Plan Sponsor, Plan Sponsor, Make Me a Match

The preceding paper in our Benefit Connection series explored how different defined benefit (DB) designs can impact the target date glide path in an accompanying defined contribution (DC) plan. We showed how cost‑of‑living adjustments, lump‑sum availability, and benefit accrual patterns all can materially impact glide path design in various ways.¹ 

While DC plans also may include a wide array of design features—such as auto‑enrollment, auto-escalation, and the specific investment lineup offered, just to name a few—the employer match is the feature we are most frequently asked about in relation to glide path appropriateness. 

Employer contributions also can take many forms, but the advantage of a match formula is that it tends to promote better participant savings behavior and thus may result in higher account balances over time. Matching also is the most common contribution feature for the plans on T. Rowe Price’s recordkeeping platform. In this paper, we review how three different employer match formulas influenced the optimal glide path design in our modeling analysis.²  

Plan design assessment

To assess how much glide paths should change based on various match formulas, we modeled four hypothetical plan designs, including the baseline safe harbor plan that we have used so far in the Benefit Connection series.

1. Baseline: The employer matches 100% of the first three percentage points of employee salary deferrals and 50% of the next two percentage points, for a maximum employer contribution of 4% of salary.
2. Lean: The employer matches 50% of the first four percentage points in salary deferrals, for a maximum employer contribution of 2% of salary.
3. Average: The employer matches 50% of the first six percentage points in salary deferrals, for a maximum employer contribution of 3% of salary.
4. Rich: The employer matches 100% of the first six percentage points in salary deferrals, for a maximum employer contribution of 6% of salary.

The lean, average, and rich designs collectively composed nearly half of the match formulas offered by the plans on our recordkeeping platform, with the average formula being the most common. We believe these designs represent a reasonable range in employer generosity—while acknowledging that there are match formulas that are leaner or richer than the ones modeled in our analysis. 

Figure 1 shows the hypothetical glide paths that our model identified as optimal for the four designs. Equity allocations in the lean and average glide paths were slightly above the baseline design. However, the maximum difference at any age was only 2.5 percentage points or less (Figure 2). The rich glide path featured slightly lower equity levels at age 40 and above, reflecting less need for participants in those plans to assume market risk to meet their retirement income targets, given their higher overall savings. 

Overall, the maximum equity allocation difference (between the rich and the lean glide paths) was less than five percentage points at all ages. These relatively minor differences may be surprising, given that the match formula typically has the most impact on the cost of a plan. However, as we shall see, it does not necessarily have a major influence on glide path appropriateness when the match is altered in isolation.³  

In our models, the impact of different match formulas on the optimal glide path was even further muted when the DC plan was paired with an accompanying final average pay DB plan (Figures 3 and 4), because the participant income provided by the DB plan made the employer’s DC plan contributions an even smaller share of total retirement wealth. 

This result only would have been reinforced if our model had reflected the reality that plan sponsors seeking to preserve liquidity have greater discretion to suspend their DC match formulas than to avoid the regulatory required minimum contributions for corporate DB plans or legislated funding requirements for some public DB plans. Throughout the 2008–2009 global financial crisis and more recently during the coronavirus pandemic, we saw several companies take advantage of this flexibility. 

Even when we applied the same DC match formula across the entire length of an employee’s career, it still did not materially affect the optimal glide path in our model. A temporary suspension in the employer match would have reduced this impact even further. 

Why the surprising results? 

T. Rowe Price recordkeeping data indicate that most plan participants respond rationally to incentives over their careers, meaning they start by saving at least as much as is required to maximize the employer match and then increase their salary deferrals as they age. Increased plan use of auto‑enrollment and auto‑escalation features probably has helped reinforce this behavior, which also is incorporated in our analysis (Figure 5). 

Recently, we’ve seen renewed sponsor interest in further encouraging participant savings by stretching a lower match percentage to cover a higher percentage of salary deferral—for example, by offering to match 50% of the first 8% of salary rather than 100% of the first 4%. 

Match formula design changes, in isolation, seemed to have minimal impact on the optimal glide path equity allocations indicated in our model. While we did not explicitly model the impact of different match designs on employee savings behavior, this is an area of research that we think warrants further exploration—particularly the impact of match formulas that require employees to defer a higher salary percentage to receive the full match. 

In the population set we assumed for this paper—based on data from T. Rowe Price’s recordkeeping platform as of 2020—70% of participants started out saving at least 4% of salary and 50% started out saving at least 5%. By age 30, more than 50% of participants were deferring at least 6% of salary, a level that maximized the employer contributions in all the plans we modeled. 

In the designs we modeled, the difference in the maximum employer contribution between the rich and lean plans was only 4% of salary (ranging from 2% to 6%), whereas the average participant deferral varied between 5.6% and 8.1% of salary, depending on age.⁴ Viewed through this lens, it’s not surprising that match generosity had a relatively small impact on optimal equity levels in the hypothetical plans we modeled. Most participants were deferring more than the employer match over most of their careers. This suggests that even when a plan sponsor cannot afford a generous match, there still may be opportunities to improve retirement outcomes by implementing auto features and incentivizing participant savings behavior. ⁵

Conclusions

While employers can prioritize matching contributions over other forms of compensation, our analysis indicates that participant behavior ultimately has the greatest impact on glide path equity allocations and on retirement readiness. Our recordkeeping data indicate that DC plan sponsors are having some success convincing participants of the importance of increasing savings, which ultimately reduces the standalone impact of the employer contribution on glide path design.