Veritas propter investigationem [Truth through research]
TitlePursuing Universal Retirement Security Through Automatic IRAs (2009)
AuthorsMark Iwry
 David John
InstitutionRetirement Security Project
TopicsBehavioural economics
 Auto-enrolment, opt-out
 Saving for retirement
 Workplace saving schemes
CountryUnited States
Date Published2009
Date posted on PR09 Sep 2014
Iwry, M John, D, (2009). Pursuing Universal Retirement Security Through Automatic IRAs (2009) Retirement Security Project,

PensionReforms’ summary and comments

Commentators in the United States fret about the decline of Tier 3, occupational retirement saving schemes.  This 2009 report proposes a new option: so-called ‘automatic Individual Retirement Accounts’ (IRAs).  These are already available as a tax-preferred way to accumulate financial assets for retirement.


“Roughly half of all working Americans work for employers that offer no retirement plan.  Thus about 78 million workers have no way to save on the job for the day when they stop collecting a paycheck.”


This apparently dire position seems to be made worse by  “…a national saving rate that has been declining steadily for most of the past twenty years and the unlikelihood that Social Security will be able to provide increased benefits, makes inadequate retirement saving a major national problem.”


There are generous tax breaks for both IRAs and their 401(k) occupational equivalents but these seem insufficient to attract people to put aside money for retirement.  The report proposes “an ambitious yet practical set of initiatives to expand retirement saving dramatically.”


“We propose making saving automatic—and hence easier, more convenient, and more likely.  This strategy has been shown to be remarkably effective at boosting participation in workplace-based 401(k) retirement savings.  We would extend this strategy to most employees who have no access to 401(k) plans by combining several key elements of our current system: payroll-deposit saving; automatic enrollment; low-cost, diversified default investments; and individual retirement accounts (IRAs).”


The proposal therefore draws on work done under the banner of behavioural economics.  PensionReforms has reviewed a number of reports that look at this for retirement savings – see here, here and here for examples.


The proposal envisages a raft of rules that employers would have to comply with, starting from the requirement to make an IRA available, supervising payroll deductions, choosing a provider and then default-enrolling all employees, not already members of another scheme.  Employees could opt-out but would need to sign a waiver.  Then there would be another set of rules to govern the way the IRAs operated, including ‘automatic investment choice’.


“Automatic IRAs would not crowd out or compete with 401(k) plans.  To the contrary, we would hope that successful experience with the new, automatic IRAs would lead more employers to step up to 401(k)s and then to match employee contributions—if not dollar for dollar, then perhaps fifty cents on the dollar or some other ratio.”


The key would be to make enrolment, contribution and withdrawal processes happen automatically “…unless the individual employee or employer stepped in and affirmatively chose a different course.”


PensionReforms wonders about all this.  There is an implicit assumption that citizens do not know what they want; employers don’t understand how to set-up their remuneration arrangements and that non-members aren’t saving enough for retirement.  But where is the evidence for all this?  Here, for example, is a report that suggests US citizens might be saving enough for retirement.


How do we know that individuals aren’t saving enough unless we understand what each individual’s total assets and liabilities are now; what they might be by the time the saver reaches whatever age the saver wants to stop working; how the saver manages the retirement process (full or partial retirement); what their Social Security entitlements might be and, most importantly, how much the saver thinks is needed for a reasonable standard of living in retirement (and what trade-offs that entails with today’s living standards)?


The report seems to assume that whatever the employee has through the workplace, especially if that is ‘nothing’ is everything the employee is doing about saving for retirement.  That almost certainly isn’t the case.


It does seem illogical for an employee to turn down the opportunity of higher total remuneration through the employer’s subsidies (if those exist). 


However, it does not follow that, by failing to join (or failing to maximise the subsidy or opting out) the employee is under-saving for retirement.  Proof of that can be obtained only following a close examination of the employee’s total household position, particularly where there is more than one earner.  Then we need to understand the household’s aspirations.  The report did not contemplate those possibilities and should have.  (File size 2.5 MB; 36 pp) 697