To anyone who has struggled to pay the bills, it’s no news that income inequality has been on the rise — things have been good at the top, but for many others, times are tough: Since the 1980s, the wealthiest 1% of U.S. households saw their incomes more than double, while the middle class has recently been struggling through a “lost decade” with stagnant or declining income. And for those at the bottom, the poverty rate remains stubbornly high.

Most studies on income equality use purely economic measures — everyone in the 25th to 75th percentiles is middle class, for example. However, a 2015 study from the Federal Reserve Bank of St. Louis also includes demographic dimensions, and found that middle-class declines were slightly worse when factors such as such as race, education level and occupation were taken into account. “The legacies of discrimination and other deeply rooted factors that contribute to low levels of educational attainment by blacks and Hispanics have slowed their economic and financial advances,” the researchers conclude.

A 2014 study in the Proceedings of the National Academy of Sciences (PNAS), “Economic Scarcity Alters the Perception of Race,” looks at a hidden aspect of the psychology of discrimination: How perceptions of economic conditions might shift social perceptions and exacerbate racial disparities. In their work, the researchers, Amy R. Krosch and David M. Amodio of New York University, conducted four psychological experiments. The key component was a set of pictures of faces that shifted in 10 steps from 100% “white” and 0% “black” to 0% “white” and 100% “black”; a combination of many individuals, the faces had a neutral expression, matched facial hair and structure, with hair and ears cropped out. The spectrum of images gradually faded from faces with Caucasian features to ones with more African-American features, and vice versa.

Subjects in the first three experiments were recruited from Amazon’s Mechanical Turk service; the groups were a mixture of white, Asian, Latino, Native American and non-black multiracial. In the fourth experiment, subjects were recruited from a city park. The experiments often tested how “priming” subjects — helping shift their evaluative criteria and judgments — could then alter the way they interpreted subsequent information. Researchers have long confirmed that most people demonstrate implicit bias of many kinds, and the experiments in the PNAS study tested whether a kind of economic “priming” might bring out embedded biases.

The experiments and their findings were:

• In the first experiment, 70 subjects were asked to respond to six “zero-sum” questions/statements about race and scarcity, such as “White people lose out when black people make economic gains.” They then were shown a series of randomized images on the race spectrum and were asked to categorize the individuals as either “white” or “black.” Overall, subjects identified a face as being black when it included 47% or more African-American characteristics/content, but those who agreed with more of the zero-sum statements tended to identify faces as black earlier.
• In the second experiment, 63 subjects were primed with words for 20 milliseconds prior to being shown each face; the words were related to economic scarcity, or were neutral or unrelated. The researchers then performed the same task of assigning a race to photos. When primed with words relating to scarcity, subjects identified the face as black when it included 40% or more African-American facial content. Those who had been exposed to scarcity-related words identified the face as black earlier than those who had been exposed to neutral or non-scarcity related negative words.
• In the third experiment, 62 subjects were randomly assigned to distribute either $10 out of $100 (the “scarcity” condition) or $10 out of $10 (control). While ostensibly waiting for a partner, they were shown pairs of faces and asked to choose which one was “most black.” This in fact was the actual experiment, and the money limit served to prime them for rating the images. Subjects in the “scarcity” condition described blurred images of faces as more black than those in the control group.
• In the fourth experiment, 59 white individuals were shown two faces rated as more “black” (as generated in the third experiment) and asked how they would they divide $15 between two of them. Overall, subjects gave less money to the darker, more stereotypically African-American, face than to the lighter, less stereotypically African-American face.

“Together, our results provide strong converging evidence for the role of perceptual bias as a mechanism through which economic scarcity enhances discrimination and contributes to racial disparities,” the researchers conclude. “Our findings also dovetail with recent evidence that scarcity is cognitively taxing and thus may undermine an individual’s ability to regulate unintended prejudices. Despite egalitarian beliefs, most Americans possess implicit negative associations with minority groups that are kept in check through effortful cognitive control.”

Related research: A 2012 book exploring General Social Survey data finds that segregationist sentiment continues to exist on a “nontrivial” level in the United States. Also of interest is a 2011 study from University of California, Irvine, “Cause of Death Affects Racial Classification on Death Certificates.” Based on a nationally representative sample of death certificates, the researchers found that a person who died of chronic liver disease was 2.4 times more likely to be classified as American Indian than white, while homicide victims were 4.4 times more likely to be classified as black.

Keywords: race, racism, psychology, economic scarcity, poverty, black, white, biracial, social psychology, African-American, cognition, prejudice