Table User Guide—Modeling Income in the Near Term (MINT) 8

Interpreting the Projected Effects Tables

For each retirement policy option/proposal, we provide six sets of tables showing the effects of the proposed changes on:

  • Social Security benefits,
  • Social Security taxes paid,
  • Household income,
  • Official poverty measure,
  • Benefits/tax ratio, and
  • Initial replacement rates.

The first four sets of tables provide results for the analysis years, 2030, 2050, and 2070, while the last two provide results for four birth cohorts: 1960–1969, 1980–1989, 2000–2009, and 2020–2029.

Each set of tables and what they show are discussed below.

Social Security Benefits

These tables show the projected changes in individual Social Security benefits for current-law beneficiaries aged 60 or older in 2030, 2050, and 2070.

The first two columns show the percent of the population with a benefit decrease or increase, and the next three columns show the percent change in individual Social Security benefits at three percentiles.

Social Security Taxes Paid

These tables show the projected changes in individual Social Security payroll taxes for current-law payroll taxpayers aged 31 or older in 2030, 2050, and 2070.

The first five columns are the same as the Social Security benefits tables except tax changes are shown rather than benefit changes. Columns 6–8 show the dollar amount changes in individual Social Security payroll taxes at the same percentiles as the percentage change columns.

Household Income

These tables have the same structure and population as the Social Security benefits tables except the effects on household income are shown rather than individual Social Security benefits.

Official Poverty Measure

These tables show the projected changes in poverty for current-law beneficiaries aged 60 or older in 2030, 2050, and 2070.

The first two columns show the poverty rate with and without (current law) the proposed option. The next three columns show the number in poverty (expressed in thousands) with and without (current law) the proposed option and the difference between them. The final column shows the percent change in the number in poverty with the proposed change. This percent change is calculated by dividing the change in thousands in poverty (5th column) by the thousands in poverty without the proposal (3rd column).

CORRECT interpretation: the number of people in poverty would decline by 2 percent.
INCORRECT interpretation: the poverty rate would decline by 2 percent.
Benefit/Tax Ratios

These tables show the projected changes in the benefit/tax ratios for workers born in four birth cohorts: 1960–1969, 1980–1989, 2000–2009, and 2020–2029 who have a tax record from which to calculate a benefit/tax ratio. We excluded those who paid zero payroll taxes over their lifetime and, therefore, could not have a benefit/tax ratio calculated.

The first five columns are similar to the Social Security benefits and taxes paid tables except the percent of the population with a benefit/tax ratio decrease and increase, and the percent change in the benefit/tax ratio at three percentiles are shown.

The last six columns show the distribution of benefit/tax ratios with and without (current law) the proposed option at three percentiles.

The benefit/tax ratio is a money's worth measure that is the lifetime present value of benefits divided by the lifetime present value of payroll taxes. The ratio represents how much in benefits an individual received for every dollar of payroll taxes paid.

How to interpret the ratios:

  • 0% means that an individual received no benefits despite paying payroll taxes (or $0.00 in benefits for every $1.00 of taxes).
  • 50% means that an individual received half as much benefits as was paid in taxes (or $0.50 in benefits for every $1.00 of taxes).
  • 100% means that an individual received the same amount of benefits as was paid in taxes (or $1.00 in benefits for every $1.00 of taxes).
  • 1,350% means that an individual received over 13 times the amount of benefits as was paid in taxes (or $13.50 in benefits for every $1.00 of taxes).

The present value of benefits includes all Social Security benefits the individual received, regardless of earnings record or type of benefit. The present value of payroll taxes includes all the payroll taxes that the individual paid over a lifetime. We use the Social Security Trust Fund interest rate to adjust benefits and taxes to their present values at age 62.

Initial Replacement Rates

These tables are the same as the benefit/tax ratio tables except initial replacement rates are shown for current-law beneficiaries born in four birth cohorts: 1960–1969, 1980–1989, 2000–2009, and 2020–2029. Only beneficiaries with both income and benefit records from which to calculate an initial replacement rate are included in the population. Beneficiaries with zero average indexed monthly earnings (AIME) or zero benefit at claiming (due to the earnings test or other fixed-dollar reductions) are excluded because no replacement took place. When no benefit is received, the initial replacement of earnings may take place in a later year or never, if the beneficiary dies before a benefit is paid.

The initial replacement rate represents how much of the initial AIME is replaced by the initial total monthly Social Security benefit. It is calculated by dividing the initial monthly benefit by the initial AIME.

How to interpret the initial replacement rate values:

  • 20% means that the initial benefit replaced one-fifth of lifetime earnings (or $0.20 in monthly benefits for every $1.00 of AIME).
  • 100% means that the initial benefit replaced all lifetime earnings (or $1.00 in monthly benefits for every $1.00 of AIME).
  • 150% means that the initial benefit replaced one and a half times lifetime earnings (or $1.50 in monthly benefits for every $1.00 of AIME).

The initial monthly benefit is the total individual Social Security benefit received at the person's claiming age, including any spousal, survivor, or disability benefits received. We calculate the replacement rate at claiming age, regardless of what type of benefits the beneficiary claimed.

Interpreting the Profile of Beneficiaries by Race & Ethnicity Tables

For all beneficiaries aged 60 or older, we provide six sets of tables with projections of the following information for 2024, 2030, 2050, and 2070:

  • Social Security benefits,
  • Poverty rates and numbers,
  • Household income distribution and sources,
  • Total earnings,
  • Household wealth, and
  • Health status and costs.

The tables also include the same information for four race and ethnicity groupings:

  • Hispanic or Latino, any race
  • White, non-Hispanic
  • Black or African American, non-Hispanic
  • All other races, non-Hispanic

Each set of tables and what they show are discussed below.

Social Security Benefits

These tables show the projected distribution of individual monthly Social Security benefits at three percentiles. The benefit amount is the total monthly benefit an individual would receive, regardless of the type of benefit or earnings record it came from.

Poverty Rates and Numbers

These tables show the poverty rate and number in poverty under the Official Poverty Measure and the Supplemental Poverty Measure.

The first and third columns show the rates and numbers under the Official Poverty Measure while the second and fourth columns show the same information under the Supplemental Poverty Measure. Further information about the Official and Supplemental Poverty Measures and how they relate to the aged, is available in this paper.

Household Income Distribution and Sources

These tables show the projected distribution of annual household income at three percentiles in the first three columns.

The last five columns show the projected mean share of household income from five sources:

  • Social Security benefits, which include benefits for the individual, spouse, and any children.
  • Annuitized asset income, which includes income from defined contribution plans (such as 401(k) accounts) and personal savings.
  • Defined benefit pension income, which includes the individual's and any spouse's defined benefit pension income.
  • All earnings, including covered earnings (from which Social Security taxes are withheld) and non-covered earnings (no Social Security taxes withheld) of the individual and his or her spouse.
  • Coresident income, which is the income of non-spousal coresidents in the household.

Rows may not sum to 100 percent because minor sources of income are excluded.

Total Earnings

These tables show the projected distribution of annual individual total earnings at three percentiles. Total earnings includes both covered and non-covered wages.

Household Wealth

These tables show the projected distribution of household wealth at three percentiles. Wealth includes retirement account balances and savings, but excludes household equity.

Health Status and Costs

These tables show projected health status and expenses. The first two columns cover the percent and number (in thousands) of beneficiaries who are projected to self-report fair or poor health. The third column shows family median annual health insurance premiums, and the fourth column shows family median annual out-of-pocket health expenses.

Interpreting the Profile of Taxpayers by Race & Ethnicity Tables

For all taxpayers aged 31 or older, we provide six sets of tables with projections of the following information for 2024, 2030, 2050, and 2070:

  • Social Security taxes paid,
  • Covered earnings,
  • Lifetime quarters of coverage,
  • Poverty rates and numbers,
  • Household income distribution, and
  • Household wealth.

The tables also include the same information for four race and ethnicity groupings:

  • Hispanic or Latino, any race
  • White, non-Hispanic
  • Black or African American, non-Hispanic
  • All other races, non-Hispanic

The table sets that differ from the beneficiaries by race and ethnicity profile tables are discussed below.

Social Security Taxes Paid

These tables show the projected distribution of annual individual Social Security taxes paid at three percentiles. Both the employer and employee shares of the Social Security portion of FICA (Federal Insurance Contributions Act) taxes are included.

Covered Earnings

These tables show the projected distribution of annual individual covered earnings at three percentiles. Covered earnings are wages from work that is subject to the Social Security payroll tax. The covered earnings in these tables are not capped at the taxable maximum.

Lifetime Quarters of Coverage

These tables show the projected distribution of lifetime quarters of coverage (QCs) earned by individuals at three percentiles. QCs, also known as credits, are required to be fully insured (40 QCs) under Social Security.

Interpreting the Population Characteristics Tables

Each projection table links (at the top) to the corresponding population characteristics table. From there, you can use the tabs on the right side to view each population analyzed across the MINT projections including:

We use 10-year birth cohorts to increase the sample size to a point where the characteristic subgroups could be examined.

The three columns in every population characteristics table are:

  • Unweighted sample: The number of people in the sample for each row. We use it to verify that we comply with disclosure avoidance policy (see “Sample Size Restrictions”).
  • Population (in thousands): The weighted population in thousands. A value of 71,500 means 71 million, 500 thousand.
  • Share of population: The percentage of the population in that particular characteristic subgroup. The Total row at the top of the table is always 100%. The percentages for the each subgroup underneath should add to 100%. For instance, under Country of Birth, United States may be 84% and Other Countries would be 16%, which adds to 100%.
CORRECT interpretation: 40% of the population is female, 60% of the population is male
INCORRECT interpretation: 40% of females are in this population, 60% of males are in this population

Definitions—Table Rows and Columns

Characteristic Subgroups—Table Rows 1

Total: Refers to the total population of the table.

Sex: Female or Male. Nonbinary identifications are not covered because they are not in the datasets used to build the MINT8 model.

Race/Ethnicity: We list “Hispanic or Latino, any race” first; the rest of the groups (White, Black or African American, and All other races) are non-Hispanic. Additional racial or ethnic identifications are not covered because they are not in the datasets used to build the MINT8 model.

Country of Birth: We differentiate between the United States and other countries.

Age:

  • The beneficiary population includes those aged 60 or older because 60 is the earliest eligibility age for any aged benefits under current law.
  • The taxpayer population includes those aged 31 or older because 31 is the earliest age in MINT for household income and poverty information.

Marital Status: Refers to the marital status in the year of analysis only. An individual's marital status can change in the future and may have been different in the past.

Highest Education Level: Reported number of years of education.

  • Graduate means more than 16 years of education,
  • Bachelor means 16 years of education,
  • Associate means 14–15 years of education,
  • High school means 12–13 years of education, and
  • Less than high school means less than 12 years of education.

Current-Law Poverty Status: Indicates whether the person is in a household that has income above (“above poverty”) or below (“in poverty”) the official poverty line under current law. The household income used for the official poverty measure is the same as the household income used in our results except for how asset income is counted. The official poverty measure of asset income only includes dividend income, interest income, and rental income (non-annuitized) as reported on income tax returns. In contrast, we include annuitized asset income from all household wealth held in defined contribution plans (such as 401(k) accounts) and personal savings in that year. We add the annuitized asset income to account for the expected spend-down of assets in retirement. The asset income value used for official poverty calculations generally produces a substantially lower asset income value than the household income measure.

Current-Law Household Income Quintile: Represents an individual's annual household income under current law, including:

  • household earnings;
  • asset income (annuitized), which includes income from defined contribution plans (such as 401(k) accounts) and personal savings;
  • defined benefit pensions;
  • means-tested income;
  • non-means-tested income;
  • Social Security;
  • Supplemental Security Income; and
  • non-spousal co-residents' income.

We calculate the income quintiles for each year (e.g., 2030, 2050, or 2070) for the population analyzed, determine the dollar thresholds for each income quintile, and assign each beneficiary to the appropriate quintile. The dollar ranges are available upon request.

Current-Law Benefit Type: Some Social Security benefits are based on one's own work, while others are based on the work of a current, divorced, or deceased spouse. The current-law benefit type refers to one of the following benefit types received in the specified analysis year:

  • Retired-worker only: receives only a retired-worker benefit based on his or her earnings record.
  • Widow(er) (includes dually entitled): receives a survivor benefit (may or may not also receive a lower worker benefit from his or her own earnings record, known as dually entitled).
  • Spousal (includes dually entitled): receives a spousal benefit (may or may not also receive a lower worker benefit from his or her own earnings record, known as dually entitled).
  • Disabled-worker only: receives a disabled-worker benefit on his or her earnings record and is under the full retirement age (FRA). Disabled workers convert to retired workers at FRA.

Our results do not show different benefit types a beneficiary might receive under a policy option/proposal or in a different year under current law.

Current-Law Payroll Taxes Quintile: Represents an individual's annual Social Security payroll taxes under current law. We calculate the payroll tax quintiles for each analysis year's population of payroll taxpayers aged 31 or older.

Current-Law Initial AIME Quintile: Represents an individual's average indexed monthly earnings (AIME) under current law at age 62, the earliest eligibility age for retired-worker benefits. We calculate the AIME quintiles for each birth cohort. The dollar ranges are available upon request.

Lifetime Payroll Tax Quintile: Represents the present value of an individual's current-law payroll taxes at age 62. We calculate the payroll tax quintiles for each birth cohort. The dollar ranges are available upon request.

Lifetime Payroll Tax Quintile (Shared): Represents the present value of an individual's current-law payroll taxes at age 62. For married couples, the payroll taxes paid while married are shared equally between them. For never-married individuals, this is the same as the lifetime payroll tax. In any year where an individual is not married, we count only their individual payroll taxes. We calculate the quintiles for each birth cohort. The dollar ranges are available upon request.

Measures—Column Headings

Threshold for Categorization in the “Decrease” or “Increase” Groups
(“Percent of Population with a—[decrease or increase]” Columns)

We categorize individuals as having a “decrease” in the amount being analyzed (benefits, taxes, income, etc.) when a proposal would reduce the analyzed quantity by 1% or more. Individuals are categorized as having an “increase” when a proposal would raise the analyzed quantity by 1% or more. We consider individuals with differences between −1% and 1% to be unaffected.

For example, consider two individuals with benefits under an option/proposal that are lower than benefits without the proposal—individual A has a benefit decrease of 0.8% while individual B has a decrease of 1.6%. We would consider individual A unaffected and not categorize him or her in the “decrease” or “increase” columns. However, we would categorize individual B as having a decrease.

Thus, some beneficiaries who are technically “affected” by a proposal, but who will still receive essentially the same benefit amount (or pay essentially the same taxes, etc.) are considered to be unaffected in our projections.

Percent Change Values
(“Percent Change in [item being analyzed] at the—[three percentiles]” Columns)

Understanding how to interpret the distribution of percent changes is critical to understanding the results correctly.

The formula we use to calculate the percent change for each individual is:

(option amount − current-law amount) ÷ current-law amount

From this distribution of individual percent changes, ranked from high to low, we calculate the 10th percentile, median, and 90th percentile values (see below).

CORRECT interpretation: a −5% median indicates that this is the median of the distribution of individual percent changes (half the individuals have a percent change that is higher and half have a percent change that is lower).
INCORRECT interpretation: a −5% median indicates that the median amount under the option is 5% less than the median amount under current law.

10th, Median, and 90th Percentile Values

The percentiles provide a picture of the distribution of policy option effects or beneficiaries/taxpayers financial status distributed from lowest to highest. The example below is for a percent change in benefits, but also applies to distributions of other policy options, dollar amounts, initial replacement rates, household income levels, etc.

Table example
  Percent change in Social Security benefits at the—
10th percentile Median 90th percentile
Total
 

10th percentile: “2%” means that 10 percent of the population has a benefit change of less than 2 percent, while 90 percent have a benefit change of more than 2 percent.

Median: “4%” means that 50 percent of the population has a benefit change of less than 4 percent, while 50 percent have a benefit change of more than 4 percent.

90th percentile: “21%” means that 90 percent of the population has a benefit change of less than 21 percent, while 10 percent have a benefit change of more than 21 percent.

CORRECT interpretations include:
  • 10% of this population has a benefit change of less than 2%.
  • 10% of this population has a benefit change of more than 21%.
  • 40% of this population has a benefit change of between 2% to 4%.
  • 40% of this population has a benefit change of between 4% to 21%.
  • 50% of this population has a benefit change of less than 4%.
  • 50% of this population has a benefit change of more than 4%.

Additional Notes

Dollar Amounts

All dollar amounts are presented in today's dollars, meaning that they are in real dollars (inflation-adjusted) for the year the table is produced. If a table is run in 2021, the dollars are in 2021 dollars and the table will note that in the column label. Tables run in 2024 will be in 2024 dollars, and so on.

Sample Size Restrictions

To maintain the privacy of survey respondents, our tables have built-in disclosure avoidance protections that suppress an entire characteristic subgroup if the sample size for any row in that subgroup is less than 100 individuals. This minimum sample size protects privacy so that we can show the 10th and 90th percentiles based on a sample size of at least 10.

For example, if there are only 82 widow(er)s in the Widowed row of the Marital Status subgroup, the entire Marital Status subgroup is removed from the table. We remove the entire subgroup to avoid secondary or tertiary disclosure issues. The subgroups below Marital Status in this example would automatically move up the page. A “short” table can reveal at a glance that at least one subgroup has been removed.

Columns that show percent of the population with a decrease or increase in whatever value is being shown (e.g. benefits, payroll tax, household income) have disclosure restrictions on the numerator's sample size as well as the denominator. The numerator must have either zero cases or meet a minimum numerator threshold of 10 for the table to display it. The table would suppress any characteristic subgroup that has a percentage based on a numerator of 1–9 cases. (MINT results are weighted, but for this example, everything is presented in unweighted sample sizes.)

There is an exception for low numerator situations where the table would show a 0% for any numerator from zero up to and including the minimum numerator threshold. If a particular percentage was based on seven records with a denominator of 30,000, it would produce a percentage of 0.02%. By only showing percentages in single digits, we would display this result as 0%, which would be the same value displayed for any numerator from 0–149.

This exception is important because there are policy options where 27,000/30,000 beneficiaries (90%) would receive a benefit increase while 8/30,000 receive a decrease. Without the exception, the very small decrease numerator would suppress a number of subgroups for both the increased and the decreased results, which limits the presentable results more than is necessary for disclosure avoidance.

1 All the characteristic subgroups are not in every table. Birth cohort tables (benefit/tax ratios and initial replacement rates) do not have age or marital status breakouts. Characteristic subgroups can also drop out of tables because of sample size restrictions (see “Sample Size Restrictions” for details).