TEACHERS RETIREMENT BOARD BENEFITS AND SERVICES COMMITTEE

Transcription

1 TEACHERS RETIREMENT BOARD BENEFITS AND SERVICES COMMITTEE SUBJECT: Five-Year Tiered and Single-Year Permissive ITEM NUMBER: 3 Service Credit Purchase Rates CONSENT: ATTACHMENT(S): 1 ACTION: X DATE OF MEETING: / 45 mins. INFORMATION: PRESENTER(S): Rick Reed PURPOSE The purpose of this item is to respond to a committee request to evaluate alternative five-year and single-year rates for the purchase of permissive service credit, to discuss any fiscal implications to CalSTRS adopting one of the alternative rate structures, and adopt contribution rates for permissive service credit for the period January 1, 2014 through June 30, BACKGROUND Members of the Defined Benefit (DB) Program may purchase service credit for a variety of activities that were previously undertaken by the member. These include: Certificated employment in a child care center operated by a school district or county office of education; Teaching at a state college or university in California; Teaching at either the California School for the Deaf or the California School for the Blind; Up to two years of certificated employment in California for the federal Job Corps; Certificated employment in a Native American School; Previous certificated service subject to coverage in the DB Program but prior to DB Program membership; Employment in a public school in another state; Service in the American Red Cross prior to 1957; Approved leave to participate in the federal Mutual Educational Cultural Exchange Program; Maternity/paternity leave; Military leave; Family care and/or medical leave; and Approved sabbatical leave. BS 13

2 Page 2 The contribution rate for permissive service credit was established in law by Chapter 279, Statutes of 1982, as a rate adopted by the board based on the most recent valuation of the DB Program and increased for any subsequent benefit enhancements. At its September 2003 meeting, the board adopted its current method for establishing the cost based on the actuarial value of the benefit. Additionally, effective July 1, 2004, the contribution rate for permissive service credit is based on a tiered table with a contribution rate for every tenyear age group beginning with age 20. The contribution rates determined are rounded to the nearest tenth of a percent. AB 340 (Chapter 296, Statutes of 2012), otherwise known as the California Public Employees Pension Reform Act of 2013, or PEPRA, changed the benefit structure for certain eligible new CalSTRS members. Because the new benefit structure (2% at 62) is significantly different than the pre-2013 benefit structure (2% at 60), staff recommended to the committee at its April 2013 meeting that the board adopt two sets of tiered rates, one for each distinct benefit structure, each based on ten-year age increments. Committee members raised a concern at that time that ten-year increments were too broad, noting that a member might delay a purchase until the end of the tenyear increment in order to take advantage of the static rate throughout a decade of age. To allow the staff to respond to these concerns, the committee adopted rates through December 31, 2013 based on the current ten-year age increments, and directed staff to develop one and five-year tiered alternative rates for consideration and possible adoption for the remainder of the fiscal year (January 1 through June 30, 2014). The rate that CalSTRS charges members in the current ten-year age increment schedule depends on the age group that the member is a part of when he or she buys the permissive service. Members who buy permissive service pay the same group rate regardless of where in the age group range they are in when they buy the service. For example, a member who buys service at age 30 would pay the same rate if she bought the same service at age 39. If the member bought the service at age 40, that member would pay a different rate, the rate of the age group from 40 to 49. This type of pricing arrangement naturally leads to two key questions: 1. Does this type of pricing cost the DB Program income, and 2. Is it fair to individual members? At the committee s direction, staff studied the permissive service credit purchase pricing issue, and as a result staff developed permissive service credit rates using five-year age groups and individual ages. Key results of the study are: Both single-year and five-year tiered rates are close enough to the ten-year tiered rate in their respective age group that cross-subsidization is not large enough to recommend a change. Five-year tiered and single-year rates may be different enough to an individual member from the ten-year tiered rate to affect service purchase decisions. Prior experience, however, indicate that the change in behavior is likely to be small. BS 14

3 Page 3 Overall impacts on costs to the DB Program of using five-year tiered or single-year rates compared to the current ten-year rates are difficult to determine but the total difference in program costs is expected to be small because the change when a member purchases service credit is expected to be small. CONTRIBUTION RATES UNDER A FIVE-YEAR AND SINGLE-YEAR STRUCTURE The following two tables compare both alternative rates to the rates in effect for the period July 2013 through December The purchase rates are shown as a percentage of salary for one year of service purchased. An important issue to address for any group rate is the crosssubsidization involved. Cross-subsidization occurs on an actuarial basis when one group or individual as part of a larger group pays more or less than the actuarial cost of a product. Closely related to cross-subsidization is the concept of anti-selection. Anti-selection occurs when an individual purchasing a product knows important information that would affect the price of the product that the seller does not know. For example, a 30 year old person buys life insurance knowing that he is expected to live only one year, while the seller believes that they are selling life insurance to a normal, healthy 30 year old. Five Year Tiered Rates Compared to the Current Ten-Year Age Tiered Rate CalSTRS 2% at 60 CalSTRS 2% at 62 Approved Approved Alternative Alternative Count 10-year Percent 10-year Percent 5-year age 5-year age (Since age group Change age group Change group rate group rate Age Range ) rate rate , , , , , , and over For this alternative rate, each current ten-year age group is comprised of two five-year age groups, the first five year age group is referred as the younger five-year age group and the second five-year age group is referred as the older five-year age group. For example, the current ten-year age group of 20 to will contain two five-year age groups, the younger BS 15

4 Page 4 five-year group 20 to and the older five-year age group of 25 to There is crosssubsidization between the younger and older five-year age groups. For example, in the ten-year age group 30 to , the younger five-year age group, 30 to subsidizes the ten-year age group by about 0.7 percent of salary (17.3 minus 16.6). This cross-subsidization (of about four percent of the ten-year age group rate) is not large enough to recommend that a change be made to the narrower age group. A review of the table indicates for age ranges that the rates for the younger five-year age groups are less than the ten-year tiered rate, and the rates for the older five-year age group are greater than the ten-year group rate. The primary reason for the lower rates for the younger five-year age group compared to the rate of its ten-year age group is that CalSTRS has more time to invest the payments to earn investment income to help pay for the benefits purchased. Generally, the investment return assumption is the single biggest reason for the difference between purchase rates; and that increased time to earn investment income to fund a benefit is only partially offset by the fact that the expected salary at retirement used to determine the member s benefit is usually greater for members who start teaching at younger ages, which would generally increase rates for younger age groups. As for the older five-year age group, their cost would no longer be subsidized by the members of the younger five-year age groups; as a result, their rates are more than the ten-year age tiered rate. In addition, the five-year tiered rate peaks at age group 60 to , which is in the normal retirement age range, and decreases gradually as a member s age increases beyond age 65. When members age, fewer benefits will be collected; therefore, less money is needed to be contributed. In contrast, the 20 to age group rate is the same as the approved 20 to age group rate as shown in the chart. The reason for this is that the new rate was based on historical purchase data of four members, which was not a good representation of that group. BS 16

5 Page 5 Single-Year Rates Compared to the Current Ten-Year Tiered Rate CalSTRS 2% at 60 CalSTRS 2% at 62 Approved Approved Alternative Alternative 10-year 10-year 1-year Age Percentage 1-year Age Percentage Count Age Age Group Change Group Change (Since Group Group Rate Rate Age Range ) Rate Rate , , , BS 17

6 Page , Approved 10-year Age Group Rate CalSTRS 2% at 60 CalSTRS 2% at 62 Alternative 1-year Age Group Rate Percentage Change Approved 10-year Age Group Rate Alternative 1-year Age Group Rate Percentage Change , , , , , , , , and over A similar analysis was made for the single-year rates compared to the current ten-year tiered rates. There is cross-subsidization among the members in a ten-year age group, but the differences, though greater than for the five-year rate age group, may not be large enough to recommend that single-year rates be used based on cross-subsidization. Nonetheless, for an individual member the differences in the rate can make a noticeable difference in the cost. A member who is age 50 and earning $70,000 would pay about $2,500 more for each year of service under the current ten-year tiered rate than the single-year tiered rate. The same member who was instead 59 and earning $70,000 would pay $1,750 less under the current ten-year tiered BS 18

7 Page 7 rate than under the single rate. The total difference between in the ten-year tiered rate and the single-year rate is $4,200 for each year of service purchased in this example. Increases or decreases in the cost to purchase service credit of this magnitude could, in theory, change the buying decisions of CalSTRS members. Fiscal Impact of Adopting the Five-Year Tiered or the Single-Year Rates In order to examine the potential fiscal impact of adopting a five-year tiered or single-year rate structure, it will be useful to look at the trends in permissive service purchases, excluding nonqualified service, which can no longer be purchased. Trends in Permissive Service Purchases. The following chart below show the annual number of purchases and contribution rates by 10-year age group since BS 19

8 Page 8 Total Purchase Purchase Age , ,081 3,230 4,074 5,645 1,210 1, ,034 1, ,519 1,762 1, or over Total 899 1,725 5,248 6,639 8,964 2,055 1,927 1,894 1,993 2,144 1,940 1,161 Board Approved Purchase Rates Age % 17.1% 18.1% 17.2% 17.1% 17.7% 17.1% 15.5% % 17.9% 18.3% 17.2% 17.7% 18.4% 17.9% 17.0% % 20.1% 20.8% 19.9% 20.5% 21.4% 21.0% 20.3% % % % % % 24.3% 25.3% 24.3% 25.2% 26.5% 26.2% 25.4% % 28.1% 28.5% 28.1% 28.4% 29.8% 29.6% 29.2% 70 or over 23.4% 23.2% 23.2% 23.2% 23.1% 24.0% 24.0% 24.2% As the chart and table show, the total permissive service purchases increased three-fold in when compared to and continued to increase in From through , there was a large decrease in the number of service purchases. It is likely that the single biggest reason for the surge was the board s decision, effective July 1, 2004, to establish a tiered contribution rate schedule for permissive service credit, rather than a single rate applicable to all members, regardless of age. This had the effect of increasing the future contribution rate for older members, who are the dominant purchasers of permissive service. Although the statistics appear to indicate that the increase in purchases peaked in , after the tiered rates took effect, most of those purchases were in fact submitted in , under the older rate structure, and because of the volume of purchases, were not processed until As the chart indicates, the higher rates of older tiers, especially at ages of 40 and above, has resulted in a substantial reduction in the purchase of service credit since the tiered rates took effect. In addition, many older members who would have normally waited several years to buy service under the old rate structure, purchased early to avoid the July 1, 2004, rate increase. By far, most members purchase service when they are in their fifties. The sixties are the second most purchased decade, followed by the forties. This distribution of service purchases is true for all years since 2000, even throughout the steep decline in purchases after Members purchase service during their fifties and sixties even though these are the ages when the service is most expensive to purchase. BS 20

9 Page 9 Purchases Within the Ten-Year Tiered Rate,. The following tables show the average and median ages for the purchase of service credit by age group. I. Average Age and Median Age of All Purchases Age Group Count Average Age at Median Age at Purchase Purchase , , , , >= Total 36, II. Average Age and Median Age of Tiered Rates Purchases Age Group Count Average Age at Median Age at Purchase Purchase , , , , >= Total 22, III. Average Age and Median Age of Single Rate Purchases Age Group Count Average Age at Median Age at Purchase Purchase , , , >= Total 14, Tables I, II, and III show purchases both before and after the single rate replaced by the ten-year tiered rates. According to tables II and III above, the median ages when purchases were made under tiered rates and single rate purchases are nearly the same, and are identical for the members between age 50 and 70, when almost 85 percent of the purchases since took place. The results in these tables, therefore, do not support the hypothesis that members might delay purchases to the end of the age group in order to take advantage of the ten-year age group rate. Part of the reason is members do not always have complete control over when they may BS 21

10 Page 10 purchase service credit. Unlike nonqualified service credit, members may only purchase permissive service credit for a variety of activities that were previously undertaken by the member, and until those activities or events occur, members are not able to purchase permissive service credit. Fiscal Impact of Permissive Service Purchases Under the Alternative Rate Structures. It is difficult to determine the exact fiscal impact of using a five-year or single-year rate. On the one hand, the methodology by which the rates are developed under the current ten-year increment reflects the average cost of the actual purchases made by members in the respective ten-year cohort. For example, the rate for those between 40 and years old represents the weighted average rate applicable to each 40 year old, each 41 year old, each 42 year old, and so on. As a result, even if a different contribution rate was established for each year of age, the amount collected by CalSTRS would have been the same as long as the member s decision at what age to purchase service credit didn t change as a result of the new rate structure. Based on prior experience, when the pricing structure changed from a single rate to ten-year increments, it does not appear likely that any change in member behavior is likely to be significant. Thus, it is likely that the fiscal impact to the DB Program of using one of the alternative rate structures would be small. Implementing a New Rate Structure. CalSTRS staff is in the process of expanding the table in START that drives permissive service rates for both 2 percent at 60 and 2 percent at 62 members to accommodate automatic application of permissive service purchase rates in increments as short as one year. This work will be completed by January 1, 2014, if the committee decides to adopt such a rate structure now or in the future. Consequently the technology costs associated with changing the rate structure at this time would not be significant. However, increasing the number of contribution rates would complicate communications with our members and training of CalSTRS staff. Selecting the Optimal Rate Structure The best pricing structure depends on the board s objectives. If the board s primary concern is the fiscal impact to the DB Program, then any of the pricing structures shown today would meet that goal. Nonetheless, if a change in the pricing structure would result in a change in the age at which members purchase permissive service, the single-year rate structure is the best and fastest at adapting to those changes, followed by the five-year tiered rate structure and then by the ten-year rate structure. As indicated earlier, however, prior experience does not indicate that member behavior would change significantly as a result of a change in pricing structure. If the board s primary concern is about equity among the members, then the single-year rate structure is the best at obtaining that goal, followed by the five-year tiered rate structure and the ten-year rate structure, although for over 80 percent of the members purchasing service credit, the difference in the contribution rate between the current tenyear rate structure and a single-year rate structure is less than two percentage points. BS 22

11 Page 11 If the board s primary concern is with transparency and member understanding of the cost of the benefit, then the ten-year tiered rate structure best meets this goal followed closely by the five-year rate structure, because it is the easiest to communicate. RECOMMENDATION Although adopting a single-year or a five-year rate would be more equitable than the current ten year rate, the differences are not significant, and the greater the number of tiers of pricing, the more complicated it is to communicate the rates to members. Moreover, given how the current ten-year rates are computed, changing to a single-year or five-year rate structure would not affect DB Program costs, assuming there would be no significant impact on the age at which members purchased their service credit. Consequently, staff recommends the board continue to use the current ten-year rate structure, and that the board adopt the same rates for the balance of the fiscal year that it adopted for the period July 1, 2013 through December 31, Average actuarial cost as percent of pay Age group CalSTRS 2% at 60 (Eligibility before January 1, 2013) CalSTRS 2% at 62 (Eligibility after December 31, 2012) % 13.8% % 15.0% % 18.3% % 22.9% % 27.1% 70 and above 24.7% 24.7% Total 26.5% 23.4% BS 23

12 BS 24

13 Attachment 1 Benefits and Services Committee Item 3 PROPOSED RESOLUTION OF THE TEACHERS' RETIREMENT BOARD BENEFITS AND SERVICES COMMITTEE SUBJECT: Contribution Rates for Permissive Service Credit RESOLUTION NO. WHEREAS, Section of the Education Code specifies use of the Contribution Rate for Additional Service Credit be used to purchase permissive service credit in the Defined Benefit Program; and WHEREAS, Section of the Education Code defines the development of the contribution rate for additional service credit, and gives the Teacher Retirement Board the authority to determine the Contribution Rate for Additional Service credit; and WHEREAS, a study has been performed to develop a recommended contribution rate for additional service credit based on age tiered rates; and WHEREAS, in September 2003 the Teachers' Retirement Board adopted a policy that the Contribution Rate for Additional Service Credit be determined based on the actuarial cost of the permissive service credit purchased after an actuarial evaluation to be set based on 10-year age ranges, and separated into two sets, based on the two distinct benefit structures; and WHEREAS, members who are first hired on or after January 1, 2013 are subject to a lower benefit structure than those hired before January 1, 2013; and WHEREAS, the board had previously adopted contribution rates for period July 1, 2013 through December 31, 2012; and WHEREAS, the Teachers Retirement Board, has renewed the actuarial study, therefore, be it BS 25

14 RESOLVED that the Teachers Retirement Board adopts the following Contribution Rates for Additional Service Credit by age ranges effective January 1, 2014 for the 2% at 60 benefit structure and the new 2% at 62 benefit structure: Contribution Rates Age Range 2% at 60 2% at % 13.8% % 15.0% % 18.3% % 22.9% % 27.1% 70 and above 24.7% 24.7% Adopted by: Benefits and Services Committee On Reviewed by: JACK EHNES Chief Executive Officer Brian J. Bartow General Counsel BS 26