . Berkeley: University of California Press.

Dunaway, R. Gregory, Francis T. Cullen, Velmer S. Burton, and T. David Evans. 2000.“The Myth of Social Class and Crime Revisited: An Examination of Class and Adult Criminality.” Criminology 38 (2): 589-632.

Gramsci, Antonio. 1971. Selections from the Prison Notebooks. New York: International.

Hawkins, Darnell F. 1995. (ed.). Ethnicity, Race, and Crime: Perspectives Across Time and Place. Albany, NY: SUNY Press.

Humphries, Drew. 1999. Crack Mothers: Pregnancy, Drugs, and the Media. Columbus: Ohio University Press.

Jenkins, Philip. 1992. Intimate Enemies: Moral Panics in Contemporary Great Britain.. Hawthrone, NY: Aldine de Gruyter.

Jenkins, Philip. 1994. Using Murder: The Social Construction of Serial Homicide. Hawthrone, NY: Aldine de Gruyter.

Jenkins, Philip. 1996. Pedophiles and Priests. New York: Oxford University Press.

Jenkins, Philip. 1998.

Molesters: The Cycle of Sex Offender Panics New Haven: Yale University Press.

Madriz, Esther. 1997. Nothing Bad Happens to Good Girls: Fear of Crime in Women’s Lives. Berkeley: University of California Press.

Mann, Coramae Richey and Marjorie S. Zatz. 1998 (eds.). Images of Color, Images of Crime: Readings. Los Angeles: Roxbury Publishing.

Michalowski, Raymond. 1985. Order, Law and Crime. New York: Random House.

Potter, Gary W. and Victor E. Kappeler. 1998 (eds.). Constructing Crime: Perspectives On Making News and Social Problems . Prospect Heights, IL: Waveland Press.

Roberts, Dorothy E. 1993. “Crime, race, and Reproduction.” Tulane Law Review 67 (6): 1945-1977.

Schwartz, Martin D., and Dragan Milovanovic. 1996. (eds.). Race, Gender, and Class in Criminology: The Intersection. New York: Garland Publishing.

Totten, Mark D. 2000. Guys, Gangs and Girlfriend Abuse. Petersborough, Ontario, Canada: Broadview Press.

Wright, Bradley, R. Entner, Avshalom Caspi, Terrie E. Moffitt, Richard A. Miech, and Phil A. Silva. 1999. “Reconsidering the Relationship between SES and Delinquency: Causation but not Correlation.” Criminology 37 (1): 175-194.

Zatz, Marjorie S., and Edwardo L. Portillos. 2000. “Voices from the Barrio: Chicano/a Gangs, Families, and Communities.” Criminology 38 (2):369-402.


Summarizing the New US Census Bureau Report on Income and Poverty: The Rich Continue to Get Richer

Summarizing the New US Census Bureau Report on Income and Poverty: The Rich Continue to Get Richer

Michael J. Lynch

The US Census Bureau released new figures on the economic health and well being of Americans on August 29th in its annual report. Below I summarize some of the important aspects of this report. To view this report:

1. Real median household income rose 1.1% in 2005 to $46,326. Real median income is an inflation adjusted measure which indicates the income amount that divides US families at their midpoint, with one half of families earning less than $46,326, and one half of families earning greater than that amount.

2. Although real median household income rose last year, the rise was not sufficient to overcome the impact of the recession that ushered in the 21st century in the US. Real median family incomes in the US in 2005 remained 0.5% lower than real median family incomes in 2001.

3. Real median income values, however, provide a misleading indicator of how widely US citizens share in recent economic gains. For example, while real median family income increased, so too did economic inequality. Economic inequality is evident in several additional indicators.

4. Median family incomes vary significantly by race. The 3 year moving average for families of different racial and ethnic backgrounds were reported as follows: Whites, Non-Hispanic ($50,784); African Americans ($30,858); Hispanics ($35,967); Asians ($61,094).

5. In 2005, the poorest 20% of families earned only 3.4 % of all household income. The top 20% of families earned 50.4% of all household incomes. Clearly, this indicates a wide disparity in the distribution of income.

6. For the top 20% of households, average incomes rose by 2 percent, and the mean annual income for this group of families is now $ 159, 583.

7. Mean income for the bottom 20% rose at a lower rate of 0.6%. The mean family income for those in the lowest 20% of household incomes rose by only $68 to $10,587.

8. The report indicated that the GINI coefficient of income inequality rose to 0.469 in 2005. The higher the GINI coefficient, the more unequal the distribution of income. The 2005 GINI is the largest inequality figure recorded by the US Census Bureau in the 40 years it has issued annual reports.

9. The small rise in income for the lowest 20% of income earning families helped reduce the proportion of the population that lives in poverty by 0.1% (to 12.6%). The 2005 poverty rate, however, was 1.3 points higher than the 2001 poverty rate (11.3%), which indicated that the poorest Americans have had much more difficulty recovering from the early 21st century recession. Today, 36.85 million Americans still live in poverty.

10. Despite the rise in median family income, the median income for both men and women declined. Men’s median income fell by 1.8% to $ 41,386, while women’s median incomes fell 1.3% to $ 31,858. On average, women still earn significantly less than men (77 cents for every dollar earned by men). It should be noted that the rise in the male/female wage level that has occurred since the mid-1980s is largely the result of men’s wages falling relative to women’s wages, and not the result of a real gain in women’s wages relative to men’s wages. Also, the discrepancy between the decline in individual wages (women/men) versus the rise in family income is the result of income generated from investments for families.

11. According to US Census Bureau documents (, poverty thresholds are as follows: for persons under age 65 ($9,827/yr); for persons over age 65 ($ 9,060/yr); for a family of 4 ($19,484) (for other family sizes, use link). It should be noted that the poverty level value set by the US Census Bureau for individuals under age 65 is slightly less ($885) than a person who earns minimum wage ($ 5.15/hour) would make working 40 hours a week for 52 weeks.

12. There are important racial differences in poverty that need to be considered. During 2005, the poverty rate for Whites declined slightly (0.1%), to 10.5%. The African American poverty rate remained constant, though they continue to be adversely affected, and the proportion of African Americans who live in poverty was 24.7%. Like the White poverty rates, the Hispanic poverty rate fell by 0.1%. However, like the African American poverty rate, the Hispanics poverty rate remains significantly higher than the White poverty rate at 21.8%.

13. Poverty rates for other groups also rose. For female headed households, the poverty rate rose by 0.6 points to 31.1%. Likewise, the poverty rate for those over 65 rose by 0.3 points to 10.1%. Poverty rates for children, however, fell by 0.2 points to 17.6%.

14. The number of Americans without health insurance increased to 46.6 million, or by 1.3 million people during 2005.

15. Important regional variations exist in reported family income patterns. The rise in incomes was above the national average in Northeastern States (2.9%) and in Westerns States (1.5%), and below the national average in Midwestern (-0.4%) and Southern (0.1%) states.