ࡱ>  bjbj:: (2PP|O'''''t'.jgUUUU000,............./2b..00004..hUUC.hhh8UU,.h0,.hhh5A'hLY.0.h2h2hhh\000 :    The Determinants of Defined Contribution Retirement Plan Participation Over Time By: Joelle Saad-Lessler and Teresa Ghilarducci This study analyzes the impact of unemployment spell length and incidence on changes in defined contribution (DC) plan participation, controlling for changes in demographic, economic, and spousal characteristics, as well as changes in plan details using longitudinal data on the same individuals from 2009-2012. We apply a Oaxaca-Blinder decomposition on the results and find that the most important factors in determining the change in DC plan participation between 2009 and 2012 were the change in employer contributions to DC plans, the change in spousal participation, and the increase in the number of weeks spent unemployed or out of the labor force. These findings underscore the importance of keeping track of the impact of unemployment on subsequent DC plan participation. Draft 12-27-2013 Introduction The current retirement savings structure in the U.S. depends on active participation in a retirement plan. Unfortunately, 47% of American workers did not have access to a retirement plan at work in 2012. Moreover, even when they were sponsored for a retirement plan at work, 15% of American workers chose not to participate in such a plan. This means that 55 percent of U.S. workers did not participate in employer-sponsored retirement plans in 2012, either because their employer did not offer one or the employee did not participate for voluntary or involuntary reasons. This paper seeks to better understand the determinants of retirement account participation. In particular, we focus on defined contribution (DC) retirement plans, since these are the most prevalent type of sponsored plans, and since defined benefit plan participation is typically mandatory. Previous research analyzed the roles of individual and household characteristics, life events, DC plan design, and the business cycle on DC plan participation. Our research focuses on the impact of individual non-employment spell incidence and duration on DC plan participation, controlling for changes in individual and household characteristics, life events and plan design. We estimate these impacts using a first differences model applied to longitudinal data from the Survey of Income and Program Participation (SIPP) for 2009-2012. This allows us to control for unobserved heterogeneity in individual savings behavior and for unobserved individual preferences. We then apply a Oaxaca-Blinder decomposition on the coefficient estimates to evaluate the roles that changes in individual and household variables, life events, plan design and unemployment spells have had on the nations DC plan participation rate over time. Recent research found that job tenure has declined in the U.S. economy; as a result, workers experience more job switches, with periods of unemployment in between. Moreover, dips in the business cycle translate into unemployment spells for workers. The results of our research highlight the long-term impacts of these unemployment spells on retirement preparedness, because a dip in participation after an unemployment spell impacts long term retirement preparedness many years after the spell has ended. The response of retirement account participation to dips in the earnings profile is an important determinant of whether workers can save enough for retirement using a DC savings vehicle. Literature Review There is an extensive literature on the determinants of retirement plan participation. Studies find that participation depends on lifetime earnings, age and education (Dushi, Iams and Tamborini 2011, Bassett, Fleming and Rodriguez 1998). Job tenure also matters (Even and MacPherson 2004, Munnell, Sunden and Taylor 2002), as well as having kids, marginal tax rates, contribution limits and home ownership (Joulfanian and Richardson 2001, Engelhardt and Kumar 2007). In addition, plan details, such as the availability and level of an employer match, the possibility of taking out a loan against the retirement account balance, choice over the investment portfolio, eligibility to participate and whether workers are auto-enrolled are enormously important in predicting whether workers choose to participate (Laibson and Madrian 2004, Cunningham and Engelhardt 2002, Papke 2004, Purcell 2009). More recent studies have also looked at the impact of business cycles on retirement plan participation (Butrica and Smith 2012, Muller and Turner 2011, Poterba, Venti and Wise 2013). This paper focuses on the impact of non-employment spells and their length on the participation decision of individuals, controlling for changes in individual and household characteristics, life events and plan design. People save for retirement for different reasons ranging from a precautionary motive, to a bequest motive, or to smooth consumption over the life cycle (Browning & Lusardi 1996.) Some people are target savers while others follow rules of thumb. Because peoples savings preferences are different in ways that are unobserved, the role of unobserved heterogeneity in studies of saving behavior is paramount. One way to control for the unobserved heterogeneity is to correct for it directly. This assumes that we know exactly how the unobserved heterogeneity affects saving behavior. For example, Karamcheva and Sanzenbacher (2010) control for selection into a job that offers a DC plan at work and find that ignoring the selection overestimates the fraction of uncovered people who would participate, and underestimates the impact of income on participation. This method fails if selection into jobs that offer a DC plan is NOT the only dimension along which unobserved heterogeneity affects savings. A more effective way to address unobserved heterogeneity is to use a fixed effects model with longitudinal data on individuals. The majority of studies of retirement account participation rely on cross sectional data, because of data availability constraints. Muller and Turner conduct a similar longitudinal analysis of participation and contribution behavior using PSID data from 1999-2005. However, the data they use do not include information on retirement plan details, such as employer matching, loan availability, and investment choices; Moreover, their panel ends in 2005. Butrica and Smith also conduct a longitudinal analysis using SIPP data from the 1996, 2001, 2004 and 2008 panels linked to administrative data on earnings and contribution amounts. Because their data constitute a synthetic panel built on pooled data from multiple years, they cannot run a fixed effects model to control for unobserved worker heterogeneity. In addition, they only have information on unemployment spells when respondents do not work at all during the year. In such a case, the Social Security Earnings record clearly indicates that the respondent was unemployed. But what about shorter unemployment spells lasting a couple of weeks? We use the most recent SIPP data from the 2008 panel, waves 3 and 11, which follow the same individuals from April-July 2009 through December 2011-March 2012. This data include information on job characteristics, plan details, individual and household characteristics, in addition to tracking whether individuals lost their job at any time between the two waves and for how many weeks. After estimating the impacts of all these on retirement plan participation, we apply a Oaxaca-Blinder decomposition to evaluate which changes were most instrumental in explaining the change in retirement plan participation from 2009 through 2012. Data and descriptive statistics This study uses data from the third and eleventh waves of the 2008 panel of the Survey of Income and Program Participation that were fielded in April-July 2009 and December 2011-March 2012. We narrow the sample to only include individuals ages 20-69 who were in the sample continuously from the third wave through the eleventh wave, who had a job working at least 30 hours per week, were wage and salary workers, were sponsored for a retirement plan at work in the third wave and in the eleventh wave and who did not work in the government sector in either the third or the eleventh wave. Then we look at participation at both points in time. We find that 7.6% of the sample participated in 2009, but did not participate in 2012, 24.8% of the sample did not participate in 2009, but participated in 2012, and 67.6% did not change their participation behavior between the two time periods; the latter consisted of 17% who did not participate, and 83% who participated in both periods. The SIPP survey asks those who do not participate in a retirement plan at work their reasons for not participating. Table 1 lists the reasons given, and it reveals that the top reasons were that the employee didnt work long enough for their employer to be eligible, couldnt afford to contribute, didnt want to tie up money, or some other reason. We look at periods of non-employment and find that 6.9% of the sample experienced some non-employment between the two periods; 7% of the sample experienced unemployment at least one week over the sample period, while 6.7% were out of the labor force at least one week over the sample period. The length of these non-employment spells ranged from 1 to 111 weeks. We evaluate when these non-employment spells occurred and find that 75% were in December 2010-November 2011, within a year prior to the second wave of data. The latter is important, because non-employment spells that happen within one year of a new job are likely to negatively affect participation because of eligibility rules for participating. In fact, an analysis of the reasons for non-participating for those who experienced an unemployment spell at any time over the sample period, or specifically in the past 12 months finds that not having worked long enough at this employer jumps to 40% and 50%, respectively, for those who stopped participating. We also look at changes in plan details. The trouble with plan details is that they are only available when workers participate in a plan. This limits our ability to learn whether changes in plan details impacted workers decision to begin participating or to drop participation. We address this problem by generating a variable that contains average plan details by industry, US state and year for all those who participated in a retirement plan at work. We use this variable as a proxy for plan details for each worker based on the workers industry and state of residence. We find that 53% of retirement plans in which employees participated stopped letting workers choose their investment portfolios, while 45% started allowing their workers to choose their investments. 54% of retirement plans in which employees participated stopped letting workers take out a loan against their balances, while 43% started allowing their workers to take out a loan. 38% of retirement plans in which employees participated stopped offering an employer match, while 58% started offering a match. 59% of workers indicated that they lost DB coverage, while 41% gained new DB coverage. The size of the employer contribution fell for 26% of the sample, while it rose for 74%. 40% of those who participated stated that their employer stopped offering something other than a standard percentage of salary contribution toward their plan, while 43% reported that their employer started offering something other than a standard percentage of salary contribution toward their plan. Model We posit a model where participation is a linear function of a persons fixed characteristics, as well as those elements that vary over time, such as their economic characteristics, the characteristics of their employment, their household (in particular, their spouse), details from the DC plan offered by their employer, and their employment history.  EMBED Equation.3  (1) Where  EMBED Equation.3  is a 0-1 indicator of DC plan participation for individual i at time t,  EMBED Equation.3 represents the individuals fixed characteristics, both observed and unobserved,  EMBED Equation.3  are the variable characteristics, and  EMBED Equation.3 is the error term. In first differences, (1) becomes  EMBED Equation.3  Where  EMBED Equation.3  (2) The fixed characteristics drop out, leaving only the variable elements. This equation is estimated using OLS. Its results are efficient and consistent as long as the fixed effects that drop out are correlated with the independent variables that are included in the model. Once we estimate the  EMBED Equation.3 s, we plug them into a Oaxaca-Blinder equation to evaluate the importance of each included independent variable in explaining the variation in participation rates over time. The Oaxaca-Blinder decomposition is:  EMBED Equation.3 = EMBED Equation.3  (3) where the first part of the right hand side is the explained portion of the change, because it is attributed to changes in the variables included in the model. The second half of the right hand side is the unexplained portion, due to changes in the effects of these variables on the participation rate. We only have estimates of b that are free of any unobserved individual heterogeneity from the 2009-2012 years because those are the only two years of longitudinal data on retirement plan details in the SIPP data. We will thus use the estimates of b we have, and assume that the bs do not change over time. The decomposition equation becomes  EMBED Equation.3 = EMBED Equation.3  (4) We use this decomposition to evaluate the roles of the variables included in the model in explaining the changes in DC plan participation over the years 2009 to 2012 using the detailed breakdown of the explained Oaxaca-Blinder component as discussed in Even and MacPherson (1990):  EMBED Equation.3  (5) . Regression Analysis In this section, we conduct an empirical investigation of the impact of non-employment spells and changes in retirement plan details on retirement plan participation using a first difference regression, controlling for changes in earnings, household income, firm size, union coverage, home ownership, citizenship, education, birth of children, changes in marital status, health insurance coverage, benefits receipt, and spousal participation, work status and retirement plan contributions. As discussed above, all plan detail information derives from predicted plan details based on each workers industry and state of residence for each year. We conduct the analysis for all wage and salary workers ages 20-69. If we assume that the determinants of participation are the same whether the employee drops participation or starts participating, we can code the change in participation to be zero if there was no change, 1 if the worker began participating, and -1 if the worker stopped participating. Then we can run a first difference regression of the change in participation on all the other changes to identify the impact of each change. Alternatively, if the impacts of the variables included in the model differ for those who drop participation from those who begin it, we would study the two decisions separately, and code the probability of starting participation vs. not changing, and, separately, the probability of ending participation vs. not changing. Preliminary results are listed in table 2 and they show that the likelihood of participating increases with the birth of each child, as was found in Butrica and Smith; this is driven by an increase in participation, not a drop in the likelihood of stopping existing participation. Participation also increases when workers move to bigger firms; this makes sense since bigger firms tend to have better retirement plan coverage, and more of the workforce participates, which may induce a peer effect. Participation increases when workers get health care coverage; this may be an indicator that there is less of a liquidity constraint on the worker, or it may be a correlate of having access to better DC plan coverage. On the other hand, receipt of government transfers is associated with a decreased likelihood of participating, likely because there is then a liquidity constraint. Increasing ones education seems to lead to a drop in participation, but only for those who already have a college degree and are getting further education; this is probably due to liquidity constraints associated with going back to school. Looking at spousal characteristics, we find that the likelihood of participating increases significantly when the other spouse participates; this is the phenomenon of savers marrying other savers. On the other hand, participation drops as the amount contributed by the other spouse increases. This indicates that spouses share a pot of money, and when one spouse contributes more to their plan, that leaves a lower likelihood for the other spouse to participate. Finally, the likelihood of participating decreases when the other spouse works; were not really sure how to understand this effect. Focusing on the impact of plan details, we find that allowing workers to choose their investment portfolio, take out a loan against the DC plan balance, and offering a match have no significant impact on the likelihood of participating in a DC plan at work. This runs counter to the findings of other studies in the literature. On the other hand, having DB plan coverage decreases the likelihood of participating in a DC plan; this makes sense because we are specifically looking at the likelihood of participating in a DC plan at work. When workers are already participating in a DB plan, there is less reason for them to save additionally using a DC plan. The size of the employer contribution to the DC plan, as well as whether the employer offers some other, non-salary based contribution to the DC plan have a positive effect on participating, and they decrease the likelihood of dropping participation; these are the only plan details that have a significant impact on DC plan participation. This implies that there is a role that employers can play in increasing workers participation in the DC plans they offer. Looking now at the effect of unemployment spells, we find that the incidence of non-employment increases the likelihood of dropping participation. But it also increases the likelihood of taking up new participation, leaving the overall effect insignificant. Whether the unemployment spell was recent or not has no impact, but that may be because the whole sample period is two years, which is short enough to be considered recent anyway. Finally, the number of weeks spent unemployed strongly decreases the likelihood of participating, and that effect comes primarily through workers who used to participate but then dropped participation. The number of weeks spent out of the labor force has the same effect as the number of weeks spent unemployed. Decomposition analysis The regression analysis identifies whether each of the variables included in the model has a statistically significant impact on DC plan participation. However, this does not reveal how economically important each of the impacts are. Table 3 lists the most important factors in DC plan participation changes from 2009-2012. It shows that out of a 14 percentage point increase in DC plan participation from 2009-2012, the biggest driver was the increase in employer contribution, followed by an increase in spousal participation rates. The increase in the number of children born also drove an increase in participation, as well as the increased health insurance coverage. The work status of a spouse brings about a decrease in participation, but as the number of working spouses dropped during the sample period, this resulted in a slight increase in participation. Receipt of government benefits is associated with decreased DC plan participation; but as the incidence of such receipt dropped over the sample period, this was responsible for a dip in participation. The availability of an employer non-salary contribution to a DC plan brings about an increase in participation; but the incidence of such non-salary contributions fell, leading to a drop in participation. Larger firm size also leads to higher participation, but here also, a drop in average firm size over the years brought with it a drop in participation. Weeks spent out of the labor force and weeks spent unemployed both are associated with a drop in participation. There was an increase in both weeks spent unemployed and out of the labor force, resulting in a significant decrease in participation. The most important of these factors in determining the change in DC plan participation between 2009 and 2012 were the change in employer contributions to DC plans, the change in spousal participation, and the increase in the number of weeks spent unemployed or out of the labor force. These findings underscore the importance of keeping track of the impact of unemployment on subsequent DC plan participation. Conclusion This study is the first to conduct an analysis of the impact of unemployment spell length and incidence on changes in participation, controlling for changes in demographic, economic, and spousal characteristics, as well as changes in plan details using longitudinal data from 2009-2012. Our analysis finds that life events, like the birth of a child, positively impact participation, but divorce, marriage or widowhood have no impact. There is also evidence that there are liquidity constraints that prevent participation, because acquiring education for those who already have a college degree is associated with a reduced likelihood of participating, and because receiving government transfers reduces the likelihood of participating. We find that spousal participation strongly predicts positive participation, but there seems to be a sharing of resources in the household, such that as the contribution of one spouse increases, the participation of the other spouse decreases. We also find that the incidence and size of the employer contribution does increase worker participation, but details of the DC plan, such as the choice over investment portfolio, the availability of a match, and allowing workers to take out a loan, do not matter. Finally, we find that non-employment spells negatively impact participation when workers find another job with access to a DC plan. We conduct a Oaxaca-Blinder decomposition on the results and find that the most important factors in determining the change in DC plan participation between 2009 and 2012 were the change in employer contributions to DC plans, the change in spousal participation, and the increase in the number of weeks spent unemployed or out of the labor force. These findings underscore the importance of keeping track of the impact of unemployment on subsequent DC plan participation. Table 1. ReasonDid not participate both periodsStarted participatingStopped participatingNo one in my type of job is allowed in the plan0.951.664.23Don't work enough hours, weeks or months per year3.81.332.24Have not worked long enough at employer to be eligible10.788.0417.41Started job too close to retirement0.160.170Too young0.480.170Cannot afford23.148.0411.19Don't want to tie up money9.034.154.98Employer doesn't contribute, or contribute enough2.691.741.24Don't plan to be on job long enough0.630.661.24Don't need it2.060.751.74Have an IRA or other pension plan coverage0.9511.99Spouse has a plan0.950.580.75Haven't thought about it3.962.241.74Other reason8.724.066.47 Table 2Total Change in ParticipationIncrease in Participation vs. no ChangeDecrease in Participation vs. no ChangeCoef.Std.Coef.Std.Coef.Std.Employee/Employer characteristicsNumber of kids0.03990.0204*0.03180.0173*-0.00690.0134 Firm size (Big firm -100+ workers vs small firm - <100)0.03400.0193*0.02160.0167 -0.01190.0123 Change in education for those with a college degree-0.14410.0819*-0.04660.0726 0.10160.0536*Receipt of benefits status-0.03940.0231*-0.02690.0198 0.01170.0149 Health insurance coverage0.13360.0365***0.09190.0324***-0.04720.0242*Spousal characteristicsIndicator that the spouse works-0.05360.0243**-0.05270.0207**0.00320.0157 Indicator that the spouse participates in a retirement plan at work0.14040.0210***0.11760.0180***-0.02540.0134*Amount of spouse's contribution to his/her retirement plan as a % of salary-0.00210.0018 -0.00390.0015**-0.00190.0011*Unemployment historyNumber of weeks spent unemployed-0.00340.0012***-0.00090.0012 0.00220.0008***Number of weeks spent out of the labor force-0.00280.0016*-0.00040.0016 0.00200.0010**Incidence of non-employment-0.00910.0403 0.10630.0365***0.12460.0262***Incidence of non-employment in past year-0.04300.0510 -0.06250.0464 -0.02070.0328 Changes in plan detailsAllows employees to choose their investment portfolio0.01520.0588 0.00000.0503 -0.01020.0381 Allows employees to take out a loan against their retirement plan balance0.00650.0506 0.03820.0432 0.03270.0325 Offers an employer match0.00850.0313 -0.00580.0267 -0.01270.0199 DB plan coverage-0.10080.0819 -0.12860.0698*-0.03280.0527 Employer contribution as a % of salary0.01570.0060***0.00330.0051 -0.01170.0037***Employer offers some other contribution to the plan0.64820.3273**0.22270.2772 -0.43580.2116**Intercept0.13200.0085***0.21620.0073***0.08660.0053***Observations538149884218R-Squared0.030.020.04 Table 3. Oaxaca-Blinder Decomposition of change in DC plan participation 2009-2012 Impact on DC plan participationBetadXEmployer contribution as a % of salary0.101***positivepositiveIndicator that the spouse participates in a retirement plan at work0.096***positivepositiveNumber of kids0.051**positivepositiveHealth insurance coverage0.005***positivepositiveReceipt of benefits status0.004*negativenegativeIndicator that the spouse works0.003**negativenegativeEmployer offers some other contribution to the plan-0.002**positivenegativeFirm size (Big firm -100+ workers vs small firm - <100)-0.012*positivenegativeNumber of weeks spent out of the labor force-0.030*negativepositiveNumber of weeks spent unemployed-0.061***negativepositive References Bassett, William F., Michael J. Fleming and Anthony P. Rodrigues. 1998. "How workers use 401(k) plans: the participation, contribution, and withdrawal decisions." National Tax Journal 51(2): 263-289Beshears, Choi J.J., Laibson D. and Madrian B.C. 2007. "The impact of employer matching on savings plan participation under automatic enrollment." NBER working paper 13352Browning, M., & Lusardi, A. 1996. Household saving: Micro theories and micro facts. Journal of Economic Literature, 34(4), 1797-1855.Butrica, Barbara A. and Karen E. Smith. 2012. "401(k) participant behavior in a volatile economy." Center for retirement research working paper #2012-24.Choi J., Laibson D. and Madrian B. 2004. "Plan design and 401K savings outcomes." National Tax Journal , Volume: 57, Issue: 2, p275-298Cunningham Christopher R. and Gary V. Engelhardt. 2002. "Federal tax policy, employer matching, and 401K saving: evidence fro HRS w-2 records. National Tax Journal Vol. LV No. 3, pp. 617-645Clark Robert, Jennifer A. Maki and Melinda Sandler Morrill 2013. "Can simple informational nudges increase employee participation in a 401(k) plan? NBER 19591Dushi Irena, Howard M. Iams and Christopher R. Tamborini. 2011. "Defined contribution pension participation and contributions by earnings levels using administrative data." Social Security Bulletin, Vol. 71, No. 2.Dushi Irena and Howard M. Iams. 2010. "The impact of response error on participation rates and contributions to defined contribution pension plans." Social Security Bulletin, Vol. 70, No. 1Dworak-Fisher Keenan. 2008. "Encouraging participation in 401(k) plans: reconsidering the employer match." BLS working papers #420Engelhardt G. V. and Kumar A. 2007. "Employer matching and 401(k) saving: evidence from the Health and Retirement Study." Journal of Public Economics, Vol. 91, Issue 10, pp. 1920-1943.Engen, Eric M; Gale, William G; Uccello, Cori E. 2005. Effects of Stock Market Fluctuations on the Adequacy of Retirement Wealth Accumulation. Review of Income and Wealth 51.3: 397-418.Even William E. and David A. Macpherson 2004. "Determinants and effects of employer matching contributions in 401K plans.", working paperHuberman Gur, Sheena S. Iyengar and Wei Jiang. 2007. "Defined contribution pension plans: determinants of participation and contribution rates." Journal of financial services research 31(1), 1-32Ippolito Richard A. 2002. "Stayers as workers and savers: toward reconciling the pension-quit literature." Journal of Human Resources, Vol. 37, No. 2, pp. 275-308Joulfanian, David and David Richardson. 2001. "Who takes advantage of tax-deferred savings programs? evidence from Federal Income Tax data." National Tax Journal LIV(3): 669-688Karamcheva Nadia and Geoffrey Sanzenbacher. 2010. "Pension participation and uncovered workers." Unpublished draft, Boston CollegeKiichi Tokuoka . 2013. "Saving response to unemployment of a sibling." Journal of Economic Behavior and Organization 89 (2013): 58-75Kusko Andrea L, James M. Poterba and David W. Wilcox. "Employee decisions with respect to 401(k) plans, in Living with Defined Contribution Pensions: Remaking Responsibility for Retirement, edited by Olivia S. Mitchell and Sylvester J. Schieber 98-112. Philadelphia, Pennsylvania: Pension Research Council, the Wharton school of the University of Pennsylvania, and the University of Pennsylvania Press, 1998 Mitchell O.S., Utkus S. P. and Yang T. 2005. "Turning workers into savers? Incentives, liquidity, and choice in 401(k) plan design", NBER working paper 11725.Muller, Leslie A, and John A. Turner. 2011. "The persistence of employee 401(k) contributions over a major stock market cycle: evidence on the limited power of inertia on savings behavior." UpJohn working paper 11-174. Kalamazoo, MI: W.E. UpJohn Institute for Employment Research. Munnell Alicia H., Annika Sunden and Catherine Taylor. 2002. "What determines 401K participation and contributions?" Social security Bulletin, Vol. 64, no. 3, 2001/2002 Munnell Alicia H. and Annika Sunden. 2006. "401K plans are still coming up short." Issue Brief No. 43. Chestnut Hill, MA: Center for Retirement Research at Boston College. http://crr.bc.edu/briefs/401_k_plans_are_still_coming_up_short.html.Papke, Leslie E. 2004. "Choice and other determinants of employee contributions to defined contribution plans." Social Security Bulletin 65(2): 59-68Poterba, James, Steven Venti and David A. Wise. "The financial crisis and savings in personal retirement accounts", (2013), NBER working paper ONB-13-02Purcell Patrick, 2009. Retirement plan participation and contributions: trends from 1998-2006. CRS report for Congress 7-5700. Washington DC: Congressional research ServiceSmith, Karen E., Richard Johnson, and Leslie Muller. 2004. "deferring income in employer sponsored retirement plans: the dynamics of participant contributions". National Tax Journal. 57(3): 639-670.Turner Jon, Leslie Muller and Satyendra K. Verma. 2003. "Defining participation in defined contribution pension plans. Monthly Labor Review, Aug 2003  Each panel of SIPP data follows the same individuals over approximately 4 years. Prior to the 2008 panel, the retirement expectations module was only fielded once per panel. The 2008 panel is the first time that the retirement expectation module, which contains details on retirement plans, was fielded twice in the same panel. This allows us to run a longitudinal analysis including details on retirement plan details.  These are years when the retirement expectations module was fielded in the SIPP 2008 panels.  Only statistically significant effects were listed. Among the insignificant variables were changes in earnings, buying a home, changes in household total income, changes in household net worth, divorce, marriage or widowhood, and changes in education for workers with a HS degree or with some college.  Firm size is coded as small for firms with less than 100 employees, and big for firms with 100+ workers. This is due to the fact that the coding of firm size in the data changed from 2009 to 2012. This new coding is the only way to accord data on form size across the two years. Z J j  M n z Y y OP˲qbSSDDDDDh[?JhO OCJOJQJaJh(hO OCJOJQJaJhVhO OCJOJQJaJh)hO OCJOJQJaJhO O>*CJOJQJaJhO OCJOJQJaJ0h;aB*CJOJQJaJfHphq 0h)B*CJOJQJaJfHphq h)CJOJQJaJhF`CJOJQJaJhVhF`CJOJQJaJhF`>*CJOJQJaJ Z noH#%%%g(gdO OPQR3<=A{239GJ+++/宒֮y]6hhO OB*CJOJQJaJfHphq 0hO OB*CJOJQJaJfHphq 6h)"uhO OB*CJOJQJaJfHphq 0hO OB*CJOJQJaJfHphq h)"uhO OCJOJQJaJh(hO OCJOJQJaJhO OCJOJQJaJh[?JhO OCJOJQJaJ!g()+/0515756678R8b9=:b:: =Y=s>>>>.AbDHK|OkRlRgdO O//6666666666666ˮˎl]R@/]'hO OOJQJ!jh"U2hO OEHOJQJU#jN h"U2hO OOJQJUVh"U2hO OOJQJjh"U2hO OOJQJUCjhF\shO OB*CJEHOJQJUaJfHphq ?jSN hO OB*CJOJQJUVaJfHphq 9jhO OB*CJOJQJUaJfHphq 0hO OB*CJOJQJaJfHphq 6hhO OB*CJOJQJaJfHphq 66777777 7!7475767777777¥…c¥C?jzN hO OB*CJOJQJUVaJfHphq Cjn h^;hO OB*CJEHOJQJUaJfHphq ?jfV hO OB*CJOJQJUVaJfHphq 9jhO OB*CJOJQJUaJfHphq 0hO OB*CJOJQJaJfHphq h"U2hO O6CJOJQJhO OCJOJQJh"U2hO OCJOJQJ77777777788'8(8)8eE#Cj#h^;hO OB*CJEHOJQJUaJfHphq ?j/N hO OB*CJOJQJUVaJfHphq Cjh^;hO OB*CJEHOJQJUaJfHphq ?j9fV hO OB*CJOJQJUVaJfHphq 0hO OB*CJOJQJaJfHphq 9jhO OB*CJOJQJUaJfHphq Cj hF\shO OB*CJEHOJQJUaJfHphq )8*85868I8J8K8L8w9x99999b:c:v:w:ɩgE=9.j N hO OUVhO OjhO OUCj*h^;hO OB*CJEHOJQJUaJfHphq ?jfV hO OB*CJOJQJUVaJfHphq Cjh^;hO OB*CJEHOJQJUaJfHphq ?jfV hO OB*CJOJQJUVaJfHphq 0hO OB*CJOJQJaJfHphq 9jhO OB*CJOJQJUaJfHphq w:x:y:z:{::::::;;;;<<<<<< =!=4=5=6=7=8=9=L=M=N=O=U=µͭ́ti\j:3hzhO OEHUj N hO OUVj-hzhO OEHUj N hO OUVjhO O0JOJQJUhO OOJQJhwhO OOJQJhO OOJQJjA&hzhO OEHUj N hO OUVhO O0hO OB*CJOJQJaJfHphq jhO OUj hzhO OEHU U========s>t>>>>>>>>DDEEaabbbbmmaRCRh7AhO OCJOJQJaJh9hO OCJOJQJaJhO OCJOJQJaJ=jhO O0JB*CJOJQJUaJfHphq 0hO OB*CJOJQJaJfHphq h;ahO OOJQJ!j9hT|hO OEHOJQJUjN hO OOJQJUVjhO OOJQJU!jhhO O0JOJQJUhO OOJQJhhO OOJQJlRRnSZZZZ`aaaaaaaaaaaab b-bCbYb$d$Ifa$gdO Od$IfgdO OgdO OYbZbbbbbqcRRR$d$Ifa$gdO Od$IfgdO OkdhA$$Ifl \N*`644 laZp(bbbbbbqcRRR$d$Ifa$gdO Od$IfgdO OkdB$$Ifl\N*`'''644 laZp(bbc#cHcTc_ckczccccccd,d;dJdvddddddddde eeeeeeeeffffff8g9gggggghhzh{hhhiiitiuiii)j*jjjjjĹĹh7AhO OCJOJQJaJh7AhO OCJaJ"h7AhO O5CJOJQJ\aJhO OCJOJQJaJh9hO OCJOJQJaJh7AhO OCJOJQJaJCbbccc#cqcRRR$d$Ifa$gdO Od$IfgdO OkdB$$Ifl\N*`'''644 laZp(#c$cHcMcRcTcqcRRR$d$Ifa$gdO Od$IfgdO OkdC$$Ifl\N*`'''644 laZp(TcUc_cdcickcqcRRR$d$Ifa$gdO Od$IfgdO Okd>D$$Ifl\N*`'''644 laZp(kclczccccqcRRR$d$Ifa$gdO Od$IfgdO OkdD$$Ifl\N*`'''644 laZp(ccccccqcRRR$d$Ifa$gdO Od$IfgdO OkdE$$Ifl\N*`'''644 laZp(ccccccqcRRR$d$Ifa$gdO Od$IfgdO OkdcF$$Ifl\N*`'''644 laZp(ccd"d'd,dqcRRR$d$Ifa$gdO Od$IfgdO OkdG$$Ifl\N*`'''644 laZp(,d-d;d@dEdJdqcRRR$d$Ifa$gdO Od$IfgdO OkdG$$Ifl\N*`'''644 laZp(JdKdvd{d}ddqcRRR$d$Ifa$gdO Od$IfgdO OkdH$$Ifl\N*`'''644 laZp(ddddddqcRRR$d$Ifa$gdO Od$IfgdO Okd?I$$Ifl\N*`'''644 laZp(ddddddqcRRR$d$Ifa$gdO Od$IfgdO OkdI$$Ifl\N*`'''644 laZp(ddddddqcRRR$d$Ifa$gdO Od$IfgdO OkdJ$$Ifl\N*`'''644 laZp(dddddddddeeeeqoooooooooookddK$$Ifl\N*`'''644 laZp( ee e+eZeeeeeeeeakdL$$Ifl \$ (;N'6    `'644 laZp(d$IfgdO O eeeeeeeeeeeeeeff ff$d$Ifa$gdO OFkdP$$IflN'`'`'644 laZp FfmNd$IfgdO OfffMfTf[f]fdfkfnfvf}ffffffffffffffFfDZFfvV$d$Ifa$gdO OFfRd$IfgdO Ofg gggg$g'g.g5g8g9gSgZgageglgsgwgggggggFfaFf^$d$Ifa$gdO Od$IfgdO OggggggggggggggggggghhEhLhShWhFf|i$d$Ifa$gdO OFfed$IfgdO OWh^hehihqhxhzh{hhhhhhhhhhhhiiiiiiFfqFfJmd$IfgdO O$d$Ifa$gdO Oiiiiii=iEiLiPiXi_ibiiipitiuiiiiiiiiiFfx$d$Ifa$gdO OFftd$IfgdO Oiiiiiijj jjjj%j)j*jSj[jbjejmjtjwjjjjFfPFf|d$IfgdO O$d$Ifa$gdO Ojjjjjjjjjjjjjjjjjjkk kkkkakFf$d$Ifa$gdO OFfd$IfgdO OFfjjkkkkkk*l+llllll2m3mUmVmumvmmmmmmmn9n}nnnnnno2oRonoooop@p\p}ppppoqȹȝ%h3qphO OB*CJOJQJaJphhO OB*CJOJQJaJphhO OCJOJQJaJhi?LhO OCJOJQJaJhO OCJOJQJaJ"h7AhO O5CJOJQJ\aJh7AhO OCJOJQJaJh7AhO OCJaJ1akhkokrkykkkkkkkkkkkkkkkkkkkklFfVFfd$IfgdO O$d$Ifa$gdO Olllll l'l*l+lRlYl`ldlklrlul}lllllllllFfFf$$d$Ifa$gdO Od$IfgdO Ollllllllm mmmm m'm.m2m3m@mEmFmGmLmMmNmFfFfd$IfgdO O$d$Ifa$gdO ONmSmTmUmVm`memfmgmlmmmnmsmtmumvmwmmmmmmFf*Ff\d$IfgdO O$d$Ifa$gdO Ommn"n&n/n8nqaPBBBd$IfgdO O$d$Ifa$gdO Od$IfgdO Okde$$Ifl\j.#'j'~ 'F'#644 lalp(8n9n}nnnnnYK:KKK$d$Ifa$gdO Od$IfgdO Okd8$$Iflrj.#'j'T'*'F'#644 lalp2nnnnnnnYK:KKK$d$Ifa$gdO Od$IfgdO Okd+$$Iflrj.#'j'T'*'F'#644 lalp2nnnnnnnYK:KKK$d$Ifa$gdO Od$IfgdO Okd$$Iflrj.#'j'T'*'F'#644 lalp2nnooo(o1oYK:KKK$d$Ifa$gdO Od$IfgdO Okd$$Iflrj.#'j'T'*'F'#644 lalp21o2oRoXo[odomoYK:KKK$d$Ifa$gdO Od$IfgdO Okd$$Iflrj.#'j'T'*'F'#644 lalp2monooooooYK:KKK$d$Ifa$gdO Od$IfgdO Okd$$Iflrj.#'j'T'*'F'#644 lalp2oooop ppYK:KKK$d$Ifa$gdO Od$IfgdO Okd$$Iflrj.#'j'T'*'F'#644 lalp2pp@pGpIpRp[pYK:KKK$d$Ifa$gdO Od$IfgdO Okdݳ$$Iflrj.#'j'T'*'F'#644 lalp2[p\p}pppppYK:KKK$d$Ifa$gdO Od$IfgdO Okdд$$Iflrj.#'j'T'*'F'#644 lalp2ppppoqYWWId$IfgdO Okdõ$$Iflrj.#'j'T'*'F'#644 lalp2oqpqrrrWMkd>$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap oqpqyqqqrrrr=r?r@rQrrrrrss@sAs}ssssssssXtYtttttttuu+u,u1u7uuuvv vvvvvvv+w-wFwGwMwmwwwwwxx2x9xpx񾫾񫾫%h3qphO OB*CJOJQJaJphhO OB*CJOJQJaJph%h3qphO OB*CJOJQJaJphhO OB*CJOJQJaJphh3qphO OCJOJQJaJCrr@sAssWMkdN$$IfZ$622 734ap d$IfgdO OMkdƷ$$IfZ$622 734ap sstt+uWMkd^$$IfZ$622 734ap d$IfgdO OMkdָ$$IfZ$622 734ap +u,uvvvWMkdn$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap vvFwGwxWMkd~$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap xxxxIyWId$IfgdO OMkd$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap pxxxxxx7y8yIyJyRyVyyyzzzzKz{z}zzzzz#{R{i{j{{{{{{{|W|t|u||||||||$}%}2}~~~ͺͺͪͺͺͺͺ"hO O>*B*CJOJQJaJph(hJ_hO O>*B*CJOJQJaJphhO OB*CJOJQJaJph%h3qphO OB*CJOJQJaJphh3qphO OCJOJQJaJ%h3qphO OB*CJOJQJaJphhO OB*CJOJQJaJph3IyJyzzzUGd$IfgdO OMkd$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap zzi{j{{WMkd$$IfZ$622 734ap d$IfgdO OMkd&$$IfZ$622 734ap {{t|u|~WMkd$$IfZ$622 734ap d$IfgdO OMkd6$$IfZ$622 734ap ~~~~WMkd$$IfZ$622 734ap d$IfgdO OMkdF$$IfZ$622 734ap ~\~~~km]^uv}ghFGzMNͺͮyh&AhO OOJQJ!jhhO O0JOJQJUhhO OOJQJhO OjhO O0JUhO OCJOJQJaJ%h3qphO OB*CJOJQJaJphh3qphO OCJOJQJaJ%h3qphO OB*CJOJQJaJphhO OB*CJOJQJaJph,uvgWMkd$$IfZ$622 734ap d$IfgdO OMkdV$$IfZ$622 734ap ghWMkd$$IfZ$622 734ap d$IfgdO OMkdf$$IfZ$622 734ap FGWMkd$$IfZ$622 734ap d$IfgdO OMkdv$$IfZ$622 734ap M܇WUSSSSUMkd$$IfZ$622 734ap d$IfgdO OMkd$$IfZ$622 734ap P܇݇އhO OCJOJQJaJhz82hO OOJQJhO OjhO O0JUhO OOJQJhLOhO OOJQJ &1h:pO O/ =!"#$%Dd P  S A? " Bt_˞G *3APD TH_˞G *3A X xVn7*n:41$h/6JChR\TG(ࠧܧ-z3K!ݴ H.:@w;~F н_' x.OdŒ}~j:!ѷh -\^^"tOso Mi_ogqD:\N |LJ,/BeeZu5xj"P(|j=ܕ-+V#Y+֪{n殴d4WBgbذa7&/:fX,^OqI^O(IÂ8mcԸl44bӘ$)mշMbdO x2{!gjZu|P S7N9I MQM:y|C9f]#QG\>vpD?`sII'iE_ԯ5G_9J(K!,]axj y"E0w~Vr4}S5>ɆL KջШ(ۥ$ _N4_d OrtR8 T>V8S 8YU9#_,QD6)FvK~9aڏφa&V$Oq0Όq5mjTčy^#y" WָJ͌P5+vAg40$6&X3ѣ41IOdXe]`3v컌7J.Ij"OMG(&;Q͂lD/ZZvM!eBeȾ75R͈ҁE_ׅ ]G5wS%XLm_+tu5 VnZv1 \[L,`TّUiY3U bDU,6ؤGszct2>e9 n g UI7',JYhe4BrR^OJ֟%0ũD4=Tk ̼x[̓Fp8`]wU:fs?!N]\HR% *f=(RәֆW`: sWWp JoWVB_LnᠽG9NR6I[Z͙={7wo 뎵h-BaSt-6~|xMK@׽\9.'VDd 0  # ABefNmۤ*a%@'\ TefNmۤ*a%@' xtxVn7di&A(PFPZqP}0Pv\Yc @;O[CcOQjipv!|0^B&h߿}60^Es~A_YcE7#OZ],[ke t~Uz@Bk.D iks@/j5:=My[XkЧM9)H!H,𡐗\M%Z91ŀH>R6NIP>KF^l˄V3irf9َfce:QY P,nt0M68ƂuD5ihS@[7VS:Mu.dC0i1!9C;"2jҔzm]c&Z5),d)q*JOOUzHN:D*F@;I8z9v2AH1daɂ!@Njl Y(pGI ]1wYLwa ܌YwfNK nLNweRhk0^Pެ0-0KАgSQ8 /$ <7'8Ê@m=^KuDd @P  S A? " BhK؊qA39 4U ThK؊qA39 4U  sxVn5nmHHDzwI7WeE&RwƳkeڞͮBJwO]ċ\r؞Lڨ0Ώ_>f{y-_#3E7ϧ ƫz/3B+t~b29B+6O.Z].X~}t?/=B)&Tra"KZ{O/kh:>Oظ<:tưKY{!qqwX<ǺbF@ՍAu{~:9h(*tN =(NƧˤB^ck™]STM/SN66 -  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~Root Entry@ F,6AData WordDocument?(2ObjectPoolB3,Aҙ6A_13235036990 F,Aآ,AOle CompObjfObjInfo !$',/2569<?BCDEGHIJKLNOPQSFMicrosoft EquationDNQE Equation.3ation.39q]  Y it =a i +bX it +e itEquation Native y_1323501464+ Ft-A%-AOle ObjInfo Equation Native 7_1449575588!FiG-AV-AOle  CompObj f  Y it FMicrosoft Equation 3.0 DS Equation Equation.39q5  iFMicrosoft EquationDNQE EObjInfo Equation Native  6_1323503738:F u-A-AOle CompObjfObjInfoEquation Native 7_1449575737F-A'-Aquation.3ation.39q  X it FMicrosoft Equation 3.0 DS Equation Equation.39q|;  itOle CompObjfObjInfoEquation Native :_13235036635F-A.-AOle CompObjfObjInfo FMicrosoft EquationDNQE Equation.3ation.39q  Y it -Y it-1 =b(X it -X it-j )+m itEquation Native _1449575899&#Fr.AW!.AOle CompObj"$ f FMicrosoft Equation 3.0 DS Equation Equation.39qZبL  it = it " it"1 FMicrosoft Equation 3.0 DS Equation Equation.39qObjInfo%"Equation Native #v_1449576095(FjF.AV.AOle %CompObj')&fObjInfo*(Equation Native ))_1323502736-F.Al.A 5 FMicrosoft EquationDNQE Equation.3>  (2Y t -2Y t-j )Ole *CompObj,.+NObjInfo/-Equation Native .Z_1323502805 2F.A.AOle 0CompObj131NObjInfo43FMicrosoft EquationDNQE Equation.3  (2X t -2X t-j )b t +(b t -b t-j )2X t-jEquation Native 4_13235030097FP/A/AOle 7CompObj688NFMicrosoft EquationDNQE Equation.3M  (2X t -2X t-j )b tFMicrosoft EquationDNQE Equation.3ObjInfo9:Equation Native ;i_1323794385<FNObjInfo>@Equation Native A 1Table2  Exp j =(2Y t -2Y t-1 )*b(2X jt -2X jt-1 )b(2X jt -2X jt-1 ) j  Oh+'0 y-ESs'Fi19#31"ejJk9m]c"hZ&S3G,+==<%F7r,BV4q2x#lyS;>< `uCPjk$EDg^Ea~$ Pj&̚Ě,wiU`Cg%A@)%rX bHJwE"SxkSoR\6a)sJviRFaQLj)MJI95Sg<3o9͈* ܓ78›e(se+WznZ4EE vG,&*$WN{ۘ lh5/2cD߸6 \bjO`2{r2pq40عK@&حV݀!MLoX؂897@#wT[0CvbzCDTw@SxBe+ !5)5{y~t\9 -$Lpxr͠QFa=⸇.H./,P2}hpܸ>]Fk[/|B€zapK-\N="Dd h\  c $A? ?#" `2r62eV$;٪&NE`!F62eV$;٪&@H|xuPNQ=3 . 4XF 1WLdV&VƂ֎O2B,H\ϽtN2{T;-A_"EM4sz"E_jұY? bs `>fDHI L gMo pf/5d{7ȌsaԻ&!߼f?tǧFk*?|;]fr&0Gk/܅b.{s"t'رUq/aw_\/oVP Dd @P  S A? " B i 4 OKg T i 4 OKz @ xWnF(*DHpQUY0wB"-AD**Y{5x[lH @w Ю|9gΌ֏H~.e":yy[¶gQy~A#]ճ8B ,=NNNdɑ%yjSOHkX'cp _k ÚD@IZ 6.MI:f1XumDa|#.8>?(/4KBHۛO VIA5y2eU "Xdw„{Yj'E/d ؛x]YnGr%V _%BQ&#^GITYH&3H@ox̥KL|h]H `+ytKG~]3&|k‹}Yc$P/!OxdU֌|q?"F֘glpŠ#H%ܸ+nvkjޠ%**Ёx11Kap94Im'g0s1dؘdqOxMQXr@i’6LU5 ܌X4@(gAt!6?]FY6tn[T=EzKJM(NDY_jĎ?Ka!Ȫʣ X6tRks PUG[K^C ini+)kP/@wY#K`.;_iѧ|_6 Ƽ -E;31;(ZiRn٩fuw49 $ ULp€G)> G"ٮܛ%TwEb;Lr33M pHhjy=cQax؝`]=mk;۶;x:8HFm"Ֆ>!ZD)kJrS4~Ƥ\R=cYކfJRu&3Ѩ :_qc8moP6`uН)]A{TvީBh%݈Z(o 嵻&2.3./[( KGnχ泥n=/-Q+:wuDd dh\  c $A? ?#" `2Q BD$PќT`!Q BD$PќT@^ |gxKPƿ{N  "RBWlw[!bX3;;t$ϹI z;{ @RU.Pr"ſMQTU˱W"l.`e/|z +>%췼ӆѝL+&<츭s=zCަ Ni؛}Ey ) ǝ3aeOA1]NV XuNp >{47PV&&zD/ֿc~j{ɜ [*~]ZV*w⼁:JMj]௼C!~ez6~s\i*z8ÔDd @\  c $A? ?#" `2DUSV(Z, n`!USV(Z,:R xcdd`` $X bbd12,(ㆫab`j`3H1g`Yǀh1 憪aM,,He`P&lebabMa``犴(b@LZ6 P.P16J`t`r*$%\`v0y{ngdbR ,.Ie`ԡCga`P:#Dd TP   S A ? " BRsx]&/[b TSRsx]&/ 0L K!xV_OF__hB VCU@-Y;*!*"EEZoiCY:ofgvg?H>sCsYtCSK Go9GOfg}t!} (-t߼cn)9z BnT9HtO^_i4f:dxG [4,sBw1D1@:i Xd\GgIUVep<(yl^gQֿfA %8pF%e{)4QK>EŜ큤~1eGѾe2hpLv쒜#Q-KD>ǚL[2LwƓ(Æh$Fv8.hM|C1f0nY-uFk2j VJ! F' ʡ->!ʰ†h<e.<!ٹs81;4%6{ $ՌؠݰQj1\3$1 ^>"?$EϢL˷ﳧ&j]&B($R!@.!"v]3$&(ఠ#0+F:,䲵ո<^㑈e>^<Ǻ ,gEZU3E4(YQU2L|Qp ayT ڂ=UQM49S.a)kLS J5}ja50V4X4U uM` k&/1ٕۃ\d-DCQT#VrMP 0}_|ϘMaLV|Yff% 1YordM'bütv6l=Kl7 =nE޲3*E 4 AVUdZE26Q9YUqX2D2(!z֖"Y.u19Ѓ lَ:1C}pꃀWkYJf5%`okS3R4)#Jl+G'iye=^N Iu@}5]sC)h!J7MXV0p,sk"S9;2 qj䖡iSA|wbH9͎=AxMs(C j!_/iF\7%B4 ļ@CWH97[_M=aas nVՓ/:3NBp}sAg%\g:?-] {WtDd T TP   S A ? "  B2ŕ {$,& Tŕ {$,KxWoDoB[J m` *kz"TD&R¬=NxnJEƱ '/ qƙ6ed~_\B?K _Z?yD|?C$3`!" C[ukǠb}QM^όeͨ;KX,w87NZ2 F6,y B拹]xj Uk_g"w6r:cX X= _ێvb/8KN6 5`z=;]ӸlpӘG,21-Sߛ$KICi6p$˨&<+3kpopsx[X:)Xq:cp_D7Q߅+Np N1;XqdѠaٽ!Rz\ʐY̡FqiS,ن헷487Ju;64whlt+MQ"d$Kpb4A2;g9n%>/wȯ(5BWr&LlXJO3&q?c F]C2CcW1Pv>Q,I:.ּW$Yi@si] "< gAǔRZ0r‑Y]Q0 `@G?1'&I*ep&3i5H4NTq>M5Դܱ糧ɝʰji5'66&9&6${ǹ㑌w<{i>,Ju͟挱LL҄}1\%L&+b=XCP=Lj"_YXW_:ZYeݮã[|Zٔޛ{SܤIc B}hY:8CLJh[;C 2VUk0(Y;w]LЁ ^'WyX0>J5;O1M0);m?,GYøD. yqacCnoV8\ 8.BjL94uCx|8NY=:5] yӄarvu^n@=t떐X|!і> H+m>7ʼ^S+e$tyv|P#*JF;hjf_04Żmx{-d\0Gm3[zøWo( ?Gף9.\EWoߡ934#Dd TP   S A ? "  BRsx]&/[[- TSRsx]&/ 0L K!xV_OF__hB VCU@-Y;*!*"EEZoiCY:ofgvg?H>sCsYtCSK Go9GOfg}t!} (-t߼cn)9z BnT9HtO^_i4f:dxG [4,sBw1D1@:i Xd\GgIUVep<(yl^gQֿfA %8pF%e{)4QK>EŜ큤~1eGѾe2hpLv쒜#Q-KD>ǚL[2LwƓ(Æh$Fv8.hM|C1f0nY-uFk2j VJ! F' ʡ->!ʰ†h<e.<!ٹs81;4%6{ $ՌؠݰQj1\3$1 ^>"?$EϢL˷ﳧ&j]&B($R!@.!"v]3$&(ఠ#0+F:,䲵ո<^㑈e>^<Ǻ ,gEZU3E4(YQU2L|Qp ayT ڂ=UQM49S.a)kLS J5}ja50V4X4U uM` k&/1ٕۃ\d-DCQT#VrMP 0}_|ϘMaLV|Yff% 1YordM'bütv6l=Kl7 =nE޲3*E 4 AVUdZE26Q9YUqX2D2(!z֖"Y.u19Ѓ lَ:1C}pꃀWkYJf5%`okS3R4)#Jl+G'iye=^N Iu@}5]sC)h!J7MXV0p,sk"S9;2 qj䖡iSA|wbH9͎=AxMs(C j!_/iF\7%B4 ļ@CWH97[_M=aas nVՓ/:3NBp}sAg%\g:?-] {WtVDd TP   S A ? "  B^04X9-;~3 T^04X9-;v A KTxW_oD__CZB,ڪRyU_Hڎz'Hh"Rg7.B%">B)x+|>a5X؞ofg3)54D:gBO]?u OrxG>ܳ󰟢w'"V:°yѰ'*/'"+\7ٱǬ/7ߋʹ<@{g\;O>)ULs"V/oau~u5Vj?rFxHXvA ?1Kp8J1mO E`68N=l&bbiR7S::K.x͋v`IrhKAPh<e.;!չs>81;_ɼa%SI<,*E~I8☋:+ G/}e㙖o;OpȪM`u ]L.FhKD4:֛^ǧ6 G YqdrOn%1uX#0sz 8e9+҂ e,dFU2bFND%t?(ͣjd"jrət LYcjPS 6 Eh 4mjk2}UY11-`094Hm'v&o+V>*dW44K,aF :[榳eeKm>52aГ&_dG[h˺1Ӫ@o P2Tq  k=5 Paɠrob a׫&'3 ]A[YJf5%#Z75i)ÏiMHޕ)nZjY6BBt@r vL~H~Z9$ՠtӄen cf[ٴ؉ilS,% `GYީN+_16{;]뇯]zy24 q-@Ȑo4K#^7~!E!+͊ ZGa+n?إ/GwRI^$VJb;|s|x6sb3g4'Bϋ3c# \}$4[rO7ѱ׷8D!} 񛻻a[=F#3'wx$Q6(S:Qf}yAe5e2X >~|\Z T0S%ihJNϤ[&ZwEKY=t1Qu~Ok"{^ƐWjNZUm"";fgs]  'kx4 4%M6PɟXgGم ^'["zoF8 1A";6R=i8豶S؅҃_JEaݔ'We1m־ Fb:e!YEd rB;ҫܧHWgؾ&R>l&%.Mh"G ]uUfASZIY^JKu ^HG;& TSQiŕ "|P%'\J,qXp"PP갠*yu`+lL17 v6Vl{D]7a,`GMj٨I(brG@1rܒVL|=%F4Yu^4<&UK?:*e&vz,VY/)oVK7aV$*K^uC:3 TjijuT0xrSm0 cXu%{aV-O)`, 8e󯰈3 *62"O ]7/E*j`AZJڕy|d]\Nj9u9TDʀ^ z4L)L[8a.*Wj XKF4o18;_SЗN~?aa8ݛ(c5te MFӂi0:8Bc2l5I C8J ]kZuhLJ Kΐ>6сik5cI/u1`XYvzEtq25}O/@bͶL,)ZPJtp-yg|ݨ޾xun{寵ڊ-MmaN-,BH./FH۬ɁWd, EGTG> v.O7/6666666666666666666666666666666666666666666666666hH666666666666666666666666666666666666666666666666666666666666666662 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~8XV~ OJPJQJ_HmH nH sH tH J`J Normal dCJ_HaJmH sH tH DA D Default Paragraph FontRi@R 0 Table Normal4 l4a (k ( No List H@H c0 Footnote Text dCJaJ<< c0Footnote Text Char@&`@ c0Footnote ReferenceH*N"N 7A0 Balloon Text dCJOJQJaJJ1J 7A0Balloon Text CharCJOJQJaJPK!K[Content_Types].xmlj0Eжr(΢]yl#!MB;BQޏaLSWyҟ^@ Lz]__CdR{`L=r85v&mQ뉑8ICX=H"Z=&JCjwA`.Â?U~YkG/̷x3%o3t\&@w!H'"v0PK!֧6 _rels/.relsj0 }Q%v/C/}(h"O = C?hv=Ʌ%[xp{۵_Pѣ<1H0ORBdJE4b$q_6LR7`0̞O,En7Lib/SeеPK!kytheme/theme/themeManager.xml M @}w7c(EbˮCAǠҟ7՛K Y, e.|,H,lxɴIsQ}#Ր ֵ+!,^$j=GW)E+& 8PK!\theme/theme/theme1.xmlYOoE#F{o'NDuر i-q;N3' G$$DAč*iEP~wq4;{o?g^;N:$BR64Mvsi-@R4Œ mUb V*XX! cyg$w.Q "@oWL8*Bycjđ0蠦r,[LC9VbX*x_yuoBL͐u_. DKfN1엓:+ۥ~`jn[Zp֖zg,tV@bW/Oټl6Ws[R?S֒7 _כ[֪7 _w]ŌShN'^Bxk_[dC]zOլ\K=.:@MgdCf/o\ycB95B24S CEL|gO'sקo>W=n#p̰ZN|ӪV:8z1f؃k;ڇcp7#z8]Y / \{t\}}spķ=ʠoRVL3N(B<|ݥuK>P.EMLhɦM .co;əmr"*0#̡=6Kր0i1;$P0!YݩjbiXJB5IgAФ޲a6{P g֢)҉-Ìq8RmcWyXg/u]6Q_Ê5H Z2PU]Ǽ"GGFbCSOD%,p 6ޚwq̲R_gJS֣9)嗛(:/ak;6j11太x~<:ɮ>O&kNa4dht\?J&l O٠NRpwhpse)tp)af] 27n}mk]\S,+a2g^Az )˙>E G鿰L7)'PK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-!K[Content_Types].xmlPK-!֧6 1_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!\theme/theme/theme1.xmlPK-! ѐ' theme/theme/_rels/themeManager.xml.relsPK] 45<=5NQ2P/67)8w:U=bjoqpx~RTVWXYZ[_vg(lRYbbb#cTckcccc,dJdddddeeffgWhiijakllNmm8nnnn1omoop[ppoqrs+uvxIyz{~gSU\]^`abcdefghijklmnopqrstuwxyz{|}~...... /4/6///////0'0)050I0K0w111b2v2x2z222 5456585L5N5s666:::::::::::::8@0(  B S  ?||ei}).049B"8Bgn5<//`6j6<<]]]]]]7^9^eegghhpixizi~iiiii-j4jjjAkEkJkQkYk`kkk,m1mCmGm[mdmn nn"nnnGoQoppJqRqSqVqbqiqrr*r1rrrjstsssssssutztttbujuv$vnwtwwwwwwwwwvx}xxxxxhymyyzzz-|6|:|?||{ 22KKee&f.f/f7fffffffffffffg'g(g0g[gcgdglggggghh hhIhQhRhZhhhhhkk8q:qs$sYs]sssyy{{|:::::::::::::::::::::::::::::::::[H I' >)UQ+#eO OYZ Z-ZCZYZZZZZZZZZZZZZ[[[#[$[H[M[R[T[U[_[d[i[k[l[z[[[[[[[[[[[[[[[\"\'\,\-\;\@\E\J\K\v\{\}\\\\\\\\\\\\\\\\\\]] ]+]Z]]]]]]]]]]]]]]]]]]]]]^^ ^^^^M^T^[^]^d^k^n^v^}^^^^^^^^^^^^^^_ ____$_'_._5_8_9_S_Z_a_e_l_s_w_________________________``E`L`S`W`^`e`i`q`x`z`{````````````aaaaaaaaaaa=aEaLaPaXa_abaiapatauaaaaaaaaaaaaaabb bbbb%b)b*bSb[bbbebmbtbwbbbbbbbbbbbbbbbbbbbbbcc ccccachcocrcyccccccccccccccccccccddddd d'd*d+dRdYd`dddkdrdud}dddddddddddddddde eeee e'e.e2e3e@eEeFeGeLeMeNeSeTeUeVe`eeefegelemeneseteueveweeeeeeef"f&f/f8f9f}ffffffffffffffffffggg(g1g2gRgXg[gdgmgngggggggggh hhh@hGhIhRh[h\h}hhhhhhhhoipijjjj@kAkkkll+m,mnnnnFoGoppppIqJqrrrrisjsssttutvvvvwwuxvxgyhyyyzzF{G{|||||@P@UnknownGTimes New Roman5Symbol3 Arial7 Verdana7CalibriC̟Lucida Grande"qhGPb2UHq244y= 3QHP $P_2!xxThe New SchoolJoelle Lessler