FLORIDA TAX REVIEW


FLORIDA TAX REVIEW 
Volume 9 2008 Number 2 

 

119 

 

SLOUCHING TOWARDS A CONSUMPTION TAX AND THE  

END OF RETIREMENT INCOME SECURITY 
 

 by 

 

 Norman Stein

 

 

I.   INTRODUCTION. ............................................................................ 121 

II.  OVERVIEW OF CURRENT TAX – SUBSIDIZED 

 RETIREMENT SYSTEM .................................................................. 125 

III.   THE WHITE HOUSE SAVINGS PROPOSALS .................................. 129 

  A. Dividend Exclusion. ................................................................. 129 

B. New Personal Savings Accounts. ............................................ 129 

1. Retirement Savings Accounts. .................................... 130 

2. Lifetime Savings Accounts .......................................... 131 

C. Employer Retirement Savings Accounts. ................................. 132 

IV.  CRITIQUE OF WHITE HOUSE PRIVATE  

 SAVINGS PROPOSALS .................................................................... 134 

A. Savings Initiative and Employment-Based  

Retirement Plans….. ...................................................................... 135 

B. Efficacy of Bush Savings Proposals for Creating Retirement 

  Income Security for Middle and Lower Income Workers .............. 144 

   1. Obstacles to Rank-and-File Utilization of Bush 

Savings Proposals ............................................................. 145 

2. Production of Retirement Income ............................... 147 

C. Savings Initiative and Rate of Return on  

  Retirement Savings.. ....................................................................... 150 

 1 Loss of Investment Opportunities ............................... 150 

2.   Fiduciary Protections ................................................. 150 

    3. Education ................................................................... 151 

4. Costs ........................................................................... 152 

5. Effect of the Proposed Income Exclusion of Stock 

 Dividends on Investment Returns ............................... 152 

  D. Budget and Tax Considerations ................................................ 153 

                                                 

 Douglas Arant Professor of Law, University of Alabama School of Law. I 

presented a version of this paper at a 2003 symposium on public policy for 

retirement security at the Ohio State Moritz College of Law. I express my deep 

appreciation for the inordinately helpful comments of Professors Grayson M. P. 

McCouch at San Diego and James D. Bryce at Alabama. 



120 Florida Tax Review [Vol. 9:2 

 

V.         WEALTH ACCUMULATION V. RETIREMENT  
 INCOME SECURITY  ....................................................................... 155 

VI.       THE ERSA AND NATIONAL RETIREMENT POLICY .................... 159 

VII. CONCLUSION…………………….. ............................................... 162 



2008] Slouching Towards a Consumption Tax 121 

 

 

SLOUCHING TOWARDS A CONSUMPTION TAX AND THE 

END OF RETIREMENT INCOME SECURITY 
 

 by 

 

 Norman Stein 

 

 I. INTRODUCTION 
 

In 2003, the Bush Administration proposed two ideas
1
 that I (and 

others) feared would undermine the retirement income security of millions of 

Americans.
2
 The first idea was to create two new tax-favored personal 

savings vehicles: the retirement savings account and the lifetime savings 

account.
3
 These tax-benefited accounts, which would have existed outside 

the employer-based retirement system and replace individual retirement 

accounts, would each have permitted annual contributions of $7,500 each, or 

$15,000 in the aggregate.
4
 (An individual could also have set up lifetime 

                                                 
1. See, e.g., Department of the Treasury, General Explanation of the 

Administration’s Fiscal Year 2004 Revenue Proposals, Feb. 3, 2003 (“General 

Explanation”), reprinted in Tax Notes Today, Feb. 4, 2003; Bush Proposes Tax Free-

Savings Plans, available at usgovinfo.com/library/weekly/aataxfree.htm; Q&A about 

Bush Tax-Free Savings Proposals, at usgovinfo.com/library/weekly/aataxfree 

_qu.htm. 

2. The 2003 proposal was opposed by pension trade groups generally 

aligned with business interest (such as the American Society of Pension Actuaries, 

see, ASPA, Statement for the Record, Committee on Ways and Means, Hearing on 

the Presidents’ Economic Growth Proposals, Mar. 5, 2003; and the Profit-Sharing 

401(k) Council of America, see, Martin A. Sullivan, Economic Analysis: Does the 

Trickle-Down Theory for Pensions Hold Water?, 98 Tax Notes 1180 (2003) 

(hereinafter, Trickle-Down Theory for Pensions), and liberal organizations generally 

aligned with the interests of consumers, Leonard Burman, “Key Thoughts on RSAs 

and LSAs,” available at www.urban.org/url.cfm?ID+1000600 (Urban Institute); 

Robert Greenstein and Joel Friedman, “President’s Savings Proposals Likely to 

Swell Long-Term Deficits, Reduce National Savings, and Primarily Benefit Those 

With Substantial Wealth, available at www.cbpp.org/1-20-04tax.htm (Center on 

Budget and Policy Priorities); “Progressivity and Saving: Fixing the Nation’s 

Upside-Down Incentives for Savings,” Statement of Peter R. Orszag, Joseph A. 

Pechman Senior Fellow, The Brookings Institute, Before the House Committee on 

Education and the Workforce, Feb. 25, 2004, www.edworkforce.house.gov/hearings/ 

108th/fc/pensions022504/orszag.htm Brookings Institute). Former Congressman 

Rob Portman, a Republican who had a strong interest in pension policy and was 

generally considered close to the White House, also opposed the proposal. See, 

Trickle-Down Theory for Pensions, 98 Tax Notes at 1180. 

3. General Explanation, supra note 1, at ¶¶ 414-417. 

 4. Id. 



122 Florida Tax Review [Vol. 9:2 

 

savings accounts for family members without earned income, into which he 

could contribute $7,500 annually.) The contributions were to be made on an 

after-tax basis, but the earnings would be exempt from income tax, a 

schemata based on the Roth IRA.
5
 The RSA would impose penalties on pre-

retirement withdrawals; the LSA would not.
6
 These ideas did not find favor 

in the Congress and were generally opposed by the business community, 

organized labor, and liberal think tanks, for reasons I will summarize 

shortly.
7
 Nevertheless, the White House

8
 and members of Congress

9
 have 

introduced similar proposals in 2004, 2005, 2006, 2007, and 2008, and the 

President’s Advisory Panel on Federal Tax Reform adopted retirement and 

lifetime savings accounts as part of its blueprint for reform, which was 

released in late 2005. Thus, the idea of these new accounts retains political 

and intellectual currency, like the proverbial bad nickel, and they, or 

variations on them, are almost certainly to survive the end of the Bush 

presidency. 

 The second new idea advanced by the White House in 2003 was the 

exclusion of stock dividends from taxation.
10

 Depending on which press 

account you credit, this idea resulted either from President Bush’s 

serendipitous courtesy drop-in on a panel at his Waco economic summit in 

which Charles Schwab advocated elimination of the so-called double tax on 

corporate income through exclusion of dividends,
11

 or from a direct remark 

by Schwab to him on the same topic.
12

 A toned-down variation on this idea 

was enacted by Congress in 2003, with dividend income taxed at a maximum 

15% rate, the same rate that now sets the ordinary ceiling on taxation of 

                                                 
 5. Id. 

6. Id. Cf. IRC § 408A (treatment of Roth IRAs). 

7. See, supra note 2. 

8. See, Department of Treasury, General Explanation of the 

Administration’s Fiscal Year 2008 Revenue Proposals (Feb. 2007); Department of 

Treasury, General Explanation of the Administration’s Fiscal Year 2007 Revenue 

Proposals (Feb. 2006); Department of Treasury, General Explanation of the 

Administration’s Fiscal Year 2006 Revenue Proposals (Feb. 2005); Department of 

Treasury, General Explanation of the Administration’s Fiscal Year 2005 Revenue 

Proposals (Feb. 2004). 

9. See, S. 547, 109th Cong., 1st Sess. (Mar. 8, 2005)(Senate bill providing 

for Retirement and Lifetime Savings Accounts); H.R. 1161, 109th Cong., 1st Sess. 

(Mar. 8, 2005) (House bill providing for Retirement and Lifetime Savings 

Accounts). 

10. General Explanation (2003), supra note 1, ¶ 33. 

11. See, Cox News Service, Charles Schwab Plays Cagey With Bush, 

August 17, 2003, available on Lexis. 

12. See, Bill Saporito, Get Ready for Class Warfare, Time Magazine, Jan. 

20, 2003, at 32. 



2008] Slouching Towards a Consumption Tax 123 

 

 

capital gain.
13

 Some observers see the 15% tax rate not as closing the door 

on the debate on either dividend taxation or the maximum capital gains rate, 

but rather as a foot in the door toward the elimination of any tax on dividend 

income or capital gains.
14

 And again, the President’s Advisory Panel on 

Federal Tax Reform included a plan that would exclude all corporate 

dividends from income.  

 On the surface, neither the expanded savings account idea nor the 

exclusion of dividends from income appears to threaten a person’s ability to 

save for retirement. Indeed, the RSA/LSA proposal would seem to increase 

opportunities to save for retirement, and the elimination of taxes on 

dividends would seem to make savings, including retirement savings, 

generally more attractive.
15

 But I will suggest in this paper that the 

proposals, individually and collectively, would negatively influence the 

creation of retirement savings for many and perhaps most working 

Americans.  

 Several of the arguments I develop in this paper relate to my 

assessment that providing tax-benefited investment alternatives to employer-

sponsored retirement plans would make it less attractive for employers to 

sponsor retirement plans, resulting in fewer and less generous plans. There is 

reason to think that middle- and lower-income people would not use the new 

savings arrangements sufficiently to replace the retirement savings they 

would lose because of the White House proposals’ negative effect on 

employer-sponsored plans.
16

 Moreover, the new savings arrangements, by 

and large, would not require preservation of assets for retirement,
17

 nor 

would they provide mechanisms to help budget the use of retirement assets 

so that they are not exhausted before death.
18

 They would provide fewer 

spousal protections.
19

 They would subject participants to higher 

administrative costs and might adversely affect the rate of return many 

working people will realize on their savings.
20

 Finally, the tax treatment of 

the new savings arrangements may complicate the possible future task of 

                                                 
13. IRC § 1(h)(11) (qualified dividends treated as part of net capital gain). 

14. Wesley Elmore, Senate Finance Panel Examines Extension of Popular 

Tax Cuts, 108 Tax Notes 45 (2005). 

15. This is certainly the view of the Department of Treasury, which in its 

press release on the proposals contended that “these bold new accounts will give 

more hardworking Americans the chance to save so they can enrich their lives and 

strengthen their retirement security.” See, Bush Proposes Tax Free-Savings Plans, 

supra, note 1. 

16. See text accompanying notes 120-130, infra. 

17. See text accompanying notes 131-141, infra. 

18. See text accompanying notes 133, 142-143, infra. 

19. See text accompanying note 134, infra. 

20. See text accompanying notes 144-152, infra. 



124 Florida Tax Review [Vol. 9:2 

 

remedying the old age-security problems that I suggest will almost certainly 

be exacerbated if the White House proposals are ever adopted.
21

 

 The White House proposals also reflect a theme that Patricia Dilley 

and I explored in an earlier article: the continuing seismic shift in 

government retirement policy from collective retirement income security to 

individual wealth creation.
22

 The White House proposals would carry this 

theme further than anything that has emerged from Congress since the 

enactment of the income tax. In this respect, the proposals share an 

ideological mooring with the various proposals to privatize Social Security, 

which would also sacrifice the idea of shared societal responsibility for the 

income security of older Americans to the idea of individual creation of 

wealth.
23

 

 On a broader and perhaps even more significant theme, as we amble 

down the road towards a tax system that favors savings over consumption,  

we also amble towards interference with our principal strategy to help 

working people save for retirement: using tax benefits to secure employer 

sponsorship of retirement savings plans covering a broad cross section of 

American workers.  

 The White House proposals did include one idea whose broad 

framework (but not necessarily its details) is attractive: the Employer 

Retirement Savings Account (“the ERSA”).
24

 The ERSA would redesign and 

streamline the complex and often irrational rules for the numerous employer-

sponsored defined contribution plans, creating a single type of defined 

contribution format for such plans. This approach would build on rather than 

abandon the existing employment-based retirement regime and, if stripped 

from the other Bush proposals, could provide a vehicle for debate over how 

best to balance the employer interests in administrative simplicity and low 

compliance costs with the societal goal of providing meaningful benefits for 

individuals who otherwise would not save sufficiently for retirement.  

 This article proceeds as follows: the section immediately below 

provides an overview of the current employment-based retirement system, 

followed by a section describing the White House tax proposals as initially 

proposed in 2003, as modified in subsequent White House Budget proposals, 

and as again modified by the President’s Advisory Panel on Federal Tax 

Reform. The next section describes the ways in which the proposals (or 

                                                 
21. See text accompanying notes 153-158, infra. 

22. See, Norman Stein & Patricia Dilley, Leverage, Linkage, and Leakage: 

Problems with the Private Pension System and How They Should Inform the Social 

Security Reform Debate, 58 Wash. & Lee L. Rev. 1369 (2001); cf. Edward Zelinsky, 

The Defined Contribution Paradigm, 114 Yale L.J. 451 (2004). 

23. Patricia Dilley, Taking Public Rights Private; The Rhetoric and Reality 

of Social Security Privatization, 41 B.C.L.Rev. 975 (2000). 

24. General Explanation (2003), supra note 1, at ¶ 448. 



2008] Slouching Towards a Consumption Tax 125 

 

 

similar proposals) would, if enacted, undercut the retirement income security 

of many working Americans. The succeeding section provides a reflection on 

the ideological underpinnings of the White House proposal. The final 

section, using the ERSA proposal as a starting place, provides some ideas to 

improve the ability of working people to prepare for retirement through 

incremental reforms to the current system. The final section also briefly 

discusses the shape that fundamental reform should take if we are to abandon 

the current employer-based system – which would likely be the effective 

result of a serious move toward a consumption tax – in favor of new ways to 

create retirement income security for American working people. 

 

II. OVERVIEW OF CURRENT TAX-SUBSIDIZED  

RETIREMENT SYSTEM 
 

 The Internal Revenue Code endows certain advance-funded 

retirement arrangements with the benefit of relatively pure income tax 

deferral.
25

 Most of the deferral is initially traceable to investment return on 

employer contributions to employer-sponsored plans in both the private and 

public sector, although individuals may also contribute to individual 

retirement accounts
26

 or employer plans that accept employee contributions, 

such as the now-ubiquitous 401(k) plan.
27

 The associated tax expenditures 

for these retirement savings arrangements is projected to exceed $100 billion 

annually for the next fiscal five years
28

 (the period for which tax 

expenditures are projected). 

 The commonly accepted rationale for the income tax expenditure is 

to help as many Americans as possible create income security for that period 

of life when they are no longer supporting themselves with wage income.
29

 

Given that the mechanism for the tax expenditure is tax deferral, that the 

benefits of tax deferral correlate with marginal tax rates, and that individuals 

with high marginal tax rates are the most likely class of individuals to save 

adequately for retirement without governmental incentives, the tax 

mechanism might strike us as an irrational means of effecting the goal of 

                                                 
25. See, e.g., Gary Boren & Norman Stein, Qualified Deferred 

Compensation Plans, Chapter 1; Michael J. Graetz, The Troubled Marriage of 

Retirement Security and Tax Policy, 135 U. Pa. L. Rev. 851 (1987). 

26. IRC § 408 (IRA); IRC § 408A (Roth IRA). 

27. See, e.g., IRC § 401(k). For a thoughtful discussion of some of the 

issues surrounding 401(k) plans and savings behavior, see, Alicia Munnell & Ankia 

Sunden, Coming Up Short (2004). 

28. Joint Comm. On Taxation, Estimate of Federal Tax Expenditures for 

Fiscal Years 2007-2011 (Sept. 24, 2007). 

29. See, e.g., Dan McGill, Fundamentals of Private Pensions 75 (7th ed. 

1996). 



126 Florida Tax Review [Vol. 9:2 

 

increasing retirement savings for as many working people as possible. 

Understood another way, however, this upside-down tax subsidy is an 

arguably rational component of a two-step governmental strategy to enlist 

the private sector in building retirement savings for lower- and moderate-

income workers. 

 The strategy is, first, to make the tax benefits of employer-sponsored 

plans sufficiently attractive to the tax-sensitive people who own and manage 

businesses so that they will decide to set up plans to capture tax benefits for 

themselves, and, second, to require such plans, once established , to provide 

meaningful benefits not only to the people who set them up, but also to 

lower- and moderate-income workers.
30

 The Code effects the latter part of 

the strategy through a series of statutory provisions, most prominently the 

nondiscrimination rules, which require plans to cover a percentage of a 

firm’s non-highly compensated employees
31

 and to provide them with 

benefits comparable, as a percentage of pay, to the benefits earned by the 

highly compensated.
32

 The pension economist Alicia Munnell described this 

elaborate two-step in this way: “the rationale for favorable tax treatment of 

qualified plans is that retirement benefits for rank-and-file employees will 

exist if Congress provides tax incentives that induce higher paid employees 

to support the establishment of employer-sponsored pension plans.” 
33

 

 This carrot/stick pension tax regime has been subject to criticism.
34

 

Many firms respond to the tax incentive to create plans, but manipulate the 

labyrinth complexities of the nondiscrimination rules to minimize benefits 

for rank-and-file employees and maximize them for more affluent 

employees, who in the absence of employer-sponsored plans would 

presumably save for retirement on their own, at least to some extent.
35

 

Moreover, some firms (particularly marginal firms) simply do not respond to 

the incentives and fail to sponsor pension plans.
36

 Thus, the system is both 

                                                 
30. See, e.g., Daniel I. Halperin, Tax Policy and Retirement Income: A 

Rational Model for the 21st Century, Search for a National Retirement Income 

Policy 157 (1987). 

31. IRC § 410. 

32. IRC § 401(a)(4). 

33. Alicia Munnell, The Economics of Private Pensions 51 (1982). 

34. See, Bruce Wolk, Discrimination Rules for Qualified Retirement Plans: 

Good Intentions Confront Economic Reality, 70 Va. L. Rev. 419 (1983). The regime 

has been called “Rube Goldbergian.” See, Trickle-Down Theory for Pensions, supra 

note 2. 

35. Id. 

36. Employee Benefits Research Institute, 2004 Small Employer 

Retirement Survey (only 28% of employees at firms of fewer than 100 are covered 

by retirement plans; only 20% of employers with fewer than 25 employees sponsor 

retirement plans).  



2008] Slouching Towards a Consumption Tax 127 

 

 

over-inclusive in that it provides benefits for those who would save for their 

own retirement without the tax incentives, and under-inclusive because it 

fails to provide meaningful benefits to many low- and middle-income 

workers.  

 But the system does provide at least some benefits for many millions 

of people, with a coverage rate of approximately 50% of the full-time private 

workforce between ages 25 and 65.
37

 While one can argue whether the 

system is a cup half-full or half-empty, by the numbers we can say that the 

employer-sponsored retirement system provides an important source of some 

retirement income security for many individuals. 

 The societal goal of creating adequate retirement income security 

through tax-benefited savings arrangements requires more than wide 

participation: it also requires, first, that plan assets are invested prudently and 

free of conflicts of interest
38

 so that investment returns are maximized, and 

second, that the savings once created are used to provide retirement 

income.
39 

 
The Internal Revenue Code, and especially Title I of ERISA, 

establish a scheme of fiduciary regulation based on traditional trust law and 

insights from modern portfolio theory, which at least in theory should 

maximize investment return.
40

 In particular, Title I of ERISA provides that 

plan fiduciaries act with appropriate prudence and loyalty to participants 

when they make investments,
41

 and that they diversify plan investments in 

most situations.
42

  

 There are three related concerns to ensure that retirement savings are 

in fact used for retirement: first, that the savings are not invaded prior to 

retirement; second, that the savings are not exhausted prior to death; and 

third, that in appropriate circumstances the savings are available to the 

surviving spouse of the pensioner. 

                                                 
37. Alicia Munnell & Annika Sunden, Private Pensions: Coverage and 

Benefit Trends 1, available at www.outfuture.org/articles/200109270824.pdf. 

38. See, ERISA § 404(a). 

39. See, IRC §§ 401(a)(9), (11) & (13). 

40. See, ERISA §§ 403-410. See, generally, Daniel Fischel and John 

Langbein, ERISA’s Fundamental Contradiction: The Exclusive Benefit Rule, 55 U. 

Chi. L. Rev. 1105 (1988).  

41. ERISA § 404(a)(1), (2). 

42. ERISA § 404(a)(3). Title I of ERISA also includes a series of rules 

prohibiting certain transactions between a plan and parties with a pre-existing direct 

or in some cases indirect relationship with the plan. See, ERISA §§ 406-408, 29 

U.S.C. §§ 1006-1008. In addition to the Title I prohibitions, the Internal Revenue 

Code imposes an excise tax on such transactions. IRC § 4975. For an early, but still 

quite useful, summary of the rules, see, Arthur H. Kroll & Yale D. Tauber, Fiduciary 

Responsibility and Prohibited Transactions Under ERISA, 14 Real. Prop Prob. & 

Trust J. 658 (1979). 



128 Florida Tax Review [Vol. 9:2 

 

 The Internal Revenue Code includes a series of rules that address, 

although imperfectly, each of these concerns.
43

 To limit pre-retirement 

access to retirement assets (i) longstanding treasury regulations prohibit 

distributions from pension (but not profit-sharing) plans prior to retirement, 

death, disability, or separation from service;
44

 and (ii) the Internal Revenue 

Code imposes an excise tax on most plan withdrawals prior to age 59.5 from 

any plan (unless the distribution is transferred to another plan or individual 

retirement account).
45

 To assist individuals to manage their retirement 

savings, pension plans (but not profit-sharing and 401(k) plans) must provide  

that the normal form of benefit is a life annuity (although if the plan permits, 

participants can elect other forms of benefits, including single-sum 

distributions).
46

 To protect the spouse of a pensioner, pension plans must 

provide that the life annuity for a married participant has a survivor benefit 

for the spouse (although the participant, with spousal consent, can elect other 

forms of benefits if permitted by the plan, including a single life annuity).
47

  

 Finally, the Internal Revenue Code has provisions designed to ensure 

that the tax expenditures are limited to the amount needed to create 

retirement security and are not so extravagant that they result in plans that 

function more as general tax shelters and estate-planning devices than 

retirement plans. The most important of these provisions place outer limits 

on the amount that may be contributed annually to a defined contribution 

plan and on the size of the retirement annuity payable from a defined benefit 

plan, for any given individual.
48

 The theory here is that the tax subsidy 

embedded in plan savings should be used to furnish reasonable rather than 

unduly lavish retirement income levels.
49

 In addition, the Internal Revenue 

Code includes minimum distribution rules that require individuals to begin 

drawing down their retirement savings in their retirement in order to limit the 

utility of tax-deferred retirement accounts as a tax-benefited estate planning 

tool.
50

  

 

                                                 
43. For a thorough treatment of the Internal Revenue Code’s distribution 

rules, see, Diane Bennett, et. al., Taxation of Distributions from Qualified Plans (2nd 

ed. 2002). 

44. Treas. Reg. § 1.401(a)-2. 

45. IRC § 72(t). 

46. IRC § 411(a)(7). 

47. IRC § 417. 

48. IRC § 415. See, generally, Norman P. Stein, Simplification and IRC § 

415, 2 Fla. Tax Rev. 69 (1994). 

49. Id. at 4-6. 

50. IRC § 401(a)(9). See, generally, Bennett, supra note 43; William J. 

Turnier, Grayson M.P. McCouch, et.al., Family Wealth Management 651-652 

(2005). 



2008] Slouching Towards a Consumption Tax 129 

 

 

III. THE WHITE HOUSE SAVINGS PROPOSALS 
 

 The White House budget proposal in 2003 included three proposals 

that would have implications for the private, generally employment-based, 

savings arrangements that currently assist Americans in creating income 

security in their retirement. The White House proposals were: (1) the 

exclusion of dividends from income taxation; (2) the creation of two 

personal, non-employment based, tax-advantaged savings vehicles, the 

Retirement Savings Account and the Lifetime Savings Account; and (3) the 

creation of the Employer Retirement Savings Account.  

 

A. Dividend Exclusion 

 

 The White House initial budget proposal included an exclusion for 

dividends. The initial purpose of the exclusion was two-fold: 1) to eliminate 

the so-called double tax on corporate earnings, and 2) to provide incentives 

for corporations to pay dividends, which a White House press release argued  

 

would promote more transparent corporate governance and more honest 

statements of income.
51

 Apparently in response to complaints by the 

corporate managerial class (who might not favor tax incentives to pay 

dividends
52

), the White House proposal was revised so that corporations 

could also avoid double taxation through a shareholder basis adjustment for 

any retained earnings.
53

 Congress did not adopt the White House proposal as 

proposed, but did set an effective maximum tax rate of 15% on both 

dividends and capital gain from the sale of stock.
54

 This provision will sunset 

for tax years beginning after 2008, unless renewed by Congress.  

 

B. New Personal Savings Accounts 

 

 The White House proposed two new personal savings arrangements: 

the retirement savings account, which is designed to assist people to 

accumulate retirement assets, and the lifetime savings account, which would 

allow people to save for any purpose on a tax-favored basis. These accounts 

                                                 
51. See, Bush Proposes Tax Free-Savings Plans, supra note 2. 

52. See, Jennifer Arlen & Deborah M. Weiss, A Political Theory of 

Corporate Taxation, 105 Yale L. J. 325 (1995) (exploration of why corporate 

management does not lobby for integration of corporate and individual income 

taxation). 

53. See, e.g., David Early-Hubelbank, Bush Administration Proposes 

Eliminating Double Tax on Corporate Earnings, Corporate Tax Bulletin (2003), at 

http://pmstax.com/corp/div0301.shtml. 

54. Tax Reform Act of 2003, at § 1(h).  



130 Florida Tax Review [Vol. 9:2 

 

would replace the current individual retirement account, to which individuals 

below a certain income level can currently contribute up to $4,000 per year 

from earned income.
55

  

 The new arrangements would replace the two types of IRA tax 

arrangements, which are sometimes referred to as “traditional” and “Roth” 

IRAS. In a traditional IRA, neither contributions nor investment earnings are 

currently taxed, but withdrawals are subject to tax.
56

 In a Roth IRA, 

contributions are currently taxable, but investments earnings and withdrawals 

are excluded from taxable income.
57

 As with employer plans, early 

withdrawals from regular IRAs (prior to the year in which the IRA owner 

attains age 59.5) are generally subject to a 10% penalty tax, and distributions 

must commence on a ratable basis beginning no later than the April 1 after 

the IRA owner attains age 70.5.
58

  

 Both the RSA and LSA differ in important ways from IRAs. Three 

key differences between IRAs and both RSAs and LSAs are: (i) there would 

be no maximum income limits for contributions to the new savings 

arrangements; (ii) contributions to the new savings arrangements would have 

to be made on a Roth (that is, after-tax) basis, with account distributions 

excludable from income; and (iii) there would be no minimum distribution 

rules requiring that distributions commence during an owner’s lifetime.
59

 

Other differences between IRAS and the new savings arrangements are 

discussed below: 

 

 1. Retirement Savings Accounts 

 

 Under the 2003 budget proposal, an individual would have been able 

to contribute up to $7,500 from earned income to an RSA.
60

 In subsequent 

budgets, this amount was reduced to $5,000, perhaps in a nod to deficit 

control, but also perhaps in capitulation to some usual White House allies 

who initially opposed the proposals as a threat to employer provided health 

                                                 
55. IRC § 408, IRC § 219. The $4,000 limit increases to $5,000 in 2008 and 

thereafter will increase to reflect increases in the cost of living. Id. 

56. IRC § 408. 

57. IRC § 408A. 

58. IRC § 408(b), (c). Roth IRAs are also subject to the excise tax on early 

distributions, but an owner of a Roth IRA is not taxed to the extent distributions 

reflect simply already taxed contributions. IRC § 408A(d)(4). Moreover,

distributions from Roth IRAs are not subject to the minimum distributions rules. IRC 

§ 408A(c)(5).  

59. See, General Explanation (2003), supra note 1, at ¶ 119- 20. 

60. Id. at 119.  



2008] Slouching Towards a Consumption Tax 131 

 

 

care.
61

 An individual would also have been able to contribute up to the 

maximum amount for a non-working spouse.
62

 Distributions of investment 

income would have been subject to a 10% excise tax if made before the 

RSA-owner has attained age 58.
63

 Roth IRAs would have been automatically 

converted into RSAs.
64

 Traditional IRAs could be converted to RSAs on an 

elective basis by paying income tax on the converted amount and payment of 

the tax could be spread over a four-year period.
65

  

 

 2. Lifetime Savings Accounts 

 

 In the 2003 budget, an individual would have been able to contribute 

$7,500 annually to an LSA; there was no requirement that the individual 

have earned income to make such contribution.
66

 Subsequent budgets have 

reduced the figure to $5,000.
67

 Moreover, an individual can contribute to any 

other person’s LSA, although contributions by or on behalf of any particular 

individual could not in the aggregate exceed the maximum amount.
68

 Thus, a 

married couple with two children could have contributed $30,000 annually to 

LSAs for family members (if the limit were, as initially proposed, $7,500 per 

individual).  

 There were no restrictions or excise taxes imposed on withdrawals 

from LSAs. It has thus been suggested by some financial columnists that 

should the Bush proposals be adopted, individuals would generally want to 

contribute the maximum to an LSA before contributing to an RSA, since pre-

age-58 withdrawals from an RSA could result in imposition of excise 

taxes.
69

 Although an LSA could be used to save for any purpose, the White 

House proposal did not eliminate special-purpose savings vehicles currently 

                                                 
61. United States Department of Treasury, “Bush Again Pushes for RSA, 

LSA and ERSA in Fiscal 2006 Budget,” at www.pressreleases.newspap.com/pr/ 

20052/pr206421.html. See, Karen C. Burke and Grayson M. P. McCouch, Lipstick, 

Light Beer, and Back-Loaded Savings Accounts, 25 Va. Tax Rev. 1101, 1123 

(2006). 

62. General Explanation (2003), supra note 1, at 120.  

63. Id. 

64. Id. 

65. Id. 

66. Id. at 119-20.  

67. United States Department of Treasury, “Bush Again Pushes for RSA, 

LSA and ERSA in Fiscal 2006 Budget,” available at 

www.pressreleases.newspap.com/pr/20052/pr206421.html. 

68. General Explanation (2003), supra note 1, at 120. 

69. See, e.g., Christine Dugas, et. al., How Bush’s Retirement System 

Overhaul Hits Home, USA Today, Feb. 23, 2003, at 2B.  



132 Florida Tax Review [Vol. 9:2 

 

available under the Internal Revenue Code (such as Coverdale education 

accounts, qualified state tuition plans, and medical savings accounts).
70

  

 

C. Employer Retirement Savings Accounts 

 

 One of the potential virtues of defined contribution plans (compared 

to defined benefit plans) is their relative simplicity of form.
71

 The regulatory 

structure that governs such plans, however, sometimes perverts this 

simplicity of form into extreme complexity, in two ways: first, by providing 

extraordinary flexibility through which firms can design plans that minimize 

participation and benefits for rank-and-file employees, and second by 

imposing administrative and compliance complexity as the toll charge for 

such flexibility.
72

 These two types of complexity (although related) can be 

conceptualized as design complexity, on the one hand, and administrative 

complexity on the other.  

 To illustrate, consider an employer that wants to minimize benefits 

for rank-and-file employees. The employer could design its plan to exclude 

(in some cases) more than 50% of its rank-and-file employees while covering 

all of its highest paid employees, and could provide benefits that are 

significantly higher (as a percentage of pay) for the latter than for the 

former.
73

 The tools an employer might use to so design a plan could include 

cross-testing,
74

 a 401(k) structure,
75

 integration with Social Security,
76

 

                                                 
70. General Explanation (2003), supra note 1, at 120. See, IRC § 529 

(qualified state tuition plans); IRC § 530 (Coverdale savings accounts); IRC § 223 

(health savings accounts). There have been a number of articles highly critical of 

health savings accounts, and particularly the tax treatment of such accounts. See, 

Leonard E. Burman, New Health Care Tax Proposals: Costly and 

Counterproductive, 110 Tax Notes 779 (Feb. 9 2006); Amy B. Monahan, The 

Promise and Peril of Ownership Society Health Care Policy, 80 Tulane L. Rev. 777 

(2006); Norman Stein, The HSA: Health Savings Accounts or Health (Policy) 

Sabotaged Again, 64 NYU Institute on Federal Taxation, Employee Benefits and 

Executive Compensation (2006). In 2006, the White House proposed a dramatic 

expansion of the tax benefits available to HSAs. See, Berman, infra. 

71. See, generally, Edward A. Zelinsky, The Defined Contribution 

Paradigm., 114 Yale L. J. 451 (2004). 

72. Norman Stein & Peter Orszag, Cross-Tested Defined Contribution 

Plans: A Response to Professor Zelinsky, 49 Buff. L. Rev. 629 (2001) (discussing 

permissible means of discrimination under Internal Revenue Code). 

73. Id. at 634-39; IRC § 410(b)(3).  

74. Id. 

75. IRC § 401(k). Section 401(k) permits employees to decide whether to 

participate, and how large a percentage of their compensation to contribute, with 

generally less exacting nondiscrimination requirements. Id. 



2008] Slouching Towards a Consumption Tax 133 

 

 

providing small benefits for its lowest-paid workers and large benefits for 

higher-paid rank-and-file workers,
77

 use of a complex average benefits test,
78

 

a separate-line-of-business test,
79

 catch-up contributions,
80

 the adoption of 

different plans for different employees,
81

 and the careful fashioning of a 

vesting schedule.
82

 These tools all impose complexity in the design and 

administration of a plan.  

 Some have argued that the law’s complexity discourages some firms 

from sponsoring plans,
83

 but this is not precisely accurate, since an employer 

can mostly or entirely avoid the complexity by adopting a plan covering all 

employees under a uniform and simple benefit formula. It would be more 

accurate to say that some firms are discouraged from adopting defined 

contribution plans because design and compliance complexity makes it 

expensive to set up plans that aggressively favor highly-paid employees in 

coverage and benefits.  

 In some sense, the holy grail of private pension reform would be to 

quantify the maximum tolerable amount of benefit and coverage disparity 

between highly and non-highly-paid employees and then put forth a single 

plan template that accommodates, with minimum complexity, up to but no 

more than this amount of disparity.
84

 This is, I think, the idea of the ERSA, 

                                                                                                                   
76. IRC § 401(l) (describing Social Security integration requirements); see, 

also, Stein and Orszag, supra note 72, at 635; see, generally Nancy J. Altman, 

Rethinking Retirement Income Policies: Nondiscrimination, Integration, and the 

Quest for Worker Security, 42 Tax L. Rev. 435 (1987) (providing excellent 

discussion on Social Security integration).  

77. See, generally Michael W. Melton, Making the Nondiscrimination 

Rules of Tax-Qualified Retirement Plans More Effective, 71 B.U.L.Rev. 47 (1991). 

78. IRC § 410(b)(2)(A). See, Stein & Orszag, supra note 71, at 636-37. 

79. See, IRC § 414®). 

80. See, IRC 402(g)(1)(c). This section permits a section 401(k) plan to 

accept an additional $5,000 contribution from a participant who is at least age 50, 

without nondiscrimination testing of any sort. Id. In many cases, the employees over 

50 who will be able to afford this contribution will be higher paid employees. 

81. See, Treas. Reg. § 1.401(a)(4)-3. See, generally Melton, supra note 77; 

Stein & Orszag, supra note 71, at 637-38. 

82. See, Stein & Orszag, supra note 72, at 638.  

83. See, e.g., Joseph Bankman, Tax Policy and Retirement Income: Are 

Pension Plan Anti-Discrimination Provisions Desirable?, 55 U. Chi. L. Rev. 790, 

806 (1988).  

84. There is at least one problem with this approach: under current law, a 

firm will incur various professional and compliance costs if it wishes to design a 

plan that aggressively favors higher paid employees; the methodologies described in 

the text can act as a toll charge for such plans. If the statute is modified to make it 

simpler to minimize coverage and benefits for rank-and-file workers, it is plausible 

that some firms, no longer facing a toll charge, will reduce coverage for such 



134 Florida Tax Review [Vol. 9:2 

 

which would create a simplified regulatory structure for all defined 

contribution plans. It would do so, at least in the Administration’s 2003 

proposals, by eliminating several complexity-creating tools that under 

current law are used to create high levels of disparity between highly and 

non-highly compensated employees, i.e., integration, the average benefits 

test, and cross-testing.
85

 These changes would reduce the amount of disparity 

between the two groups of employees that employers can effect today 

through plan design.  

 The rules governing the ERSA, in tacit recognition of the loss in plan 

sponsor ability to use these tools to favor higher-paid employees, would 

increase the permissible degree of disparity between the two groups for 

elective contributions and eliminate the “top-heavy rules” – special rules 

requiring minimum contributions and more accelerated vesting for 

participants in plans whose benefits are primarily allocable to certain “key” 

employees.
86

  

 

IV. CRITIQUE OF WHITE HOUSE PRIVATE SAVINGS PROPOSALS 
 

 The White House private savings initiative (the RSA, LSA, and 

dividend exclusion) would strongly push the country towards a consumption 

tax base,
87

 which has a political constituency primarily among those 

                                                                                                                   
employees. It is also plausible that inclusion of such express statutory provisions will 

result in benefits consultants designing template plans, which will act as the default 

for most employers, even those who might have been inclined to provide more 

generous benefits for rank-and-file employees if that had been presented as an 

option. 

85. In the most recent incarnation of the ERSA, however, cross-testing 

would be permitted. See, General Explanation of the Administration’s Fiscal Year 

2006 Revenue Proposals Dept. of Treas., at 129 (Feb. 2008). 

86. IRC § 416. The accelerated vesting provisions of Section 416 are 

moving toward the scrapheap of irrelevancy, as Congress improves the overall 

minimum vesting rules, which currently require all defined contribution plans to 

satisfy the same vesting rules as section 416. IRC § 411(a)(2)(B). In addition, cash 

balance plans are now subject to 3-year cliff vesting, a more demanding standard 

than imposed on plans because they are top-heavy. IRC § 411(a)(13). 

87. There is a rich and growing literature on a consumption tax. The classic 

articles in the legal academic literature include William D. Andrews, A 

Consumption-Type or Cash Flow Personal Income Tax, 87 Harv. L. Rev. 1113 

(1974); Alvin C. Warren, Jr., Fairness and a Consumption-Type or Cash Flow 

Personal Income Tax, 88 Harv. L. Rev. 931 (1975); William D. Andrews, Fairness 

and the Personal Income Tax: A Reply to Professor Warren, 88 Harv. L. Rev. 947 

(1975); Alvin Warren, Would a Consumption Tax Be Fairer than an Income Tax?, 

89 Yale L.J. (1980); Barbara Fried, Fairness and the Consumption Tax, 44 Stan. L. 

Rev. 961 (1992). As I will discuss briefly in the text of the article, the White House 



2008] Slouching Towards a Consumption Tax 135 

 

 

concerned more with national economic growth than with distribution and 

among those who believe that tax equity is judged better on a lifetime than 

an annual basis.
88

 The proposals might, as their advocates predict, increase 

aggregate national savings and might improve tax equity by certain metrics 

of fairness. In doing so, however, they would likely break the back of the 

employer-sponsored pension system, on which approximately half of 

working Americans rely for retirement savings.
89

 The proposals would do 

this by reducing the number of employer-sponsored plans and by likely 

reducing both the contribution levels and investment return for rank-and-file 

employees who participate in the plans that remain. Moreover, there is 

reason to believe that the private savings initiatives would not adequately 

replace the retirement income security that most working Americans would 

lose from employer-sponsored plans.  

 This section of the paper provides the analysis that undergirds this 

pessimism about the effects of the White House proposals on retirement 

income security. The section considers (i) possible effects of the private 

savings proposals on employer sponsorship and employee use of 

employment-based retirement plans; (ii) the prospects that middle and lower  

income workers will use the new savings initiatives to create new retirement 

savings to compensate for the diminishment of the importance of employer-

sponsored retirement plans; (iii) the effect of the savings initiative on 

investment return on retirement savings for middle and lower income 

Americans; and (iv) some tax and budget consequences of the savings 

initiative. 

 

A. Savings Initiative and Employment-Based Retirement Plans 

 

 I have already suggested that scholars and policymakers believe that 

firms generally sponsor plans in response to preferences of firm owners and 

                                                                                                                   
savings proposals move us toward a consumption tax in an incomplete, and ill 

conceived, manner. 

88. See, Joseph Bankman & Barbara Fried, Winners and Losers in the Shift 

to a Consumption Tax, 86 Geo. L. J. 539 (1998); Michael J. Graetz & Deborah H. 

Schenk, Federal Income Taxation, Principles and policies 36-39 (4th ed. 2002). 

89. See, supra note 2. At least one group (The American Society of Pension 

Professionals and Actuaries) that initially opposed the RSA and LSA proposals has, 

in light of a reduction in contribution limits and modifications to the ERSA proposal, 

now supports the White House savings initiative. Compare ASPA, Statement for the 

Record, Committee on Ways and Means, Hearing on the Presidents’ Economic 

Growth Proposals, Mar. 6, 2003, http://aspa.org/archive/gac/2003/030503-eco-

propsal.htm (Last visited Jun. 29, 2007) with Administration Announces Revised 

Savings Proposals – Changes Made to Address ASPA’s Concerns available at 

http://www.aspa.org/archivepages/gac/2004/2004-02-02-savingsproposals.htm. (Last 

visited Jun. 29, 2007). 



136 Florida Tax Review [Vol. 9:2 

 

other highly-paid employees for tax-advantaged deferred compensation, with 

the nondiscrimination rules then requiring that less well-paid employees also 

be provided with some benefits.
90

 To appreciate the significance of the tax 

benefits to highly paid individuals in an employer’s decision to establish and 

maintain a plan, it is useful to think about the question first in a world 

without an income tax. What are the considerations that would go into plan 

sponsorship in such a world? 

 I begin with the observation that deferred compensation plans 

predated the income tax (and thus the tax benefits that our income tax regime 

confers on qualified plans), so there are certainly non-tax reasons for firms to 

sponsor pension plans. One of the early motivations for pension plans was to 

provide for the orderly superannuation of industrial and construction 

workers, which was thought not only to ensure the retirement of older 

employees with declining productivity, but also to raise the morale of active 

workers and thereby increase their productivity.
91

 Some also saw pension 

plans as satisfying a moral obligation of the employer to its elderly workers 

who were no longer capable of working.
92

 Historically, employers have also 

been able to use pension plans to create incentives for long job tenures, 

which can be valuable in industries that invest heavily in their human 

capital.
93

 In addition, some employees may value some degree of deferred 

compensation at more than its cost to the employer even apart from its tax 

benefits: employees may value the forced savings aspect of deferred 

compensation; may enjoy relatively high rates of return due to economies of 

scale and the professional investment management that plans offer; and, in 

defined benefit plans, may appreciate both the benefit guarantees and the 

pooling of mortality risk.  

 But employers also incur significant costs when they sponsor 

retirement plans. Sponsorship of plans imposes design, administrative, and 

compliance costs on the sponsoring firm. Employers
94

 also are subject to 

potential liability to participants if they violate ERISA’s complex fiduciary 

and other rules.
95

 Defined benefit plan sponsorship requires employers to 

guarantee investment performance, assume mortality risk, and absorb the 

                                                 
90. See, text accompanying supra notes 29-36. 

91. See, generally, Arthur Cloud, Pensions in Modern Industry (1930); 

Murray Latimer, Industrial Pension Systems (1932). 

92. See, Lee Aquier, Old Age Dependency in the United States (1912). 

93. See, John Langbein & Bruce Wolk, Pension and Employee Benefit Law 

29-32 (3rd ed. 2000). 

94. Employees who are assigned discretionary duties in administering a 

plan can also be held personally liable under ERISA. See, ERISA § 3(21)(defining 

fiduciary), and 502(a)(3) and §§ 502(a)(3), 409 (creating personal liability for 

fiduciaries in certain situations). 

95. See, ERISA §§ 502(a)(2), (a)(3), 29 U.S.C. §§ 1132(a)(2), (a)(3). 



2008] Slouching Towards a Consumption Tax 137 

 

 

volatility of periodic funding obligations and the resulting need to sometimes 

make substantial contributions to plans in lean business years.
96

 And for 

some employers, the Internal Revenue Code’s nondiscrimination rules, 

which require plans to provide benefits to some non-highly compensated 

employees, impose perhaps the steepest cost of plan sponsorship.
97

 Some 

such employees benefit little from the qualified plan tax deferral and some 

(but certainly not all) will have a strong preference for immediate 

compensation.
98

 Thus, some employees for whom the employer provides 

deferred compensation will not value the compensation at its cost to the 

employer, which will increase the cost of total compensation for those 

employees.  

 In addition, the non-tax benefits of employer-sponsored retirement 

plans have become somewhat attenuated over the last two decades, 

particularly with respect to a firm’s ability to design a plan to advance human 

resource strategies. ERISA’s vesting rules have limited the employer’s 

ability to use a plan to encourage long job tenures.
99

 Moreover, 

demographers predict that labor markets will begin to tighten as baby 

boomers reach retirement age
100

 and thus some firms may have a reduced 

interest in retirement plans as a strategy to encourage older employees to 

retire.  

                                                 
96. IRC §§ 412, 430 (funding requirements for defined benefit plans). 

97. See, IRC § 401(a)(4); Treas. Reg. § 1.401(a)(4).  

98. Alicia Munnell, The Economics of Private Pensions (1982).  

99. IRC § 411(a)(requiring that vesting begin no later than five years after 

employment); IRC § 416(b) (requiring vesting to begin no later than three years for 

“top-heavy” plans). Moreover, the Pension Protection Act now requires that all 

defined contribution plans conform to the rules previously applicable to top-heavy 

plans, IRC § 411(a)(2)(B), and requires cash balance and other hybrid defined 

benefit plans to provide 100% vesting after three years, IRC § 411(a)(13). 

100. Michael W. Wyand, Aging Issues: Retirement of ‘Baby Boom’ 

Generation Poses Potential Problems for Employers, GAO says bna’s Pension & 

Benefits Rep. 2919 (2001) citing Older Workers: 

Demographic Trends Pose Challenges for Employers and Workers (Report 02-85) 

(2001)), Peter Coy & Done Brady, Old. Smart. Productive.; Surprise! The Graying 

of the Workforce is Better News Than You Think, Business Week 78 (Jun. 27, 

2005), Stacy Polos & Dimitri S. Nightingale, Employment and Training Policy 

Implications of the Aging Baby Boom Generation, Urban Institute, Jun. 1, 1997, 

available at http://www.urban.org/Template.cfm?NavMenuID=24&template=/ 

Taggedcontent/ViewPublication.cfm&PublicationID=6558. There has also been 

some interest in phased retirement, where older employees would reduce their hours 

and replace the resulting lost income by beginning to receive benefits from employer 

sponsored retirement plans. See, e.g., Phased Retirement, available at 

http://www.workforce.com/section/02/feature/23/47/31/. 



138 Florida Tax Review [Vol. 9:2 

 

 It should also be said that non-tax explanations for whatever 

employee preference for deferred compensation exists, perhaps never strong, 

are weaker now than in 1974, when Congress enacted ERISA. The wide 

availability of mutual funds has made it possible for small investors to 

benefit from professional money management without the intervention of an 

employer-sponsored retirement plan. Moreover, the market decline of the 

early 2000s may have made some employees wary of defined contribution 

plans,
101

 and defined benefit plans, which shield employees from market and 

mortality risk, are themselves in decline, in large measure because of their 

costs to the firm, especially in industries with aging workforces.
102

 And in an 

increasingly mobile workforce, individuals may be less inclined to want what 

might be a relatively short-term employer to be entrusted with custody of 

their retirement savings.  

 This brings us to the question of whether employer-sponsored 

retirement programs would exist in the absence of the tax deferral that those 

plans currently effect. At least for small firms, the conventional 

understanding – that firms generally establish plans to satisfy a tax-driven 

preference for such plans by firm owners and highly-paid employees – seems 

accurate. Small firms are not likely to derive meaningful non-tax benefits 

from retirement-plan sponsorship (other than in the recruitment and retention 

of relatively highly paid employees who value the tax savings), and the fixed 

costs of plan sponsorship cannot be spread over a large workforce. Without 

the tax benefits realized by firm owners and highly-paid employees, small 

firm sponsorship of retirement plans would certainly be less common than it 

now is, perhaps far less common.  

 For large firms, the answer is less clear. Such firms are more likely 

than small firms to be able to use plans to advance human resource goals and 

have more potential plan participants among whom to spread the fixed costs 

of plan sponsorship. Moreover, we inhabit an existing world, not a world of 

pure theory, and in that existing world almost all large firms sponsor 

retirement plans. Firm culture and experience with retirement plans may 

result in increased employee appreciation for such plans. And many large 

                                                 
101. See, Fidelity’s Study Defines New Directions for Vendors and 

Sponsors, DC Plan Investing (Nov. 9, 2004) (overall participation rates for 401(k) 

plans declined by 2% between 2002 and 2003); Kathy Chu, Employee 401(k) 

Participation Slips – only 76% of Eligable Workers Utilized Plans Loast Year, Down 

From 80% in 2002. Wall St. J. (Oct. 6. 2004), at C15. (participation declined by 

4%). 

102. Retirement Policy: Decline in Use of Defined Benefit Plans May be 

Due More to Workers Than Firms, 32 BNA’s Pension & Benefits Reporter 1114 

(2005), available on Westlaw at 32 BPR 114 (citing Stephanie Aaronson & Julia 

Cioronado, Are Firms or Workers Behind the Shift Away from DB Pension Plans? 

Fin. and Econ. Discussion Series, Division of Research & Statistics and Monetary 

Affairs, Fed. Reserve Board, Working Paper No. 205-17, 2005). 



2008] Slouching Towards a Consumption Tax 139 

 

 

firms have invested resources in educating their employees about the value 

of retirement benefits.
103

 In addition, it can be difficult to take away a benefit 

that employees have been conditioned to expect. Thus, while we can 

speculate about whether in the absence of tax benefits large firms would 

decide to sponsor retirement plans if they were not already doing so, it does 

not seem likely that such firms would immediately extinguish such plans if 

the tax advantages of such plans were suddenly reduced or even withdrawn. 

Still, it should be said that if the tax benefits were withdrawn or reduced, 

there would almost certainly be diminished interest in retirement plans 

among large firms, at least in the long run. 

 The role of collective bargaining should also be noted: unions often 

bargain for retirement benefits, both for single-employer plans in large firms 

and for multi-employer plans in some industries with numerous firms and 

high rates of employee mobility between those firms.
104

 Tax benefits for 

their membership may be one reason why unions negotiate for such plans, 

but there are other reasons as well.
105

 Unions represent employees at all 

points on a demographic line through the workforce and older employees 

nearing retirement can be expected to place a high value on retirement 

income. Unions may also be motivated by the long-term welfare of their 

membership, which would include retirement income security, even when 

younger workers may not appreciate the value of retirement benefits. In 

addition, unions can, and do, educate their membership about the value of 

retirement income, which may result in some younger workers attaching 

value to retirement benefits.
106

 I would expect, then, that many unions would 

continue to negotiate for retirement benefits even if the tax advantages 

embedded in those benefits were reduced. 

 But overall, the tax benefits of qualified plans are an important 

driver of firm willingness to sponsor retirement plans, particularly for small 

firms.
107

 Equally important, under the nondiscrimination rules, the benefits 

of highly compensated employees are constrained by the benefit levels and 

participation rates of rank-and-file employees in the plan. As a result of this,  

some firms, in order to provide higher levels of benefits to highly-paid 

employees, cover more rank-and-file employees and/or set higher benefit 

levels for them than they otherwise would. 

                                                 
103. See, Hewitt Survey Reveals New Employer Trends in Retirement, 

Business Wire, Jan. 18, 2005, http://www.businesswire.com. 

104. William C. Greenough & Francis P. King, Pension Plans and Public 

Policy 44-47 (1976). 

105. Id. 

106. See, U.S. Dept. of Labor Advisory Council on Employee Welfare and 

Employee Ret. Plans, Report of Working Group on Planning for Retirement (2001) 

(noting that labor organizations educate their members about retirement planning). 

107. John Langbein and Bruce Wolk, supra note 93, at 30. 



140 Florida Tax Review [Vol. 9:2 

 

 If tax benefits for highly paid employees participating in plans were 

eliminated or reduced, we would expect, then, to see two effects on qualified 

retirement plans: fewer firms would sponsor plans (particularly smaller 

firms), and benefit levels for moderate- and lower-income employees would 

be reduced in some of those plans that remained.  

 There are two ways that the tax benefits of highly paid plan 

participants can be reduced. One way is direct: either increase the effective 

tax rate on investment return from the retirement plan, as was the case in a 

short-lived 15% excise tax on aggregate plan distributions to the extent they 

exceeded $150,000 annually,
108

 or decrease the limits on the benefits that 

can be provided to higher-paid individuals. The other is indirect: decrease the 

effective tax rate on investment return outside the retirement plan, which 

would reduce the comparative benefit of investing inside of an employer-

sponsored retirement plan. The White House savings proposals would have 

the latter effect: the 2003 proposal would eliminate income tax on dividends 

or stock appreciation caused by retained earnings and would also eliminate 

tax on any type of investment held in an RSA or LSA. The concern, then, is 

that highly paid employees would no longer have incentive to participate in 

an employer-sponsored retirement plan since they could realize similar after-

tax returns by investing in stock or by holding other types of investments in 

an RSA or LSA. 

 The Bush proposals would not, however, entirely eliminate the tax 

benefits of investing through the medium of a qualified plan, for at least two 

reasons. First, and most important, to the extent appreciation in stock value 

resulted from factors other than retained earnings, gain on the stock would be 

subject to capital gains taxation on stock held for more than one year. 

Second, even if gain on investment were entirely exempted from taxation, 

the pricing of stock should under normal market conditions adjust upward to 

reflect the tax-advantaged treatment of dividend distributions.
109

 In other 

words, returns on stock, even though nominally exempt from tax, can be 

expected to reflect an implicit tax burden. Investment assets whose returns 

were not broadly exempted from income taxation should thus pay a higher 

before-tax rate of return than the tax-free return on stock, to account for the 

explicit tax to which such returns would be subject. Since holding such assets 

in a qualified plan would defer the tax until benefit distribution, qualified 

plans could still carry a tax benefit by investing in assets generating taxable 

returns.  

                                                 
108. IRC 4980A (1996). See, generally, Bruce Wolk, The New Excise and 

Estate Taxes on Excess Retirement Plan Distributions and Accumulations, 39 U. Fla. 

L. Rev. 987 (1987).  

109. Daniel J. Mitchell, et. al., Pathway to Economic Growth and Tax 

Reform: Eliminating the Double Tax on Dividends, The Heritage Foundation, Mar. 

28, 2003, available at http://www.heritage.org/Research/Taxes/bg1640.cfm. 11111* 



2008] Slouching Towards a Consumption Tax 141 

 

 

 Thus, if the White House’s proposal were adopted and included only 

the elimination of the double tax on corporate earnings, there would likely 

continue to be some, albeit diminished, demand among high-income 

taxpayers to participate in qualified plans.
110

   

 But the Bush proposals also would create the RSA and LSA, tax-

exempt private savings vehicles. Thus, highly paid individuals could use 

RSAs and LSAs until they reached the limits of those accounts. The limits of 

the accounts in the 2003 White House proposal were nominally $7,500 per 

individual (but in subsequent proposals less), but contributions could be 

made to an RSA on behalf of a non-working spouse and contributions could 

be made to an LSA on behalf of any individual. Thus, a single person could 

have made contributions of $15,000 annually to these accounts as initially 

conceived; married couples could have made contributions of $30,000 

annually; and a married couple with two children could have made 

contributions of $45,000 annually. 

 These contributions are, however, after tax; the contribution limits 

are thus higher than nominally identical contribution limits would be for the 

more familiar pre-tax contributions to an IRA or 401(k) plan. For an 

individual with a 36% marginal tax rate, a $15,000 before-tax contribution is 

the equivalent of an after-tax contribution of $23,437.50.
111

 Another way of 

conceptualizing the difference between after-tax and before-tax contributions 

is by looking at the cost of the contribution: an after-tax contribution of 

$15,000 costs the taxpayer $15,000, while a before-tax contribution costs the 

taxpayer $15,000 plus the $8,437.50 in tax that the taxpayer must pay in the 

year of contribution. Thus, the actual contribution limit for a single taxpayer 

(with a 36% marginal tax rate) is equivalent to a $23,437.50 before-tax 

contribution; a couple’s contribution would be the equivalent of a $46,875 

before-tax contribution; and a two-child couple’s contribution would be the 

equivalent of a $70,312.50 before-tax contribution. The individual would 

have a tax incentive to contribute to an employer plan only to the extent she 

desired to invest in taxable assets in excess of these amounts. 

                                                 
110. To some extent, this would depend on how large the spread is between 

the rate of return on stock and the before-tax rate of return on assets generating 

taxable returns. As long as there is some spread, there will be some tax benefit to 

investing in such assets through the intermediary of a tax exempt vehicle, such as a 

qualified retirement plan. It is beyond the scope of this article to predict the amount 

of the spread between taxable and nontaxable assets and whether the resulting tax 

benefits to highly compensated individuals would sufficiently offset the costs of plan 

sponsorship to make plan sponsorship attractive to the employer. 

111. This is an illustration of the equivalency of exemption of capital from 

taxation (through deduction) or exclusion from income) and exemption of taxable 

income on capital from income taxation. See, generally, E. Cary Brown, Business-

Income Taxation and Investment Incentives, in Income, Employment and Public 

Policy, Essays in Honor of Alvin H. Hansen (1948). 



142 Florida Tax Review [Vol. 9:2 

 

 The universe of individuals able to invest beyond the indicated levels 

is certainly not large. Thus, to the extent that plan sponsorship, and the 

generosity of coverage and benefits in plans, hinges on the plan’s utility to 

highly compensated individuals, the result of the Bush initiatives would 

likely be fewer and less generous pension plans. As suggested earlier, 

smaller firms are likely to be particularly sensitive to reductions in tax 

preferences for highly paid individuals.
112

 I offer two examples that may be 

typical of some small-firm responses to the Bush savings proposals, if 

enacted.  

 Assume a closely held corporation employing the owner and five 

other employees. The owner’s compensation from the business last year was 

approximately $100,000, and she expects to have approximately the same 

amount of compensation this year. The owner is 55 and last year established 

a safe-harbor 401(k) plan,
113

 to which she contributed $10,000 on behalf of 

herself. The safe-harbor plan uses a matching formula, which requires a 

100% match on the first 3% of compensation, and 50% of the next 2% of 

compensation.   

 Three of the five other employees are eligible to participate in the 

401(k) plan and two of them do so. The two employees earn $25,000 each 

and elect to defer $1,250 each to the 401(k) plan. The sole proprietorship 

must, under the safe harbor matching formula, contribute an additional 

$1,000 for each of the electing employees. Moreover, if we also assume that 

the employer incurs $500 in expenses to maintain the plan, the employer will 

incur $2,500 in direct costs of plan sponsorship.
114

 In addition, we can 

assume that the plan also imposes some indirect costs, including the 

expenditure of some of her time. Finally, let us assume that the owner, who 

has little savings outside the plan, is concerned about her ability to access 

savings inside the plan. 

 If the original Bush savings proposals had been enacted, the owner 

could use the $10,000 contribution to make a $7,200 after-tax contribution to 

an LSA. The owner could now terminate the 401(k) plan, saving the $2,500 

additional plan costs. In addition, since an LSA does not penalize premature 

distributions, the owner will have unrestricted access to the LSA if she needs 

the money before retirement. The consequence of the owner’s actions is that 

the two employees who participated in the 401(k) plan will no longer receive 

matching contributions.
115

 

                                                 
112. See, text accompanying supra note 91. 

113. IRC § 401(k)(12).  

114. This may not be strictly accurate, for the employer may through the 

matching contributions attract more productive workers, but this is speculative.  

115. Note that the matching contribution not only added to the employees 

own savings, but also may have been an important incentive for the employees to 



2008] Slouching Towards a Consumption Tax 143 

 

 

 The second example is a doctor’s office, which employs the doctor 

and eight other employees. The doctor, who is age 50, earns approximately 

$300,000 annually and has been saving $45,000 annually in his retirement 

plan. He has a non-working spouse and two children. In addition to his 

retirement savings, he saves $30,000 each year outside the plan, which he 

divides equally between stocks (mostly through a mutual fund), municipal 

bonds, and real estate. He also contributes $5,000 to section 529 plans for 

each of his children. 

 His practice currently sponsors two plans: a safe harbor “match” 

401(k) plan and an age-weighted profit-sharing plan.
116

 He elects to 

contribute $20,000 to the 401(k) plan (which included a $5,000 “catch-up” 

contribution) and receives a matching contribution of $8,000, for a total 

contribution of $28,000. His practice also contributes $17,000 to his account 

in an age-weighted profit-sharing plan.  

 Four of the employees elect to participate in the 401(k) plan and six 

employees participate in the age-weighted profit sharing plan. The doctor 

contributes $8,000 in matching contributions to the 401(k) plan and $6,000 

to the age-weighted profit sharing plan for the employees. In addition, the 

doctor pays $500 for administrative expenses for the 401(k) plan and $750 

for the age-weighted profit-sharing plan.  

 If the White House’s savings proposals are enacted, the doctor could 

set up RSAs and LSAs for himself and his spouse, in which he can deposit 

$30,000. Let us assume, however, that the doctor will want to move $10,000 

of his non-plan investments into the LSAs, reducing the amount of qualified 

plan savings that he can shift to the RSA/LSAs for himself and his spouse to 

$20,000. But since the RSA/LSAs are made with after-tax contributions, the 

$20,000 before-tax contribution limit is the equivalent of a $31,250 after-tax 

contribution (the basis on which his contributions were made to the qualified 

plans).
117

 The doctor can now contribute $31,500 to the LSA/RSAs that 

would have been contributed to the retirement plans, which accounts for all 

but $13,500 of what the doctor would, but for the RSA/LSA vehicles, have 

contributed to the practice’s qualified plans. The doctor has a choice of what 

to do with the $13,500. First, the dividend exclusion, and the shift of some 

non-plan savings to the LSAs, will have increased the future effective after-

tax rate of return on those investments, which might result in the doctor 

deciding to reduce his aggregate savings. A second choice would be to shift 

some of the $13,500 to LSAs for his two children, if that fits with his family 

wealth-management strategies. This choice would allow the doctor to 

                                                                                                                   
save at all, since they would not have received the matching contributions if they had 

not contributed to the 401(k) plan themselves. 

116. For details on age-weighted profit-sharing plans, see, Orszag & Stein, 

supra note 72.  

117. See, supra note 109, and accompanying text. 



144 Florida Tax Review [Vol. 9:2 

 

achieve the same deferral that he accomplished through qualified plans 

without the use of any qualified plan. If the doctor makes this choice, his 

employees will lose the $14,000 that the doctor had been contributing for 

them and the advantages of a workplace savings program. 

 But suppose that the doctor neither wishes to reduce his savings rate 

nor establish LSAs for his children; in that case, he will still have $13,500 he 

will want to defer to a qualified plan. One option would be for him to 

establish a SIMPLE 401(k) plan, which requires a matching contribution 

equal to the first 3% of compensation.
118

 The doctor could effectively 

contribute the entire $13,500 to the plan and his contributions on behalf of 

his other employees to this plan would be $4,800. He would not need to 

make any contributions to the age-weighted profit-sharing plan, which he 

could freeze or terminate. The other employees would thus lose 66% of the 

employer-provided benefits that would have otherwise been made to their 

accounts. Moreover, some of the doctor’s employees might reduce their 

elective contributions in response to the reduction in the employer matching 

contributions.  

 This article does not make a claim that all firms, or even all small 

firms, would discontinue plan sponsorship or reduce employer-funded 

benefits for rank-and-file employees in response to the White House savings 

proposals. Nor does it argue that firms that do not currently have retirement 

plans would never adopt them if the Bush savings proposals are enacted. As 

the article earlier observed, there is an array of reasons that firms sponsor 

plans, of which the tax incentive for highly compensated employees is only 

one.
119

 But for many firms it is the primary reason, and the Bush proposals, 

by sapping out of the tax law much of the vitality of the tax incentives for 

employer-sponsored plans, could be expected to reduce both their numbers 

and their generosity to ordinary working Americans.  

 

B. Efficacy of Bush Savings Proposals for Creating Retirement Income 

Security for Middle and Lower Income Workers 

 

 In the previous subsection, I suggested that the White House savings 

proposals, by providing tax-advantaged savings opportunities outside 

qualified retirement plans, would reduce the incentives for highly paid 

individuals to save in qualified plans. I also suggested that this would result 

in a universe in which fewer firms sponsored qualified retirement plans and 

that some of the remaining plans would be less generous. The retirement-

policy question that follows is to what extent rank-and-file employees would 

use the new private savings vehicles to replace the lost retirement income 

from their employer-sponsored retirement plans.  

                                                 
118. IRC § 401(k)(11). 

119. See, supra notes 102-05 and accompanying text. 



2008] Slouching Towards a Consumption Tax 145 

 

 

 The answer, I fear, is that much of what is lost will not be replaced. I 

have two concerns in particular: first, that many of the rank-and-file 

employees who today participate in qualified plans would not utilize the 

Bush personal savings accounts; and second, that much of what is saved in 

the Bush personal savings accounts will be consumed prior to retirement. 

 

 1. Obstacles to Rank-and-File Utilization of Bush Savings Proposals 

 

 There are reasons to suspect that rank-and-file employees may not 

exhibit a robust response to the Bush personal savings proposals, at least 

compared to their response to qualified retirement plans (when such plans are 

offered to them). I start with the observation that many rank-and-file  

employees are what some have referred to as “reluctant savers,”
120

 people 

who find it difficult to save for retirement on their own, because of pressing 

needs for present consumption. Indeed, as this paper earlier suggested, the 

traditional explanation for the qualified-plan tax expenditure is that qualified 

retirement plans will create retirement income for reluctant savers.
121

  

 Qualified plans can create retirement income for such savers in 

either of two ways. In plans funded exclusively with employer dollars, the 

Internal Revenue Code’s nondiscrimination rules require coverage of some 

rank-and-file employees automatically, without action on their part and 

without any corresponding reduction in their immediate compensation. In 

such plans, then, reluctant savers are forced to save for retirement.  

 Other plans, primarily 401(k) plans, do not force employees to save: 

in such plans, an employee must elect to participate, and there is a direct cost 

to the employee who does so elect: lower immediate compensation. But the 

law encourages, and arguably coerces, participation among reluctant savers. 

Employees are in part encouraged to participate because their 401(k) 

contributions enjoy the benefit of tax deferral.
122

 More important, to enjoy 

qualified tax status, 401(k) plans must comply with a special set of 

nondiscrimination rules, which can be satisfied through annual testing that 

compares the deferral rate of the highly compensated employees with the 

deferral rate of rank-and-file employees.
123

 In effect, the amount of income 

that highly-paid employees can defer in a 401(k) plan depends on the amount 

that non-highly compensated employees defer. Thus, many firms encourage 

employee participation in their 401(k) plans through employer-provided 

matching contributions and educational programs on the value of saving for 

                                                 
120. Daniel Halperin, Employer-Based Retirement Income – the Ideal, the 

Possible, and the Reality, 11 Elder L. J. 37 (2003)(noting problems of reluctant 

savers). 

121. See, supra notes 29-37, and accompanying text. 

122. See, supra note 29-36, and accompanying text. 

123. IRC § 401(k)(3).  



146 Florida Tax Review [Vol. 9:2 

 

retirement.
124

 And the Internal Revenue Code also offers safe-harbor 

alternatives that exempt a plan from annual non-discrimination testing if the 

plan provides either a statutorily defined matching contribution or provides a 

mandatory non-matching contribution for each participant.
125

 Moreover, 

behavioral economists have observed that individuals are more likely to save 

if they can pre-commit to saving a portion of future earnings: 401(k) plans 

offer employees a pre-commitment mechanism.
126

  

 Thus, qualified plans either compel or encourage rank-and-file 

employees to save for retirement. The Bush personal savings vehicles are 

voluntary, so reluctant savers will not be compelled to save; in effect, this 

means that reluctant savers who save only because qualified plans compel 

them to will lose retirement savings if their qualified plan is terminated or cut 

back. Moreover, the Bush savings proposals will offer weaker incentives for 

voluntary savings than 401(k) plans. Indeed, the Bush savings accounts offer 

a single meaningful incentive for participation: the permanent exclusion of 

investment return on plan contributions from income taxation.
127

 They will 

not provide matching contributions and employers will not have incentive to 

sponsor savings educational programs. Perhaps most important, the Bush 

accounts do not incorporate the pre-commitment mechanism of 401(k) plans.  

 One might respond that vendors of investment products – the 

Fidelities and Vanguards of the world – would aggressively market RSAs 

and LSAs, and this is certainly probable. What is not probable, however, is 

that such vendors will be especially interested in establishing such accounts 

for small investors or that their marketing will be directed to people who will 

open small accounts. The reason for this prediction is simple: the fixed costs 

(and potential liabilities) of small accounts likely render such accounts only 

marginally profitable.
128

 Indeed, most mutual fund vendors have minimum 

                                                 
124. Alicia H. Munell & Annika Sunden, Coming Up Short/The Challenge 

of 401(k) Plans 58 (The Brookings Institution 2004). 

125. IRC § 401(k)(12).  

126. See, generally, Deborah M. Weiss, Paternalistic Pension Policy: 

Psychological Evidence and Economic Theory, 58 U. Chi. L. Rev. 1275 (1991). 

Moreover, in response to work by some behavioral economists – most notably 

Richard Thaler and Shlomo Benartzi – a number of section 401(k) plans are using 

“automatic enrollment,” in which employees must opt out rather than opt in to plan 

participation. William G. Gale, J. Mark Iwry & Peter R. Orszag, The Automatic 

401(k): A Simple Way to Strengthen Retirement Savings (Brookings Institute 2005), 

at www.brookings.edu/views/papers/20050228_401k.pdf. (Last visited Jun. 30, 

2007).  

127. It is, of course, possible to revise the Bush proposals to provide 

matching contributions.  

128. John Waggoner, Some Mutual Funds Still Let You Start Small, USA 

Today, (Aug. 23, 2001) at D., available at http://www.usatoday.com/money/perfi/ 

columnist/waggon/2001-08-24-waggon.htm. (Last visited Jun. 30, 2007). The article 



2008] Slouching Towards a Consumption Tax 147 

 

 

contribution requirements; Vanguard, Fidelity, and T. Rowe Price, for 

example, will not open an individual retirement account with less than 

$1,000 and many of their fund options require substantially higher minimum 

contributions.
129

  

 Moreover, the RSA/LSA accounts will be funded with after-tax 

dollars. Earlier this article discussed how the contribution of an after-tax 

dollar contribution to a Roth-type savings vehicle shields more investment 

income from taxation than a deductible contribution to a traditional 

retirement plan, but this is not an advantage to people who cannot afford to 

contribute beyond the contribution limits for such accounts (and this will be 

most people). Moreover, there is reason to suspect that such individuals 

generally prefer before-tax to after-tax contributions.
130

 Thus, the tax 

incentives for middle- and lower-income people to contribute to RSAs and 

LSAs may be weaker than the tax incentives for contributions to regular 

401(k) plans. The White House proposals would eliminate traditional 

individual retirement plans, where contributions are made with pre-tax 

dollars. 

 

 2. Production of Retirement Income 

 

 Qualified retirement plans are designed, generally speaking, to 

provide individuals with a stream of income following their withdrawal from 

the labor market because of retirement or disability. The Internal Revenue 

Code, through regulatory restriction
131

 and excise taxes,
132

 discourages use 

                                                                                                                   
notes, however, that some mutual funds might pursue the small investor in the 

expectation that with regular contributions and compounding of investment income, 

the small account of today will become the large account of tomorrow. 

129.See, e.g., https://flagship2.vanguard.com/VGApp/hnw/content/Account 

Serv/Retirement/ATSTradIRAOverviewContent.jsp ($2,500 minimum investment 

for most Fidelity Investment funds). 

130. There are three possible explanations for this preference: first, that 

individuals have a behavioral preference for reducing taxes in the year of 

contribution; second, that individuals may fund part of their contribution with the 

immediate tax savings; and third, that middle and lower income individuals believe 

that they will have lower marginal tax rates in retirement. I have not been able to 

find any empirical research on the question posed in the text: whether traditional or 

Roth treatment is a stronger tax incentive for moderate income individuals. But I am 

reasonably confident that my hunch that before-tax treatment is a more effective 

incentive than Roth treatment for moderate income taxpayers. Others seem to share 

this view. See, Karen C. Burke and Grayson M. P. McCouch, Lipstick, Light Beer, 

and Back-Loaded Savings Accounts, 25 Va. Tax Rev. 1101, 1141 (2006). 

131. Treas. Reg. § 1.401-1(b) (pension plan intended to pay income after 

retirement).  



148 Florida Tax Review [Vol. 9:2 

 

of such plans to provide pre-retirement income. In addition, pension plans 

(but not profit-sharing plans) must make benefits available as life annuities, 

which relieves participants from the planning necessary to prevent 

dissipation of retirement savings prior to death.
133

 Moreover, a participant’s 

spouse must consent to any benefit form from a pension plan that does not 

include a survivor’s benefit for the spouse,
134

 making it more likely that 

benefits will not be exhausted before the death of a participant’s spouse.  

 In qualified plans these rules work imperfectly: we know, for 

example, that pre-retirement leakage occurs,
135

 and we know that some 

spouses consent to distributions that do not include a survivor benefit if the 

spouse outlives the participant.
136

 This can result in participants exhausting 

retirement savings prio to their death, or married couples exhausting benefits 

before they both die. But the rules do provide at least some restraints on 

consuming qualified plan savings in a manner inconsistent with the purpose 

of the qualified-plan tax subsidy: providing retirement income.  

 The White House savings proposals do not include equivalent 

restraints on non-retirement use of assets in the new savings accounts, but 

impose weaker and in some cases no rules limiting pre-retirement 

consumption or post-retirement exhaustion of RSA/LSA assets. This would, 

I fear, exacerbate the problem of old-age poverty, at least at the margins.  

 The LSA, to which individuals could contribute $7,500 per family 

member under the 2003 budget proposal, would impose neither regulatory 

restriction nor excise tax on pre-retirement withdrawals.
137

 The RSA, while 

imposing no direct regulatory restrictions on premature withdrawals, would 

impose an excise tax on distributions before age 58, which would discourage 

some pre-retirement RSA distributions.
138

 Most individuals, however, would 

probably choose to contribute first to the LSA (precisely because it does not 

impose a penalty on early distributions) and only contribute to an RSA to the 

extent that they are able to contribute in excess of the LSA limits.
139

 Thus, 

many individuals who use the new savings proposals will be subject to 

neither direct nor indirect restrictions on premature withdrawals of income. 

                                                                                                                   
132. A 10% excise tax is imposed on most qualified plan distributions made 

before the year in which the participant attains age 59.5. IRC § 72(t).  

133. IRC § 411(a)(7); Treas Reg. 1.401-1(c)(1)(I). 

134. IRC §§ 401(a)(11), 417.  

135. Munnell and Sudden, supra note 122, at 125, et. seq. 

136. See, generally, Camilla E. Watson, Broken Promises Revisited: The 

Window of Vulnerability for Surviving Spouses Under ERISA, 76 Iowa L. Rev 461 

(1991). 

137. General Explanation (2003), supra note 1, at 120. 

138. Id. 

139. See, How Bush’s Retirement System Overhaul Hits Home, supra note 

60. 



2008] Slouching Towards a Consumption Tax 149 

 

 

Moreover, the excise tax on RSAs would only apply to withdrawals of 

investment income,
140

 not to withdrawals of the contributions themselves, 

and if we predict that the shape of withdrawal regulations will follow those 

applicable to Roth IRAs, withdrawals will be treated as coming from 

contributions rather than investment income until aggregate withdrawals 

exceed aggregate contributions. Thus, many pre-retirement withdrawals 

would be tax-free even from an RSA. Thus, compared to assets saved in 

qualified plans, assets saved in the White House savings proposals would be 

more likely to be used before a person stopped working.
141

  

 Moreover, the Bush personal savings accounts neither require nor 

induce individuals to take benefits in annuity form. While an individual 

could certainly purchase an annuity using plan assets on the commercial 

insurance market, the cost of commercial annuities is high (especially for 

relatively small annuities).
142

 Thus, many holders of RSAs and LSAs will 

have to manage their retirement savings, not only investing appropriately but 

also planning withdrawals to ensure a more or less constant income stream 

until death. This is a difficult enterprise, and there is little reason for 

optimism that most Americans are equipped to manage their assets in this 

way.
143

 Compared to qualified pension plans, then, the Bush savings vehicles 

will probably result in fewer retired individuals receiving a lifetime stream of 

income after retirement.  

 Finally, the Bush proposals do not require spousal consent to a form 

of benefit without a survivor annuity. Indeed, nothing in the Bush proposal 

promotes protection of surviving spouses. Again, this is a step backward 

from the rules protecting spouses in qualified plans.  

                                                 
140. General Explanation (2003), supra note 1, at 120. 

141. Footnote about double-bind: if you put limits, fewer people will 

contribute. 

142. See, Barry Perlman, Chris Lott, Ed Dollars, Subject: Insurance-

Annuities in The Investment FAR website, available at http://invest-

faq.com/articles/ins-annuities.html. (Last visited June 30, 2007) (Annuities usually 

have a sales load, usually have very high expenses, and always have a charge for 

mortality insurance. The expenses can run to 2% or more annually). Also see, Jeffrey 

R. Brown and Mark J. Warshawsky, Longevity-Insured Retirement Distributions 

from Pension Plans: Market and Regulatory Issues, National Bureau of Economic 

Research Working Paper Series, January 2001. Available at 

www.nber.org/papers/w8064. (Last visited Jun. 30, 2007) (Individual annuity 

markets do not offer actuarially fair prices). 

143. See, Report of Working Group on Planning for Retirement, supra note 

105. 



150 Florida Tax Review [Vol. 9:2 

 

C. White House Savings Initiative and Rate of Return on Retirement Savings 

 

 The Bush savings proposals may have the effect of reducing the rate 

of return on retirement savings of middle- and lower- income individuals. 

This section suggests five reasons why this might occur. 

 

 1. Loss of Investment Opportunities 

 

 Qualified plan investment programs take one of two forms: assets 

are either pooled and invested by plan fiduciaries on behalf of all participants 

or are invested in accordance with the individual instructions of each plan 

participant (“self-directed plans”).
144

 In the latter case, the plan generally 

offers the participant an array of available mutual funds and other investment 

opportunities, which are initially chosen by a plan fiduciary. In both types of 

plans, the employer, or rather the fiduciary selected by the employer, serves 

in an agency relationship to the plan participants and presumably in most 

cases is more competent than the participants in managing funds or selecting 

investment options. Moreover, plans, which sometimes have substantial 

assets to invest, can bargain for investment options that might not otherwise 

be available to moderate- and lower-income individuals. For example, some 

mutual funds require investors to make a substantial minimum investment 

that is beyond the means of many individuals. Pooled funds, of course, have 

access to such investments and also to venture capital and real estate 

opportunities that may not be available to small individual investors, who in 

any event may not be capable of evaluating them in a sophisticated manner. 

Even self-directed account funds sometimes offer investment funds that 

would not be available to small investors outside plans.  

 In the RSA/LSA formats, employees would not benefit from the 

agency of plan fiduciaries or plan bargaining power and would thus lose 

access to some types of investments, which would undermine their ability to 

optimize portfolio diversification (and if you believe in such things, the 

opportunity to invest in assets that offer better risk/return ratios than typical 

mutual funds).  

 

 2. Fiduciary Protections 

 

 ERISA imposes fiduciary obligations on people who choose how to 

invest plan funds and/or choose investment options for participants in plans 

where participants direct investment choices.
145

 The obligations include 

                                                 
144. See, ERISA § 404(c), 29 U.S.C. § 1004(c). 

145. See, Pension Plan Administration: Participant-Directed Accounts, 

Bureau of National Affairs. Available on Westlaw at BNA-CBG 221460. (Last 

visited Jul. 10, 2007). 



2008] Slouching Towards a Consumption Tax 151 

 

 

prudence and loyalty,
146

 which require care in selecting, and in monitoring, 

plan investments.
147

 These obligations should in theory, and generally will in 

practice, discourage plan fiduciaries from investing plan resources in 

investments that carry uncompensated risk and/or that unnecessarily 

compromise portfolio diversity.  

 In contrast to qualified plans, RSAs and LSAs will remove the 

agency of the employer-designated fiduciaries. Individuals, who sometimes 

will not have extensive training or experience in investment management, 

will be saddled with the responsibility of investment management, including 

selecting appropriate investments, monitoring portfolio performance, and 

periodically adjusting asset allocations. Such individuals might lack the 

sophistication necessary to replicate the professional investment management 

provided by or through plan fiduciaries.  

 One can respond to this criticism, by noting that participants in self-

directed qualified plans are currently required to choose among investment 

options. But at least in such plans the investment options themselves are 

selected by plan fiduciaries. Moreover, some observers of qualified plans 

have suggested that ERISA’s encouragement, if not mere tolerance, for 

participant-directed plans, has been a mistake and should be eliminated or at 

least reduced.
148

 The RSAs and LSAs move in the opposite direction. 

 

 3. Education 

 

 Employers sometimes provide investment education to employees 

participating in self-directed plans.
149

 It is not clear that vendors of 

investment products will provide similar education or advice, and if they do, 

there is at least some risk that the education may reflect the vendor’s own 

pecuniary interests.
150

 

 

                                                 
146. ERISA § 404(a), 29 U.S.C. § 1004(a). 

147. See, Pension Plan Administration: Participant-Directed Accounts, 

supra note 145. 

148. See, Susan J. Stabile, Freedom to Choose Unwisely: Congress’ 

Misguided Decision to Leave 401(k) Plan Participants to their Own Devices, 11 

Cornell J. L. & Pub. Pol’y 361 (2002). 

149. See, Pension Plan Administration: Participant-Directed Accounts, 

supra note 143. 

150. The Pension Protection Act of 2006 added a prohibited transaction 

exemption for provision of investment advice by financial firms with potential 

conflicts of interest in rendering the advice, although the exemption does include 

some safeguards against abuse. See, ERISA § 408(b)(14).  



152 Florida Tax Review [Vol. 9:2 

 

 4. Costs 

 

 Plans are able to spread fixed costs among all participants and will 

also generally be in a better position than individuals to negotiate low costs 

from fund vendors and service providers.
151

  

 

 5. Effect of the Proposed Income Exclusion of Stock Dividends on 

Investment Returns 

 

 The White House proposal to eliminate tax on stock dividends 

should, according to conventional economic analysis, provide a one-time 

(arguably windfall) gain to holders of stock and should, over time, increase 

before-tax yields on taxable investment assets (such as productive real estate 

holdings and debt instruments).
152

 However, because there will continue to 

be some tax-exempt investors who purchase some stock, the discount on 

stock yields will probably not completely reflect the tax advantage that stock 

would enjoy.  

 A rational affluent investor, in reaction to the Bush savings 

proposals, could be expected to move qualified stock holdings outside tax-

sheltered funding vehicles (such as qualified plans and RSA/LSAs) and to 

use tax-sheltered vehicles to hold investments that would otherwise be 

taxable. By doing so, such an investor could maintain a well-diversified 

portfolio while maximizing after-tax returns Investors of moderate means, 

however, might not be able to achieve such results. First, they might not have 

the same access as affluent investors to the full range of taxable investments. 

Second, a moderate-income investor might find it difficult to maintain a  

diversified portfolio without substantial investment in stock, whose return 

can be expected to be burdened with a new implicit tax reflecting the 

exclusion of dividends from the income tax base outside the plan. The effect 

would be to reduce rates of return on new savings by at least some moderate-

income individuals, either because of reduced return on stock investments or 

reduced portfolio diversification. 

 An interesting result of the exemption of dividends from tax would 

be that participants in qualified plans would be the only individuals who 

would effectively pay tax on dividends, since they would pay tax on plan 

distributions 

 

                                                 
151. The employer has a fiduciary duty to negotiate for reasonable fees and 

acceptable service levels from investment vendors. Department of Labor, Employee 

Benefit Security Administration, Understanding Retirement Plan Fees And 

Expenses, available at http://www.dol.gov/ebsa/publications/undrstndgrtrmnt.html. 

(Last visited Jun. 30, 2007).  

152. See, supra note 109. 



2008] Slouching Towards a Consumption Tax 153 

 

 

D. Budget and Tax Considerations 

 

 The Bush savings initiatives will have budgetary and tax 

implications, implications that may make it more difficult for a future 

government to address social issues, including issues of old-age poverty, 

which I suggest in this article would be exacerbated by the Bush savings 

proposals.  

 The Joint Committee on Taxation estimated that the full dividend 

exclusion proposed in the 2003 budget proposal would cost $235 billion 

revenue over the ten years following its enactment.
153

 The RSA/LSA was 

expected, in the short term, to raise revenue during its first five years and 

thereafter to lose revenue.
154

 The proposal, then, would increase annual 

budget deficits as far as the eye can see, or at least as far as the Joint 

Committee estimates extend;
155

 the increased national debt would burden 

national fiscal policy in the future. As I argued in earlier sections of this 

paper, the beneficiaries of the revenue shortfalls would generally be the 

relatively affluent individuals who are most likely to use the Bush savings 

proposals. 

 The scoring of the Joint Committee assumes static tax rates. If debt 

repayment becomes an increasingly large part of the budget, which is 

certainly a possibility, it may become necessary to increase tax rates to pay 

for governmental programs, including programs to deal with old-age income 

insecurity. The Bush proposals, however, will have constricted the tax base 

by eliminating a shareholder level tax on corporate income and by exempting 

from tax returns on investments held by RSAs and LSAs, meaning that the 

increases in tax rates on income remaining in the tax base (primarily wages) 

will be steeper than they would have to be if the tax base were broader. 

Moreover, since distributions from qualified plans would be included in the 

tax base, retirees would share in the increased tax burden to the extent that 

their retirement income is generated by qualified plans but not to the extent it 

is derived from RSAs or LSAs, or through investments in stock. This may be 

disturbing to people concerned with either horizontal or vertical equity, the 

latter because the more affluent are more likely than the less affluent to 

derive substantial retirement income outside qualified plans; the former, 

because of the disparate treatment of different sources of investment income. 

                                                 
153. Joint Committee on Taxation, United States Congress, Description of 

Revenue Provisions Contained in the President’s Fiscal 2004 Budget Proposal 123, 

JCS-7-03 (Mar. 2003). 

154. Id. See, generally Karen C. Burke and Grayson M. P. McCouch, 

Lipstick, Light Beer, and Back-Loaded Savings Accounts, 25 Va. Tax Rev. 1101, 

1123 (2006) for a discussion of the revenue effects of the savings account proposals. 

155. Id. 



154 Florida Tax Review [Vol. 9:2 

 

 It would of course be possible for a future Congress to subject the 

Bush savings accounts and stock dividends to some level of tax, but this may 

prove politically difficult. Individuals who invested in such accounts would 

be able to contend that by choosing to contribute to such accounts with 

before-tax dollars, they had entered into a contractual-type agreement 

permanently entitling them to a tax exclusion for any investment income 

produced by such accounts. This may well be a compelling political 

argument.
156

 

 Moving to reintroduce a second-level tax on corporate earnings may 

also prove difficult in the future, both on political and policy grounds. 

Politically, corporations and shareholders could be expected to oppose a 

reintroduction of a shareholder-level tax. Moreover, as a matter of first 

principles, a single-level tax on business income has unambiguous economic 

advantage, including roughly equivalent tax treatment of debt and equity 

investments in corporations, and of corporations and other forms of business, 

and these advantages would make it difficult, and perhaps unwise, to revert 

to a double tax on corporate income.
157

 Reintroducing a second level of tax 

on corporate income would almost certainly depress the value of such stock 

at a time when older individuals might be selling such stock to provide 

retirement income. 

 The tax on corporate income could also be increased directly, by 

increasing corporate tax rates. But there are practical limits on how high a 

direct tax on corporate income can be without creating strong tax-avoidance 

incentives and without disadvantaging domestic enterprises in the 

competition for capital. Moreover, as would be the case with reintroducing a 

shareholder level tax, increasing corporate rates in the future could depress 

stock values with an adverse effect on retirees.  

 To recap, the Bush savings proposals would be expensive, would 

benefit relatively affluent individuals who can be expected to save for 

retirement without governmental incentives, and to the extent we later have 

to pay for revenue costs, might be disproportionately paid for by individuals 

who would not, as a group, have derived much advantage from the proposals 

and indeed may have seen their own retirement savings diminished because 

of them. 

 There is one further troublesome aspect of the proposals that might 

become manifest in the future: the possibility that heirs of affluent 

individuals will be able to entirely escape income tax while supporting 

                                                 
156. But see, e.g., Daniel Shaviro, When Rules Change: An Economic and 

Political Analysis of Transition Relief and Retroactivity (University of Chicago 

Press, 2000).  

157. See, Council of Economic Advisers Eliminating the Double Tax on 

Corporate Income, Jan. 7, 2003 available at http://www.treas.gov/press/releases/ 

docs/exclusion.pdf. (Last visited Jun. 30, 2007). 



2008] Slouching Towards a Consumption Tax 155 

 

 

themselves on income from investments. Let me illustrate with an extreme 

example. Assume, for example, a couple whose first and only child is born in 

the year that Congress gives birth to the Bush proposals. The couple, who 

have substantial wage income, establish and make maximum annual 

contributions to RSAs and LSAs for themselves and an LSA for their child. 

Their savings outside of the plan are invested entirely in long-term stock 

holdings.  

 If we assume a 6% return on investments, at age 30 the child’s LSA 

will have accumulated $800,000. Assume also at this point that the child 

begins living on his investment income, which is now $64,000. (This could, 

of course, be supplemented by additional tax-free income from stock 

transferred to him by his parents and imputed income from a home.) Assume 

that when the child is 50, his parents die, leaving him the beneficiary of the 

LSAs and RSAs. At this point, each of the four accounts will have 

accumulated $2.3 million, for a total of $9.2 million. The RSA will have to 

be distributed eventually, but the proceeds, which will be distributed free of 

tax, can be invested in stock. (Again, the total of assets generating tax-free 

income can be increased by bequeathing the child stock.) In any event, the 

child, and future generations, could enjoy substantial annual income without 

ever paying income tax (with capital gains avoided by not disposing of the 

stock).
158

  

 Note that with respect to future generations, this is not exclusively a 

consumption tax regime; it is a tax on wage income and some but not all 

business income. Non-corporate business income passed through to an RSA 

or LSA will also receive a permanent exemption from income tax. The 

regressivity of such a tax system would, to my mind, be hard to justify, even 

if some affluent individuals continue to receive some wage income. 

Moreover, to ensure a meaningful tax on business income, there would have 

to be a thorough house cleaning of the corporate income tax, focusing on the 

aggressive tax planning that minimizes the corporate tax and deliberate 

preferences whose rationales are weaker in a single-level tax corporate tax 

system.  

 

V. WEALTH ACCUMULATION V. RETIREMENT INCOME SECURITY 
 

 The Bush Administration’s savings proposals reflect two broad and 

casually related themes: income verses consumption/wage tax base, and 

social insurance verses wealth creation.  

                                                 
158. It should, of course, be observed that corporate income would be taxed 

and thus anyone deriving dividend income or capital gain from corporate 

investments would be subject to an implicit tax. Nevertheless, to many wage earners, 

it may appear that such individuals are escaping all tax liability.  



156 Florida Tax Review [Vol. 9:2 

 

 The first theme is the tax theme of what our nation should settle on 

as the appropriate tax base. The Bush proposals endorse further movement 

from our current mixed income/consumption tax base toward a purer 

consumption/wage base model. Proponents of a consumption base model 

argue that unlike an income tax, a consumption base is neutral with respect 

to lifetime consumption patterns (by not penalizing savings) and thus a fairer 

tax base.
159

 They also argue that an income tax depresses aggregate national 

savings, and therefore that moving to a consumption tax base will increase 

our nation’s anemic savings rate.
160

  

 While I favor an income tax over a consumption tax base because I 

believe that the former is more adaptable to the equitable tax norm that taxes 

should be apportioned on relative abilities to pay, the preference for one base 

over another ultimately is a reflection of a proponent’s economic ideas and 

perhaps ethical values. There is, so far as I can tell, no mediating criteria of 

truth to resolve the debate over the preferred tax base.
161

 But any movement 

toward a consumption base, in my view, should be conditioned on a national 

debate of the relative merits of competing bases. The Bush proposals, 

however, do not squarely join this issue, but advocate the proposals simply 

as a means to help taxpayers save for retirement and other long-term 

purposes (such as education or saving for a house or a medical 

emergency).
162

 In this sense, the Bush proposals are a stalking horse for a 

move toward a consumption/wage tax, or perhaps more accurately a Trojan 

horse, disguised as a way to help taxpayers to save for retirement rather than 

as an event signaling a dramatic shift in tax base. 

 Moreover, the Bush proposals would nudge us only part way toward 

a consumption tax. As Professors Karen Burke and Grayson McCouch have 

noted, the savings proposals are based on a yield-exempt rather than cash-

                                                 
159. See, notes 87 and 88; see, also, Daniel N. Shaviro, Replacing the 

Income Tax with a Progressive Consumption Tax, 103 Tax Notes 91 (Apr. 5, 2004). 

160. Id. 

161. Moreover, a number of scholars have argued that a consumption tax 

with multiple tax rates can have the same degree of progressivity as an income tax, 

particularly an income tax that defers gain on property until realization. See, e.g., 

Daniel N. Shaviro, Replacing the Income Tax with a Progressive Consumption Tax, 

103 Tax Notes 91 (Apr. 5, 2004). 

162. Pamela Olsen, Assistant Secretary of the Treasury for Tax Policy, said 

of the RSA: The savings options proposed today will give all Americans the 

opportunity and flexibility they need to save for their retirement security and other 

needs.” Department of Treasury Press Release, “ The President’s Savings Proposals: 

Tax-Free Savings and Retirement Security Opportunities for all Americans,” (Feb. 2, 

2004), available at http://www.treas.gov/press/releases/js1131.htm. (Last visited Jun. 

30, 2007). 



2008] Slouching Towards a Consumption Tax 157 

 

 

flow consumption tax model and fail to address the issue of debt within such 

a model.
163 

 
How to ensure retirement income for those who will not voluntarily 

save is itself a difficult issue in designing a consumption tax. In our income 

tax system, we do two things to encourage such savings: first, we privilege 

qualified-plan retirement savings over immediate consumption and other 

types of savings by deferring tax (or carving out a consumption tax niche for 

retirement savings); second, through the nondiscrimination rules, we employ 

more coercive measures to force many employers to provide retirement plan 

coverge for individuals for whom we suspect tax incentives alone would be 

inadequate to cause them to save adequately for retirement. In a consumption 

tax regime, all savings are equally privileged and business owners and 

managers will thus lack tax incentives to have their firms sponsor retirement 

plans that will provide retirement income for reluctant savers. (This latter 

point is, of course, a large part of my critique of the Bush savings proposals.) 

 It is possible, of course, for a society deriving revenue primarily 

from a consumption tax base to design policies that will address the issue of 

myopic retirement savings patterns.
164

 A universal and adequate system of 

social insurance is one approach. Other approaches might provide cash or 

other incentives to employers to establish retirement plans or require all 

wage earners to save a percentage of their compensation in an individual 

account to which appropriate fiduciary protections would attach. We might 

also consider government matching or direct contributions or loans to low 

and moderate income individuals to save for retirement and/or establishing 

pre-commitment devices to improve the savings behavior of such 

individuals.
165

 We might also, as a nation, make a greater investment in 

financial literacy education.  

 The Bush proposals, however, do not propose expanding our Social 

Security program to reflect retirement-income losses that the proposed 

movement to a consumption tax might cause. Nor do the proposals 

incorporate incentives and/or mechanisms that might result in retirement 

savings by otherwise reluctant savers. This should not, of course, be 

unexpected because the Bush administration packaged the RSA and LSA not 

as a step toward a consumption tax that might diminish the retirement 

security of low and moderate wage earners but as a self-contained means of 

increasing savings, including retirement savings, for all Americans.  

                                                 
163. Karen C. Burke and Grayson M. P. McCouch, Lipstick, Light Beer, 

and Back-Loaded Savings Accounts, 25 Va. Tax Rev. 1101, 1136 et. seq. 

(2006)(excellent discussion of these and other issues).  

164. See, generally, Dallas L. Salisbury (ed), Tax Reform/Implications for 

Economic Security and Employee Benefits (Employee Benefit Research Institute 

1997). 

165. Id.  



158 Florida Tax Review [Vol. 9:2 

 

 The second broad theme of the Bush proposals is the emphasis on 

individual wealth creation in contrast to a social contract in which wage 

earners are assured adequate income security in retirement. This preference 

for individual wealth creation verses social contract is more manifest in the 

proposals by the Bush administration, and others, to convert, at least partly, 

the Social Security system from a defined benefit plan to a defined 

contribution plan. 

 But the private pension system, which the Bush proposals would 

weaken, also exhibits important social-insurance elements. In private-sector 

defined benefit plans, for example, benefits historically if not still 

customarily are distributed in annuity form, thus socializing the mortality 

risk among the employee group and limiting the extent to which benefits can 

be converted from retirement income to unrestricted and freely transferable 

wealth. Defined benefit plans also limit the extent to which employees are 

subject to investment risk, both by the employer’s assumption of that risk 

and by mandatory benefit insurance through the Pension Benefit Guaranty 

Corporation. All plans limit the ability of employees to use resources prior to 

retirement, and minimum distribution rules moderate the use of plan tax 

deferral to build tax-advantaged estates (thus helping channel tax benefits to 

the production of retirement income). Finally, and most important, the 

nondiscrimination rules ensure that at least some of the tax benefits that 

plans enjoy are used to create retirement income for moderate income wage 

earners.  

 I do not argue that the private pension system is primarily a system 

of social insurance, only that it has some elements of such a system. Indeed, 

with the strong movement toward defined contribution plans, and within that 

trend toward 401(k) plans in particular, the private pension system also 

reflects themes of individual responsibility and wealth creation.
166

 The 

private pension system, then, is probably best understood as a hybrid system 

combining ideas of collective and individual responsibility, and the creation 

of retirement income and of wealth generally. Bush’s proposals would 

nudge, or perhaps shove, the system’s emphasis further toward individual 

responsibility and creation of wealth. Whether this is a good thing or a bad 

thing is open to debate, and like the debate about the proper tax base, has no 

correct answer, although few readers of this article will mistake on which 

side of the debate my own allegiance lies.  

 What I do not think is fairly debatable is that as we move toward 

individual responsibility and wealth creation, we move further from the ideal 

of universal retirement income security. The White House, however, 

pretends otherwise, justifying its proposals as a means for increasing 

                                                 
166. See, Colleen E. Medill, The Individual Responsibility Model of 

Retirement Plans Today: Conforming ERISA Policy to Reality, 49 Emory L. J. 1 

(2000).  



2008] Slouching Towards a Consumption Tax 159 

 

 

retirement savings among all Americans, including those with low and 

moderate income. If adopted, it would be a national catastrophe for some 

such individuals as they enter retirement unless we simultaneously adopt 

new approaches to help them save for retirement.  

 

VI. THE ERSA AND NATIONAL RETIREMENT POLICY 
 

 Our current retirement system is famously, and inaccurately, likened 

to a three-legged stool.
167

 The stool’s purported legs are Social Security, 

employer pensions, and private savings. The stool, such as it is, is not sturdy. 

Many Americans accumulate little in the way of private savings, and fewer 

than half of Americans in the private work force will receive benefits, and a 

much smaller percentage meaningful benefits, from employer-sponsored 

retirement plans. Thus, for many Americans the more accurate metaphor 

might be a pair of stilts or a pogo stick. And Social Security does not provide 

sufficient income to permit a retiree to live above the poverty level without 

some other sources of income. This suggests that our national retirement 

policy, such as it is, is not much of a policy for many working Americans.  

 There are three approaches to improve national retirement policy. 

One approach would consciously move us toward a more comprehensive 

system of social insurance, either by a direct expansion of the Social Security 

program or through a mandate for employers to adopt retirement plans 

satisfying certain criteria (or a mandate for employees to put aside a 

percentage of their income for retirement). The idea of a mandated employer 

system found little political traction in years past and might not find much 

more today.
168

 A second approach is that implicitly advocated by the Bush 

White House in its personal savings proposals: treating the decision to save 

for retirement as an individual decision for each American, although using 

tax incentives to nudge tax-sensitive Americans to greater savings.  For the 

reasons suggested in this paper, I have doubts about the efficacy of this 

approach. I am, however, optimistic that this approach faces political 

obstacles as formidable as those facing the more universal system that I 

prefer. 

 This leaves a third approach: improving rather than supplanting the 

existing voluntary employer-based retirement system. Since the system is 

voluntary, it is unlikely that changes to the system can ever result in 

universal coverage. It is a flawed system and, given that it is based on the 

voluntary adoption of plans by employers and that employers will generally 

not want to contribute much to such a system on behalf of workers who 

                                                 
167. Munnell, supra note 33, at 2. 

168. Dana A. Muir, From Yuppies to Guppies: Unfunded Mandates and 

Benefit Plan Regulation, 34 Ga. L. Rev. 195 (1999). 



160 Florida Tax Review [Vol. 9:2 

 

would prefer cash compensation, it will almost certainly always remain a 

flawed system. 

 But the system can be improved, and the Bush ERSA proposal is 

right in approach even if wrong in some of its details. As this paper earlier 

described, the ERSA attempts to simplify the regulatory regime governing 

defined contribution plans by jettisoning several inordinately complex 

Internal Revenue Code provisions that, on the one hand, permit employers to 

design plans to limit coverage and benefits for lower-income employees and, 

on the other hand, require some plans that are excessively weighted toward 

key employees to provide at least minimum benefits to other employees. The 

2003 ERSA would have been subject to a single set of relatively simple 

nondiscrimination rules.  

 In effect, the proposed ERSA rules identify a comparative level of 

benefit for rank-and-file employees that is adequate to justify the tax 

expenditure for the highly compensated and then creates a simple 

nondiscrimination test that assures that comparative benefit level while 

minimizing the employer’s design and compliance costs. The Department of 

Treasury put it thus: the ERSA “simplifies qualification requirements while 

maintaining their intent of providing broad-based coverage of employees. By 

reducing unnecessary complexity, the proposal significantly reduces 

employer compliance costs.”
169

 

 The ERSA proposal has, from my perspective, three shortcomings. 

The first, and obvious, shortcoming is that the ERSA proposal is joined to 

the Bush personal account proposals, which this paper has already argued 

will dampen employer appetite for sponsorship of plans. The proposals, 

however, are severable. 

 The second problem is that the proposed nondiscrimination rules do 

not demand sufficient benefits for rank-and-file employees. The 

nondiscrimination tests are modeled after the rules for section 401(k) plans, 

and like 401(k) plans allow employers to design their plan either to satisfy a 

safe harbor or to tie the rate of deferral of highly compensated employees to 

a multiple of the average deferral rate of other employees. Both the safe 

harbor and the annual deferral-comparison test are less demanding than the 

tests under current law for section 401(k) plans. Presumably the argument for 

the less restrictive tests is the elimination of rules under current law that 

accommodate employers who wish to weight plan benefits aggressively 

toward higher-paid employees.  

 Consider first the safe harbor test. To come within a safe harbor, a 

plan would either have to make nonelective contributions equal to 3% of 

                                                 
169. Department of Treasury Press Release, “The President’s Savings 

Proposals: Tax-Free Savings and Retirement Security Opportunities for all 

Americans,” (Feb. 2, 2004), at http://www.treas.gov/press/releases/js1131.htm. (Last 

visited Jun. 30, 2007). 



2008] Slouching Towards a Consumption Tax 161 

 

 

compensation or to offer matching contributions of 50% of compensation on 

the first 6% of pay.
170

 Given that the maximum matching contribution for an 

employee would be 3% (50% of 6% of compensation), it is difficult to see 

why an employer not predisposed to generosity to employees would choose 

the non-elective deferral approach. In 401(k) plans the safe harbor match for 

employees contributing 6% of compensation would be 4.5%.
171

 Thus, most 

employers would presumably choose the matching safe harbor. 

 There are three problems with a matching safe harbor. First, the 

employer has no incentive to encourage employees to contribute to the plan 

since the employer’s cost is increased by such contributions. Second, a 50% 

match on the first dollar contributed may not be high enough to stimulate 

high rank-and-file contribution rates. In 401(k) plans, the safe harbor 

requires a 100% match on the first 3% of compensation deferred, and 50% of 

the next 3% of compensation.
172

 Third, no matter how high the match, some 

employees will not elect to defer and thus will not earn retirement benefits. 

 If an employer elects not to use a safe harbor, the deferral rate for 

highly compensated employees can equal up to twice the deferral rate for 

other employees, although there is no limit on the former if the deferral rate 

for the latter is 6% or greater. A problem with this system is that it looks to 

average deferral rates, so just as with the safe harbors, not all employees will 

have to save for retirement in order for the plan to satisfy the 

nondiscrimination rules. Moreover, a high deferral rate can be achieved 

despite low participation rates by making large non-elective contributions on 

behalf of the lowest paid employees, which is an inexpensive way of raising 

the average deferral rate for the non-highly compensated group.
173

  

 A more effective, simplified nondiscrimination test might involve a 

reverse match. Under this approach, the employer would make an initial 

contribution for all participants and highly compensated employees’ elective 

deferrals would be limited to a statutory multiple of the initial contribution 

rate the employer made.
174

 For example, if Congress set the multiple at three 

and the employer made 3% contributions, all employees, including highly 

compensated employees, would be permitted to defer an additional 9% of 

their compensation. This approach avoids complicated annual testing and has 

                                                 
170. General Explanation, supra note 2, at 126. 

171. IRC § 401(k)(12).  

172. Id.  

173. The Department of Treasury and the IRS promulgated Treas. Reg. 

1.401(k)-2, which places limits on but does not eliminate this strategy. See, 

Retirement Plans; Cash or Deferred Arrangements Under § 401(k) and Matching 

Contributions or Employee Contributions Under § 401(m) Regulations, 69 Fed. Reg. 

78144 (Dec, 29, 2004). 

174. Orszag and Stein, supra note 114, at 669-74 (proposing reverse match 

nondiscrimination rule). 



162 Florida Tax Review [Vol. 9:2 

 

the important advantage over a “matching” safe harbor of ensuring that all 

participants are allocated an initial contribution. It also ties the amount that 

highly paid employees can defer to the amount being deferred for rank-and-

file employees. 

 The third shortcoming of the Bush proposals is that they do not 

attempt to coordinate the proposed rules for defined contribution plans with 

equivalent rules for defined benefit plans. In some sense, this may be 

appropriate, since defined benefit plans are today in decline, and providing 

employers who sponsor them with design flexibility may increase their 

attractiveness if design flexibility is limited in defined contribution plans. 

But there is a danger that small firms will adopt aggressive defined benefit 

plans that will provide little in the way of benefit for lower-paid workers.  

 

VII. CONCLUSION 
 

 This article argues that the Bush personal savings proposals – the 

creation of tax-advantaged personal savings accounts and the exclusion of 

corporate income from shareholder level tax – will have negative effects on 

the retirement income security of many Americans of moderate means. The 

article suggests that such savings accounts would reduce the number of 

employer plans, reduce benefit levels for employees in those plans that 

remain, provide no alternative means for savings that will be attractive to 

adversely effected workers, and reduce the rate of investment return for non-

affluent individuals. The proposals would also impose significant revenue 

cost and might partly immunize the affluent beneficiaries of the proposals 

from tax increases that might be needed in the future to retire the national 

debt that would result from the revenue losses. Like a policy bad nickel, 

these ill-advised concepts keep turning up in proposals from the White 

House, from members of Congress, and from the President’s Advisory Panel 

on Federal Tax Reform.  

 In contrast, the ERSA, if isolated from the personal savings accounts 

and dividend exclusions, provides a helpful first step toward making useful 

adjustments to the current system. Through an ERSA-like approach, we can 

both reduce expensive compliance burdens and ensure that a reasonable 

share of the qualified plan subsidy provides meaningful levels of benefits to 

employees of moderate and low income. Of course, such a modification to 

the system would not result in either universal coverage or benefit adequacy, 

which can only be achieved through an expanded social insurance system (or 

some form of mandated savings). But it would do at least some good and no 

or little harm. 

 



2008] Slouching Towards a Consumption Tax 163