Stay Connected

The State of the Children’s Savings Field: A Look Behind the Numbers

CFED recently released A Growing Movement: The State of the Children’s Savings Field 2016. Based on a survey of CSA programs, the document highlights trends in the field. For most people, that document covers all that they want to know about the state of the CSA field. However, for my fellow data wonks and CSA aficionados, this blog offers a deeper dive into the survey results.

One of the more revealing aspects of the survey is the variations in CSA program models. To dig into these variations more, we analyzed key program features by program size. For the purpose of this analysis, we categorized programs enrolling 2,000 or more children annually as “large” and programs enrolling less than 2,000 children as “small.” (See more on annual enrollment across programs in the State of the Field document.) Here is what the analysis shows about differences and similarities in key features between these programs:


  • Large programs are more likely to use automatic enrollment — 57% of larger programs are opt-out as compared to 26% of smaller programs.
  • Large programs usually have narrower target populations — all but one of the large programs enroll children at just one point in time (birth or kindergarten). Just over half of small programs have only one enrollment point, while the other 49% allow kids to enroll at multiple points (e.g., at any grade in elementary school).


  • Large programs are more likely to offer initial deposits — 100% of large programs provide an initial deposit compared to 57% of small programs. However, the median initial deposit amount is the same for both ($50).
  • Large and small programs offer savings matches and benchmark incentives at virtually the same rates.

Account Type

  • Large programs are more likely to use 529s as their account platform — 71% of large programs use 529s compared with 43% of small programs.


  • Large programs are more likely to be at least partially government funded — 57% of large programs receive some type of government funding (local, state or federal) compared with 34% of small programs.

The analysis reveals that large and small programs tend to differ in some significant ways, particularly around enrollment. This is not very surprising given the need to streamline operations when a program enrolls thousands of children per year—even as many as 35,000. Larger programs cannot expend as many resources per child as smaller programs if they hope to keep administrative costs manageable. So, for example, automatically enrolling children based on data collected through birth or school records is more efficient for a large program than reaching out to and following up with more than 2,000 families about enrolling their children in the program.

Higher usage of 529 accounts for large programs is also not surprising, since four out of the seven large CSA programs are statewide programs. It makes sense that these programs would use the state-supported 529 college savings plan, especially given the difficulty of finding one bank or credit union that is accessible in all regions of the state. The omnibus 529 structure also supports automatic enrollment, since the program can invest an initial deposit on behalf of each child into the pooled 529 account by just using information obtained from state birth certificate records.

Overall, the structure of larger programs matches CFED’s guidance on designing CSA programs for scalability. (See a discussion of this at the beginning of Section A in Investing in Dreams.) Features such as automatic enrollment, working within the existing structure of a 529 account and receiving public funding and support, may make it easier to build and sustain large-scale programs, especially statewide programs. Even for programs that are starting with a smaller rollout, if the ultimate goal is to serve thousands of children per year, it is important to build in the features that will make the program manageable and successful at a larger scale.