Performance Trends in State-Facilitated Retirement Savings Programs
1 Introduction
A review of the data over a 3-year period of program performance — December 2019 to December 2022 — for OregonSaves, Illinois Secure Choice, and CalSavers, highlighting several positive trends.
2 Total Assets
2.1 Raw Data at a Glance
2.2 Data Cleaning and Preprocessing
In the data preprocessing and cleaning steps, several important procedures were carried out to ensure the quality and completeness of the dataset. These steps are detailed below:
1. Removing Duplicate Rows: Two duplicate rows in the dataset were identified and eliminated using the unique function in R. This process helps to ensure that each data point in the dataset is unique, preventing any duplication that might lead to inaccurate analysis or skewed results.
2. Handling Missing Values in “CalSavers”: In the “CalSavers” column, four rows contained missing values. This occurred because CalSavers did not initiate monthly reporting until June 30, 2020, and instead had a previous practice of reporting on a quarterly basis. To resolve this issue and align the data with the reporting frequency, we applied the Last Observation Carried Forward (LOCF) imputation method. The LOCF technique involves replacing missing values with the most recent observed value in the time series. This method was chosen to ensure that the imputed values accurately reflect the temporal context of the data. By utilizing the most recent available data point, we maintain the continuity of the information, addressing the missing values without disrupting the dataset’s chronological integrity. This approach enhances the dataset’s completeness and allows for more meaningful analysis and interpretation.
3. Filling in Missing “Total” Values: The “Total” column had four missing values as well because Calsavers was missing four values. To calculate the Total value for all the three states, the sum of the “OregonSaves,” “IL Secure Choice,” and “CalSavers” columns for their respective rows was computed using the rowSums function. This step allowed us to ensure that the “Total” column is complete and accurately reflects the total assets under management for each corresponding row.
After completing the data cleaning and preprocessing of the Total Assets dataset, we have generated the following dataset, which will be used for analyses.
2.3 Data Insights and Visual Interpretations
Total assets under management grew by over 1000%, from $54.8 million to $640.2 million.
CalSavers, the state with the largest number of eligible workers, grew faster than other state programs over this period, expanding total assets by more than 26,000% from $1.4 million to $373.0 million. It is important to note that the Californian program was as still in its infancy in December 2019.
Illinois Secure Choice expanded total assets by roughly 690% between December 2019 and December 2022, from $12.4 million to $98.5 million.
OregonSaves grew total assets by over 310% over this same time period, from $41.0 million to $168.7 million.
California ended with highest amount of asset under control, $373.0 million, versus $98.5 and $168.7 million for Illinois and Oregon respectively.
The asset accumulation for CalSavers demonstrates an exponential growth curve, underlined by a robust positive trend.
On a month-on-month basis, CalSavers averaged a growth rate of 18.3%. When annualized, this rate further magnifies, reflecting an average yearly growth of 219.6%.
Analyzing monthly growth percentages, December 2020 emerged as the pinnacle for CalSavers, where it recorded a monthly growth rate of 42.7%.
Important Note: A crucial factor to remember while interpreting these numbers is the preprocessing strategy employed on the dataset. The Last Observation Carried Forward (LOCF) method was utilized to handle NA values. This preprocessing, between January 2020 to June 2020, inadvertently inflated the growth percentages, making them appear higher. Thus, while some months in this range showed growth rates as high as 108% and 76%, these figures are not representative of the actual growth and are outcomes of the preprocessing strategy. The real, unskewed data was available post-June 2020, reaffirming the aforementioned peak growth rate in December 2020.
2.3.1 Dynamic Distribution of Retirement Assets Over Time
Initial figures from December 2019 depict CalSavers lagging behind its counterparts - Illinois Secure Choice and OregonSaves, in terms of total assets.
However, this narrative shifted dramatically as CalSavers displayed an aggressive growth pattern:
By March 2021, CalSavers outpaced Illinois in asset accumulation.
A few months later, by October 2021, it surpassed OregonSaves.
Post these milestones, CalSavers not only retained its lead but managed to substantially widen the asset gap with its counterparts.
2.3.2 Proportional Shifts in Retirement Program Contributions
The bar chart below displays the percentage distribution of total assets among three retirement programs over four consecutive years (2019-2022).
CalSavers: Starting with only 2.6% of the total assets in December 2019, there is a dramatic increase in its share, reaching 58.3% by December 2022. This significant growth indicates a rapid expansion of the CalSavers program.
Oregon Saves: This program had a substantial portion of the assets at 74.8% in December 2019, but its share has decreased over the years, ending at 26.4% in December 2022.
IL Secure Choice: The share of total assets for IL Secure Choice has seen some fluctuations, starting at 22.6% in 2019, increasing to 29.3% in 2020, then dipping to 20.8% in 2021, and further decreasing to 15.4% by December 2022.
Overall, the chart indicates a shifting dynamic where CalSavers has gained a significant market share between 2019 and 2022, while Oregon Saves, although starting strong, has seen a relative decrease in its share of total assets. IL Secure Choice has experienced ups and downs but remained relatively stable compared to the other two programs.
2.3.3 Trend and Seasonality in Program Asset Values
Below is a time series decomposition of asset values for different programs over the three year period. Each program exhibits a robust upward trend, signifying appreciable asset value appreciation over the observed period. Additionally, a pronounced seasonal pattern emerges, remarkably consistent across all programs, suggesting systematic fluctuations tied to specific times of the year.
The seasonal peaks are most prominent in May and November, where asset growth accelerates. Conversely, June and December present as the troughs in the seasonal cycle, where asset growth dampens, marking the lowest expansion points in value. These periodic declines could be influenced by factors such as fiscal year-ends, tax implications, or scheduled withdrawals that typically depress asset growth during these months.
While the overarching trend for asset values is positive, understanding the cyclical patterns is crucial for accurately forecasting and strategizing for future growth within these programs. It is important to note that while the seasonality effects are pronounced, the overall trend remains strongly positive, suggesting that any seasonal dips are temporary and overshadowed by the general growth trajectory of the asset values.
