Anatomy of Income Inequality in US (2016)

 The sources indicate that understanding US income inequality between 1979 and 2013 requires moving beyond single statistics toward a granular dissection of demographic subgroups. The methodology focuses on decomposing traditional inequality measures to reveal the underlying forces of age, gender, race, and education.

Limitations of Traditional Measures

Standard inequality measures, like the Gini coefficient, are criticized for being "single numbers" that gloss over the nuances of societal change. A primary methodological concern is that the standard line of perfect equality (the 45-degree line) assumes everyone's income should be equal at any given moment. However, the sources argue this is misleading because life-cycle differences—where income naturally evolves with a person’s age and experience—ensure that the Gini coefficient will never be zero, even under perfect equality of opportunity. For instance, it is calculated that if every worker followed a standard life-cycle trajectory in 2013, the Gini coefficient would still be approximately 0.13.

Evolutionary Methodologies: Paglin vs. Pyatt

To account for these life-cycle effects, the sources discuss two primary decomposition methods:

• Paglin’s Gini (P-Gini): Proposed in 1975, this method replaces the standard line of equality with a "P-reference line". This line represents equal lifetime incomes without the constraint of a flat age-income profile. However, this method was criticized for being sensitive to the width of age cohorts and for ignoring the "overlap" term, which occurs when some members of an older cohort are poorer than members of a younger cohort.
• Pyatt Decomposition Method: This is the primary method used in the current study because it successfully incorporates the overlap term. Pyatt’s approach views the Gini coefficient as a statistical game where individuals compare their income to others; the "expected gain" from switching to a higher income represents the level of inequality. This method decomposes the Gini into three effects:
  1. Changes in inequality within age groups (weighted average of internal cohort inequality).
  2. Changes in inequality between age groups (reflecting the age-income profile).
  3. Demographic dynamics (changes in the population share/size of each age group).

Data and Regression Strategy

The researcher utilizes the Luxembourg Income Study (LIS), a harmonized micro-level dataset that allows for cross-country comparisons. Methodologically, the study prioritizes personal-level data over household data. Household data is considered "shaky" for long-term inequality studies because households form and dissolve over time, making it difficult to track specific demographic groups like gender or race.

To further isolate the drivers of inequality, the sources employ regression analysis. This allows the researcher to examine how factors such as marriage rates, variance in the number of children, education levels, and immigration status specifically impact within-cohort inequality while controlling for age, gender, and racial variables.

Analogy for Understanding Decomposition: Analyzing income inequality with a single Gini coefficient is like looking at the average temperature of a whole country; it tells you the general climate but hides the fact that one coast is freezing while the other is in a heatwave. Decomposition acts like a weather map, allowing researchers to see the "within-group" storms and "between-group" shifts that actually explain the changing environment.

Analysis of US income inequality from 1979 to 2013 reveals several stylized facts that challenge the utility of a single "Gini coefficient" for the entire population. The sources highlight that while overall inequality has grown, the underlying drivers are specific to age, gender, and labor market shifts.

1. High and Stable Within-Cohort Inequality

A primary observation is that the share of overall income inequality due to within-age-group differences in the U.S. has remained consistently high and steady.

• Between 1979 and 2013, within-cohort inequality accounted for 71% to 74% of total U.S. inequality.
• This rate is significantly higher than in many other advanced nations; for example, in Denmark, this share is only between 50% and 60%.
• While other countries saw this share fluctuate or soar (such as the UK), the U.S. profile has remained remarkably constant despite a rising overall Gini coefficient.

2. The Unbalanced Rise Across Age Groups

The sources observe an uneven growth of inequality across the life cycle, with the most drastic increases occurring at the two ends of the age spectrum.

• The Elderly (65–79): Inequality within the 70–74 age group rose by 14.5%. Key factors include increased life expectancy for black Americans (specifically black men), which adds lower-income individuals to the older population mix, and a widening gender gap among the elderly. Changes in social security, such as the shift from defined-benefit to defined-contribution plans, are also suspected drivers.
• The Youth (20–29): Inequality among 20–24 year-olds grew by 9%. This is attributed to decreased economic mobility, higher rates of single-parenthood, and shifts in education policy from need-based to merit-based aid, which tends to favor youth from affluent families.
• Middle-Aged Workers: In contrast, inequality within middle-aged cohorts (35–39) remained virtually unchanged or grew very little during this 34-year period.

3. The Gender Divergence

One of the most striking observations is that inequality trends for men and women moved in opposite directions.

• Men: Inequality among American men rose sharply, with their Gini coefficient increasing from 0.36 in 1979 to 0.45 in 2013.
• Women: Paradoxically, inequality among women declined from 0.47 to 0.45 over the same period.
• Married Women: This group was the primary driver of the decline in female inequality, likely due to increased labor force participation and a reduction in the number of children, which narrowed the "family gap" in wages.

4. Labor Market and Policy Observations

The sources identify specific structural changes that explain why inequality rose for men but not for women:

• Minimum Wage: While the real value of the minimum wage fell by 22%, the percentage of women earning the minimum wage plummeted from 20.2% in 1979 to 5.4% in 2013. Because men’s share of minimum wage work remained stable, the falling real value of the wage floor increased male inequality while affecting fewer women.
• Unionization: Deunionization was much more severe for men (whose membership rates dropped from 25% to 12%) than for women (who saw a smaller decline from 14.5% to 11%).
• Technology and Automation: Computerization has replaced "routine" jobs typically held by men, while "abstract" managerial roles—where the male labor supply is less elastic—saw massive wage increases, further polarising the male income distribution.

Analogy for Understanding Stylized Facts: Viewing US inequality through a single Gini coefficient is like looking at a mountain range from a distance; you see it is high, but you miss the fact that some peaks are rising while specific valleys are being filled in. Decomposition allows researchers to see that the "male peak" is getting steeper while the "female valley" is actually leveling out.

The sources explain that the rise in US income inequality between 1979 and 2013 cannot be attributed to a single cause; rather, it is driven by a complex interplay of demographic shifts, labor market institutions, and structural technological changes. These factors often impacted different subgroups in diametrically opposed ways.

1. Demographic and Life-Cycle Factors

The sources point to the aging of the US population as a foundational explanatory factor.

• The Permanent Income Hypothesis (PIH): Inequality within a specific age cohort naturally increases as that cohort ages. Because the large baby boomer generation moved into their 50s and 60s during this period—the years where within-cohort inequality is statistically highest—their demographic weight pushed the overall Gini coefficient upward.
• Life Expectancy: Increased life expectancy for black Americans, particularly black men, contributed to higher inequality among the elderly. As more individuals from the lower end of the income distribution survived into older age cohorts, the income variance within those older groups expanded.

2. Labor Market Institutions

The decline of institutional protections significantly increased inequality for men while having a more nuanced effect on women.

• Minimum Wage: While the nominal minimum wage rose, its real value fell by approximately 22% between 1979 and 2013. This exacerbated male inequality because the share of men earning at or below the minimum wage remained stable. Conversely, the share of women earning the minimum wage plummeted from 20.2% to 5.4%, meaning fewer women were affected by the falling real wage floor.
• Deunionization: The collapse of union membership was much more severe for men than for women. Sources argue that deunionization accounts for roughly 20% of the increase in male wage inequality, whereas it had a negligible or non-existent effect on female wage inequality.

3. Structural and Technological Shifts

The sources emphasize job polarization driven by computerization and automation as a central driver of the "male vs. female" inequality divergence.

• Routine vs. Abstract Jobs: Computers have largely replaced "routine" jobs (clerical or repetitive production) while complementing "abstract" jobs (managerial and professional).
• Supply Elasticity: The supply of men for abstract, high-paying jobs is relatively inelastic, meaning increased demand for these skills resulted in massive wage spikes for those at the top. In contrast, the supply of women was more elastic as many entered the labor force for the first time, preventing similar wage polarization within female cohorts.
• Automation Exposure: Women reduced their employment in manual routine jobs at a faster rate than men, which shielded them from some of the wage-suppressing effects of automation that hit the male-dominated production sectors.

4. Gender and Family Dynamics

Changes in the "traditional" family structure played a critical role in shaping the distribution of income.

• Labor Force Participation: Female labor force participation rose by 13.3%, while male participation fell by 9.7%. Statistically, as female participation moves past 50% and approaches 100%, within-group inequality naturally begins to decline, which partially explains the falling Gini among women.
• The "Family Gap": A decline in the number of children per woman reduced the "wage penalty" often associated with motherhood. Additionally, the labor supply decisions of married women became less sensitive to their husbands' wages, decoupling their income from the highly unequal male wage distribution.

5. Factors for the Youth and Elderly

• Youth Inequality: This is attributed to decreased economic mobility and a shift in educational policy where merit-based aid (which favors affluent families) has increasingly replaced need-based aid. Higher tuition rates have further stratified outcomes for young workers.
• Elderly Inequality: Beyond life expectancy, the sources cite changes in social security policies, such as the transition from defined-benefit to defined-contribution pension plans, which shifted financial risk onto the individual and widened income disparities in retirement.

Analogy for Explanatory Factors: Trying to explain US inequality with one factor is like trying to explain why a forest is changing by only looking at the rain. You must also account for the age of the trees (demographics), the thinning of the soil (deunionization), and the introduction of a new species (automation) to understand why some parts of the forest are flourishing while others are withering.

The sources identify several key social policy changes that have significantly influenced the trajectory of US income inequality between 1979 and 2013, often acting as a "suspect" in the rising disparities within specific age and gender groups. These policies span labor market regulations, education funding, welfare restructuring, and retirement security.

1. Labor Market and Tax Policies

One of the most direct policy impacts cited is the decline in the real value of the minimum wage, which fell by approximately 22% during this period. While the nominal wage increased, the failure to index it to inflation exacerbated inequality among men, though its effect on women was mitigated because the share of women earning the minimum wage plummeted from 20.2% to 5.4%. Additionally, the Tax Reform Act of 1986, which reduced the top marginal tax rate from 50% to 28%, is noted as a primary motivator for women from high-income households to join the labor market, contributing to the "great gender convergence".

2. Welfare and Family Support

Significant shifts in the social safety net occurred in the mid-1990s, altering labor participation incentives.

• Welfare Reform: The transformation of the old Aid for Families with Dependent Children (AFDC) into the more conditional and temporary Temporary Assistance to Needy Families (TANF) in 1996 disproportionately affected women of color and incentivized labor market entry.
• EITC Expansion: The expansion of the Earned Income Tax Credit (EITC) is credited with increasing labor force participation among single mothers and older workers by providing more flexible work incentives.

3. Education and Youth Policy

The sources suggest that educational policy has shifted from being an equalizing force to one that may exacerbate inequality.

• Shift in Aid Type: There has been a significant shift in funding from need-based aid to merit-based aid.
• Stratification: Approximately 57% of merit-based aid currently goes to children from high- or high-middle-income families. Because affluent youth often do not need the same financial stimulus to pursue a degree, this policy shift can worsen the inequality of opportunity.
• Tuition and Incarceration: Rising tuition rates and high incarceration rates for black parents—which increased six-fold for black children under 18 since 1980—are also cited as social policy-related factors that have hindered economic mobility for the youth.

4. Retirement and Social Security

Policy changes regarding the elderly have shifted financial risks from the state and employers onto the individual.

• Pension Structure: There has been a dominant move from defined-benefit plans (where the employer bears the risk) to defined-contribution plans (where the individual bears the risk).
• Social Security Adjustments: The increase in the full retirement age and the reduction in participation in retirement benefits are considered drivers for the unbalanced rise of inequality among the American elderly.
• The Support Ratio: The "support ratio"—the number of workers (20–64) per retiree (65+)—is projected to fall to 2.1 by 2040. This demographic-policy challenge places a higher burden on younger workers to support the social security system, potentially driving further aggregate inequality.

Analogy for Social Policy Changes: Social policy functions like the safety netting and tension cables on a suspension bridge. Between 1979 and 2013, the sources suggest that while some cables were strengthened to help certain groups cross (like the EITC for workers), other parts of the safety netting (like the real value of the minimum wage and need-based education aid) were allowed to fray, making the crossing significantly more precarious for those at the ends of the age spectrum.