Model 2.0 – Middle for Retirement Analysis

The transient’s key findings are:

  • The Nationwide Retirement Danger Index captures the share of working-age households that would fall brief in retirement.
  • Current upgrades to the Index enable for extra correct measurement and evaluation.
  • The retooled Index confirms earlier findings: half of households will be unable to keep up their way of life once they retire.


The Nationwide Retirement Danger Index (NRRI) measures the share of working-age households that’s “at-risk” of being unable to keep up their pre-retirement way of life in retirement.  The train includes evaluating households’ projected substitute charges – retirement revenue as a proportion of pre-retirement revenue – with goal charges that might enable them to keep up their dwelling commonplace.  The important thing discovering is that roughly half of the nation’s working-age households are susceptible to falling brief even when they work to age 65 and annuitize all their property.  This result’s secure over time, with some ups and downs reflecting financial and market fluctuations.  

Since its inception, the Middle has periodically made modest adjustments to the NRRI.  Lately, nevertheless, we undertook a serious overhaul to include new analysis findings and methodological advances.  To take care of transparency of the NRRI, this transient summarizes the adjustments and presents outcomes for the recalculated Index.

The dialogue proceeds as follows.  The primary part describes the nuts and bolts of the NRRI.  The second part paperwork the key enhancements within the NRRI’s underlying methodology.  The third part presents the principle outcomes, that are typically in line with these from earlier NRRI publications.  The ultimate part concludes that retirement readiness stays a serious problem for a lot of of right now’s working-age households; they should save extra and/or work longer to enhance their prospects for a safe retirement. 

Nuts and Bolts of the NRRI 

The NRRI is constructed with knowledge from the Federal Reserve’s Survey of Shopper Funds (SCF), a triennial nationally consultant family survey.  Calculating the NRRI includes three steps: 1) projecting a substitute charge – retirement revenue as a share of pre-retirement revenue – for every SCF family ages 30-59; 2) setting up a goal substitute charge that might enable every family to keep up its pre-retirement way of life in retirement; and three) evaluating the projected and goal charges to search out the share of households “in danger” (see Determine 1 ).

Projecting Family Substitute Charges

The substitute charge calculation begins with a projection of retirement revenue at age 65.  This measure is outlined broadly to incorporate revenue from Social Safety, outlined profit (DB) plans; monetary property each in outlined contribution (DC) plans and saved instantly; and housing, which incorporates imputed lease in addition to dwelling fairness.  The asset values for the latter three parts – DC property, non-DC monetary property, and housing fairness – are derived from reported wealth within the SCF.  They’re every projected individually to age 65 based mostly on their respective wealth-to-income ratios by age, that are secure over time.  As proven in Determine 2, the general wealth-to-income ratios from the 1983-2019 SCF surveys relaxation roughly on high of each other, bracketed by 2007 values on the excessive facet and 2013 values on the low facet.  

Line graph showing the ratio of wealth to income by age from the Survey of Consumer Finances, 1983-2019

The NRRI then assumes that households convert all their property, together with monetary property, 401(ok)/IRA balances, and proceeds from a reverse mortgage, right into a stream of revenue by buying an inflation-indexed annuity.

Sources of retirement revenue that aren’t derived from reported wealth within the SCF are estimated instantly.  Particularly, Social Safety advantages are calculated based mostly on estimated earnings histories for every member of the family, listed to nationwide common wage development.  DB pension revenue is predicated on the quantity reported by survey respondents.

The remaining step is to calculate common lifetime revenue previous to retirement.  Pre-retirement revenue for owners contains earnings and imputed lease from housing.  Common lifetime revenue (with earnings, once more, listed to common wage development) then serves because the denominator for every family’s substitute charge.  This measure excludes revenue from property.

Estimating Goal Substitute Charges

To find out the share of the inhabitants that can be in danger requires evaluating projected substitute charges with a benchmark charge.  A generally used benchmark is the substitute charge wanted for households to keep up their pre-retirement way of life in retirement.  Individuals sometimes want lower than their full pre-retirement revenue since they typically pay much less in taxes, not want to save lots of for retirement, and sometimes have paid off their mortgage.  Thus, a larger share of their revenue is out there for spending.  The Index estimates the goal substitute charges for several types of households utilizing a consumption-smoothing mannequin, which is predicated on the idea that households need the identical degree of consumption in retirement as they’d earlier than they retired. 

Calculating the Index

The ultimate step is to check every family’s projected substitute charge with the goal from the consumption-smoothing mannequin.  These whose projected substitute charges fall greater than 10 % beneath the goal are deemed to be susceptible to having inadequate revenue to keep up their pre-retirement way of life.  The Index is solely the share of all working-age households that fall greater than 10 % wanting their targets.

Main Enhancements

Though the general modeling framework described above stays unchanged, the up to date NRRI contains the next main enhancements. 

Tasks retirement revenue extra precisely.  The largest change to the NRRI is modifying the projection of wealth-to-income ratios for every family to extra precisely mirror the wealth distributions noticed within the knowledge.  The earlier projection methodology was based mostly on the imply wealth development paths estimated by a linear regression strategy.  A disadvantage to this strategy is that, given rising wealth inequality, the outcomes are more and more biased towards the buildup paths of richer households.  Thus, the outdated methodology tended to overestimate the wealth of middle- and lower-income households.  In distinction, the brand new methodology initiatives wealth based mostly on median values, which makes the wealth projections at retirement higher mirror the noticed distributions.

Higher displays the shift from DB to DC plans.  The rising share of employees lined by DC plans for the reason that Nineteen Eighties implies that the extent and sample of DC asset accumulation differs throughout delivery cohorts.  To account for these variations, the brand new methodology initiatives DC property individually for 3 broad cohorts: 1) employees born earlier than 1945, who had been not less than midway into their careers when protection beneath DC plans started to increase in 1980; 2) employees born from 1945-1955, who had been early of their careers throughout the transition to DC plans; and three) employees born after 1955, whose careers principally fall within the years when DC plans had been already prevalent.

Fashions monetary debt individually.  The unique wealth projection methodology subtracts households’ non-mortgage money owed from their monetary property and initiatives the ensuing internet monetary property as a single variable.  Analyses of earlier NRRI outcomes recommend that the dynamics of monetary debt could be of curiosity on their very own.  For instance, middle-age and middle-income households noticed very restricted enhancements in retirement preparedness in 2016 partly on account of elevated non-mortgage borrowing.  The brand new methodology now initiatives monetary property and non-mortgage debt individually, permitting for extra in-depth evaluation in addition to counterfactual evaluation specializing in borrowing.

Refines the goal substitute charge mannequin.  Within the authentic methodology, goal substitute charges had been calculated and matched to noticed households within the SCF in an approximate method: targets had been calculated for 12 family classes decided by 4 family varieties (single male, single feminine, married with two earners, and married with one earner) and three revenue teams, after which assigned to noticed households utilizing these traits.  Beneath the brand new methodology, a lot richer family traits are used for calculating goal charges, permitting the projected substitute charges to be matched to tons of of targets, which yields extra correct estimates.  Particularly, matching is now based mostly on rather more fine-grained revenue teams and households’ precise DB protection and homeownership standing. 

Incorporates the Earned Revenue Tax Credit score (EITC) in substitute charges.  The outdated methodology didn’t embody the EITC within the calculation of pre-retirement revenue.  Nevertheless, the EITC is essential for low-income households throughout their working years, so the brand new methodology contains it to raised seize the revenue these households might want to substitute in retirement. 

Different adjustments embody:

  • Annuity components used to annuitize projected wealth now higher mirror the degrees and tendencies noticed within the annuity market.
  • Reverse mortgage calculations now use up to date rate of interest assumptions and principal restrict issue tables.  
  • Key mannequin assumptions and inputs, akin to wage development, rates of interest, inflation, and mortality tables are up to date.
  • The NRRI codebase has been largely moved from Stata and Excel spreadsheets to Python, permitting for extra versatile mannequin growth, improved computation, and simpler upkeep. 

The New Nationwide Retirement Danger Index

Regardless of the intensive adjustments in methodology, the general degree and time sample of the Index stay the identical as earlier than (see Determine 3).  Thus, a very powerful discovering nonetheless holds: about half of working-age households will be unable to keep up their pre-retirement dwelling commonplace.  Furthermore, the sample continues to mirror the well being of the economic system.  The Index elevated considerably from 2007 to 2010 throughout the Nice Recession, after which declined a bit from 2013 to 2019 because the economic system loved low unemployment, rising wages, robust inventory market development, and rising housing costs.  These enhancements had been modest on account of some countervailing longer-term tendencies – such because the gradual rise in Social Safety’s Full Retirement Age (FRA) and the continued decline of rates of interest – which made it tougher for households to realize retirement readiness.

Bar graph showing the previous and new National Retirement Risk Index, 2004-2019

Patterns by Family Sort

Along with the time sample, the NRRI patterns by age, revenue, and wealth are additionally typically in line with our earlier publications.  

Age.  The NRRI in 2004 exhibits a big discrepancy in retirement readiness by age group, which displays the dramatic adjustments within the retirement panorama such because the shift of pension protection from DB to DC plans, rising life expectancy, and the rise within the FRA (see Desk 1).  Because the tendencies for these underlying components stabilized over time and their affect totally materialized, the age discrepancy within the NRRI has narrowed.

Table showing the percentage of households "at risk" at age 65 by age group, 2004, 2010, and 2019

Revenue.  Households’ retirement preparedness in all revenue teams was closely affected by the Nice Recession (see Desk 2).  The center and the best thirds noticed important enchancment from 2010-2019 on account of rebounding housing and fairness costs.  In distinction, households within the backside third noticed nearly no enchancment as they’re much less more likely to personal a home and take part in DC plans, and have few monetary property.  As well as, the rise in wage development for lower-income employees, which is nice information typically because it improves their present way of life, results in decrease projected Social Safety substitute charges as a result of progressive profit method.  The rise within the FRA additionally has a very giant affect on low-income households, who rely virtually completely on Social Safety for retirement revenue.

Table showing the percentage of households "at risk" at age 65 by income group, 2004, 2010, and 2019

Wealth.  When considered by wealth, households’ retirement preparedness typically exhibits the same sample.  The discrepancy between the highest and backside wealth teams, although, is far bigger than these by revenue (see Desk 3), reflecting the truth that wealth inequality is extra extreme than revenue inequality.

Table showing the percentage of households "at risk" at age 65 by wealth group, 2004, 2010, and 2019


Since its inception, the NRRI methodology has acquired many enhancements and updates, which improve the projections of households’ wealth and substitute charges at retirement and the estimations of the goal substitute charges.  With the newest upgrades, the NRRI can extra precisely measure the retirement preparedness of working-age households and consider the affect of financial and coverage components on retirement safety. 

After recalculating the NRRI utilizing essentially the most up to date methodology, the underside line from our earlier research nonetheless holds: about half of right now’s households won’t have sufficient retirement revenue to keep up their pre-retirement way of life, even when they work to age 65 and annuitize all their monetary property, together with the receipts from a reverse mortgage on their houses.  The robustness of the outcomes confirms the retirement saving challenge confronted by right now’s working-age households, and that we have to repair our retirement system in order that employer plan protection is common.  Solely with steady protection will employees be capable of accumulate sufficient assets to keep up their way of life in retirement. 


Chen, Anqi, Wenliang Hou, and Alicia H. Munnell. 2020. “Why Do Late Boomers Have So Little Retirement Wealth?” Subject in Transient 20-4. Chestnut Hill, MA: Middle for Retirement Analysis at Boston School.

Munnell, Alicia H., Anthony Webb, and Luke Delorme. 2006. “Retirements at Danger: A New Nationwide Retirement Danger Index.” Chestnut Hill, MA: Middle for Retirement Analysis at Boston School.

Munnell, Alicia H., Anqi Chen, and Robert L. Siliciano. 2021. “The Nationwide Retirement Danger Index: An Replace from the 2019 SCF.” Subject in Transient 21-2. Chestnut Hill, MA: Middle for Retirement Analysis at Boston School.

Munnell, Alicia H., Wenliang Hou, and Geoffrey T Sanzenbacher. 2018. “Nationwide Retirement Danger Index Exhibits Modest Enchancment in 2016.” Subject in Transient 18-1. Chestnut Hill, MA: Middle for Retirement Analysis at Boston School.

U.S. Financial institution of Governors of the Federal Reserve System. Survey of Shopper Funds, 1983-2019. Washington, DC.

Supply hyperlink

Related Articles


Please enter your comment!
Please enter your name here

- Advertisement -spot_img

Latest Articles