California Secure Choice:Market Analysis & Program DevelopmentNari Rhee, PhDUC Berkeley Center for Labor Research and Education

Oregon Retirement Savings Board MeetingTigard, OR

December 14, 2015

2

Outline

• Implications of California Secure Choice Market Analysis for Plan/Program Design– Opt-out rate, default contribution rate, risk/reward

tradeoff, and access to funds

• Program Design Considerations for State-Sponsored Retirement Savings Plans

3

California Secure Choice (State Auto-IRA Initiative)

Theme Key Provisions

Coverage • Mandatory for private employers with 5+ employees that do not offer qualified plan

• Auto-enrollment for employees with individual choice to opt out

Funding • Employees contribute through payroll deduction• Board sets default contribution rate

Investment • Default investment vehicle selected by Board

Fees • Statute silent on total fees, but 1% is assumed cap

Payout • Not specified in original legislation. Stakeholder interest in lifetime income option.

4

Eligible Population• ~6.3 million workers in ~300,000 establishments• Median annual wage income: $23K

– F/T year round workers: median $36K– Family median income: ~$50K

• 1/2 are age 36 or younger• 2/3 workers of color

– Nearly half of target market is Latino

• Over 1/2 work for businesses with less than 100 employees

5

Key Market Research Findings & Implications for Plan Design

• Sources: Analysis of US BLS Current Population Survey, 6 focus groups (2 in Spanish), and online survey of 1,000 eligible workers

• Key findings:– Opt-out Rate– Risk Tolerance– Importance of Access to Funds

6

Key Focus Group Findings

• Auto-enrollment and payroll deduction popular as an easy way to save

• Some distrust of financial institutions and government, especially among low-income groups

• Low level of financial literacy – (how IRAs work, diversification, risk/return tradeoff, relationship

between risk and time horizon)

• Strong risk aversion, in theory – Most workers voice preference for “safe” investments in focus

group discussions – However, survey results show different preference when

respondents are confronted with concrete risk/reward tradeoffs

Survey: 3/4 of Uncovered Workers Would Stay in California Secure Choice if Auto-Enrolled

Total

3% Contribution

5% Contribution

56%

57%

55%

18%

16%

20%

27%

27%

26%

Stay in programStay in program, but ask your employer to change the contribution rateOpt out of program

NET: Stay in program – 73%

(n=1,000)

(n=500)

(n=500)

Q1. If you were automatically enrolled in the California Secure Choice program above at (3%/5%) of your paycheck, would you…?

NET: Stay in program – 74%

NET: Stay in program – 73%

Source: Matthew Greenwald & Associates, “Online Survey of Employees Without Workplace Retirement Plans: Report of Findings,” October 2015.

Among those who would stay in the program but ask to change their deferral rate, over half would request a higher contribution rate.

Less than 1%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

More than 10%

Don't know

0%

10%

23%

1%

25%

7%

1%

3%

0%

20%

6%

6%

3% Contribution(n=103)

Less than 1%

2%

4%

6%

8%

10%

Don't know

3%

2%

15%

21%

3%

<0.5%

8%

8%

8%

24%

6%

1%

5% Contribution(n=103)

Q1A. What percentage would you ask your employer to change the contribution rate to?

32% would lower their contribution

rate

62% would raise their contribution

rate

43% would lower their contribution

rate

55% would raise their contribution

rate

If the program had automatic escalation of 1% annually up to a maximum of 10%, a fifth would opt out and another third would just stop the increases.

Total 48% 33% 19%

Stay in the program with the increases in contribution

Stay in the program but ask your employer to stop the increases

Opt out of the program entirely

Q4. Suppose that the program automatically increased your contributions by 1% of your paycheck every year up to a maximum of 10%. Would you…

• Opt-out rates if the program included an automatic escalation feature are higher for non-Hispanic than for Hispanic respondents (23% vs. 15%).

(n=1,000)

Q4. Suppose that the program automatically increased your contributions by 1% of your paycheck every year up to a maximum of 10%. Would you…

Respondents were shown two investment fund options for the program.The retirement savings program could be set up with different fund options. Two examples are described below. Please read both the description and the chart:

A. A “Balanced Fund" that has a mix of 60% stocks and 40% bonds. This is expected to provide significant investment growth over the long term. However, performance will vary a lot from year to year, and there is a 1-in-50 (2%) chance of losing some of the principal (your contributions) after 20 years.

B. A “Money Market Fund,” an interest-bearing account that protects the principal. You will never lose your deposit, but interest rates may fail to keep pace with inflation.

Risk/Reward: About twice as many prefer a Balanced Fund over a Money Market Fund.

Option A: Balanced

fund51%

Option B: Money

Market fund23%

Makes no difference

10%

Don't know16%

• Half (51%) prefer a Balanced Fund compared to 23% who prefer a Money Market Fund. Another 10% do not have a preference and 16% don’t know.

- Preference for the Balanced Fund increases with education and income.

Base: Total, n=1,000

Q5. Do you prefer to have the money automatically invested in a low-cost fund that is:

85% of Uncovered Workers Support Auto-Enrollment Retirement Savings Program

Very good idea57%Somewhat good

idea29%

Somewhat poor idea10%

Very poor idea4%

Q2. In general, do you think automatically enrolling workers into a retirement savings program while giving them the option to choose not to participate is a:

n=1,000

13

Implications for Plan Design

• Default contribution level drives savings behavior and ultimately, retirement income adequacy.– Auto-enrollment popular in focus groups and survey

– Same tolerance for 3% and 5% contribution rate.

– Tolerance for auto-escalation as long as EEs have choice to stop it

• Workers seem to prioritize growth over principal protection. How you frame risk/reward is important.

• Lack of access to funds during emergencies is a deal-breaker for a significant share of eligible workers in CA.

Impact of Default Contribution Rate on Program Finances: Analysis for CSC

*Sequence of Annual Investment Returns as follows: 0%,0%,-10%,-10%,5%,5%,10%,10%,0%,-15%,5%,5%,5%,5%,5%

Financing requirements and program expense ratios are very sensitive to the default contribution rate.

The opt-out rate has a small to moderate impact below 50% because key variable costs are tied to the number of participants and the CSC program is large in scale.

Because the baseline model assumes that participants are defaulted to very low risk investments during first three years, the impact of adverse investment returns is only seen in later year program expense ratios.

14

Required Financing

(USD Millions)

Payoff Year

Year 1Program Expenses

as % of Assets

Year 5Program Expenses

as % of Assets

Year 10Program Expenses

as % of Assets

Baseline (5% Contribution; 25% opt-out) $73 5 3.17% 0.58% 0.37%

3% Contribution Rate $129 7 4.78% 0.79% 0.47%

10% Opt-out Rate $73 5 3.02% 0.57% 0.36%

Adverse Investment Returns* $72 5 3.17% 0.63% 0.37%

15

Traditional or Roth?

• Different tax implications

– Traditional IRA: pre-tax contributions, withdrawals taxed at retirement

– Roth IRA: post-tax contributions, tax-free withdrawal. Penalty-free withdrawal of contributions.

• Income tax filing/deduction required to claim tax benefit for traditional IRA, but not Roth.

• CA eligible workers: lower income tax bracket

• No preference in focus groups. Empirical research indicates that deferral rates are not sensitive to pre/post tax.

16

Program Design Issues That Call for Early Exploration

• Auto-enrollment and recordkeeping logistics• Interface between state agencies and Third

Party Administrator (TPA)• Enforcement

17

Auto-Enrollment & Recordkeeping Logistics

• Active vs Passive auto-enrollment?• Requiring active confirmation to start payroll

deduction can increase opt-out rate to as high as 30%

• Employers may not be best suited to track opt-outs and contribution elections– may reduce participation and create employer liability – makes auto-escalation impractical

• Passive auto-enrollment can be facilitated by Recordkeeper

18

Additional Recordkeeping Issues

• How will invalid SSNs be resolved?

• State auto-IRAs require hybrid recordkeeping model that combines features related to employer sponsored plans and individual IRAs

• Engage employers, recordkeeper, and payroll processors early

19

Engage State Agencies Early

• California: necessity for buy-in from key agencies– Dialogue with Employment Development Department (EDD) regarding

employer outreach, enforcement, and potential role in program operations

– Recent discussions with Director of Industrial Relations regarding enforcement model

– Additional legislation needed to assign enforcement authority for Secure Choice

• Agencies differ in capabilities, program models, culture, and IT requirements

• Engage agencies early to parse responsibilities, administrative and enforcement procedures, and costs

20

Enforcement• DOL requirement for mechanism to enforce EE rights

under state plans• What will be the key enforcement protocols? What

state agencies will be involved?– ER compliance with state requirements– EE complaints/grievances if not resolved directly with ER and

TPA

• Enforcement budget– Some costs may be covered by program fees; other costs may be

the responsibility of the state

21

Concluding Thoughts

• Standard tools for market analysis can yield valuable data to inform retirement savings plan design

• Fleshing out critical program and enforcement logistics takes time. Engage key entities early.

22

AcknowledgmentsCalifornia Secure Choice study team:

• Overture Financial (lead) Investment• UC Berkeley CLRE Labor research• Matthew Greenwald & Assoc. Market research• Segal Consulting Plan administration• BridgePoint Recordkeeping • RMD Pension Consulting Actuarial

Integrated Market Analysis, Program Design, and Financial Feasibility Study expected mid-January 2015