From Pandemic Disruption to a Lower Plateau
This chapter follows the national kindergarten MMR coverage trajectory from its pre-pandemic high through the 2020–2021 disruption and into the lower level reported afterward. The point is not only to measure how far coverage fell, but to decide what kind of change the record describes: a short-lived interruption, or a persistent loss that later cohorts continue to carry. That distinction helps the reader interpret a national trend without turning its timing into proof of one cause.
The starting point was high, but not permanently guaranteed
Before the pandemic, U.S. kindergarten MMR coverage was reported at approximately 96–97% in the 2018–2019 school year. The study uses 96.0% as a specific reference point for that period. By 2023–2024, the national figure had fallen to 92.7%. The updated weighted estimate for 2024–2025 remained near 92.5%.
The change from roughly 96% to roughly 92.5% is about three and a half percentage points. That is a national average, not a claim that every state lost exactly the same share. It is nevertheless large enough to alter the context in which the number should be read. The earlier level sat above the conventional 95% benchmark often used for measles protection; the recent national estimates sit below it. The benchmark is a practical reference, not a guarantee. But moving from above it to below it means the national margin is thinner, and it makes the distribution underneath the average more consequential.
A rate can decline in two different ways. It can fall briefly and then return to its former level, or it can fall and settle into a new range. The first pattern suggests an interruption whose effects were largely repaired. The second suggests that the system’s response did not close the gap before a later cohort entered the same reporting frame. The study’s central temporal observation is that the national record looks more like the second pattern: coverage declined during the pandemic period and did not fully rebound when in-person schooling resumed.
That conclusion is about the shape of the surveillance series. It does not require us to decide that every child missed the same opportunity, or that every state followed the same path. It says that the national system did not restore its earlier average within the period examined. For interpreting the title’s 92%, the persistence matters as much as the endpoint. A lower figure that is already recovering means something different from a lower figure that has become a plateau.
The disruption was followed by incomplete recovery
The study reports an approximately 1–2 percentage-point decline in MMR coverage, along with declines for other major childhood vaccines, during the 2020–2021 school year. The timing coincided with pandemic-related disruptions to routine care and school-based vaccination enforcement. Families faced altered access to health services, schools changed their operating arrangements, and ordinary administrative reminders and checks were interrupted or weakened. The available evidence supports that sequence of events and the observed decline.
The next transition is the important one. After in-person schooling resumed, coverage did not simply return to its former level. The study describes a lower plateau and suggests that missed vaccinations during the pandemic period were not fully recovered. A plateau does not mean that no child caught up. It means that enough of the gap remained, or new shortfalls appeared, for the national estimate to stay below its earlier position.
This persistence can be understood through the way kindergarten surveillance renews its denominator. Each school year brings a new group of children into the kindergarten population. A missed appointment or a disrupted school-entry check in one period may therefore remain visible later if families do not receive an effective reminder, cannot reach care, or encounter an administrative process that does not close the gap. Reopening a clinic or a school restores the opportunity to vaccinate, but opportunity alone does not guarantee completion. Catch-up requires contact with families, available services, a usable record, and enough follow-through for the child’s status to change before or during school entry.
The same mechanism works in the opposite direction. If a system resumes its old routines without identifying which children were missed, the average can stabilize below its prior level. A lower plateau is not necessarily a single event continuing unchanged; it can be the accumulated result of an earlier interruption plus incomplete repair. That is why a later national percentage should not be interpreted only as a report card on the current year. It may also reflect the system’s ability—or inability—to recover from earlier losses.
A school-year statistic is not a longitudinal diary of the children who appeared in an earlier statistic. Suppose a child was not recorded as up to date during a period when routine care and school verification were disrupted. That child might receive the missing vaccination later, remain delayed at the next school entry, or leave the reporting frame before the status is corrected. Meanwhile, a new kindergarten cohort enters the denominator. The national estimate can therefore stay low even while some earlier children catch up, because recovery and replacement happen at the same time.
This is why “catch-up” has a more demanding meaning than simply reopening services. A functioning recovery process has to find the children who were missed, contact families, make vaccination available, reconcile records, and complete the school-entry verification step. If any link is weak, the system may restore capacity without restoring coverage. The study points to reminder–recall systems and post-pandemic outreach as possible weaknesses, not as measured explanations that account for a known fraction of the gap. The distinction matters: a plausible repair target is not the same as a demonstrated cause.
The shape of the series also depends on the time window chosen. A single annual value can hide a gradual recovery, a new decline, or offsetting changes among jurisdictions. Several years of similar low values provide stronger evidence of persistence than one low observation, but even a multi-year national plateau does not show that the same children or the same places account for it. To separate those possibilities, analysts would need more detailed cohort, jurisdiction, and local records than the national summary alone provides. The supplied study can describe the national trajectory and later jurisdictional relationships; it cannot follow every child through the recovery process.
This limitation does not make the plateau an artifact. It clarifies what the plateau means. The record supports concern that the earlier national level was not restored during the period examined. It does not support a claim that every missed vaccination remained missing, that every service interruption had the same effect, or that a single intervention would automatically return coverage to its former level. A careful interpretation treats the lower plateau as a system-level outcome whose component pathways still need to be separated.
The diagram separates observed transitions from explanatory certainty. The record shows a high baseline, a decline during the pandemic period, and a later lower level. The items in the note are possible contributors discussed by the study, not stages that can be assigned a measured share of responsibility.
Timing identifies candidates, not a verdict
It is tempting to tell a clean causal story: the pandemic disrupted care, so coverage fell; schools reopened, so coverage should have recovered; because it did not, one particular barrier must have taken over. The sources do not support that degree of certainty. They support a more useful distinction between what is observed and what remains to be investigated.
Observed in the study’s description are the timing of the decline, its approximate size, and the persistence of a lower plateau. Also observed is the relationship between the later national level and the conventional benchmark. The study then identifies several plausible contributors: disruption to routine pediatric care, weakened school-based enforcement, incomplete reminder–recall systems, limited post-pandemic outreach, access to health care, and public attitudes toward vaccination. The CDC parent report adds evidence that required-vaccine coverage had decreased since the pandemic and that increased hesitancy was a possible context, but a cross-sectional attitude survey cannot reconstruct the entire national time path.
These candidate explanations are not mutually exclusive. A family may have faced a closed clinic, missed a reminder, and held a concern about vaccination. A school may have reopened while its records or verification capacity remained strained. A jurisdiction may have restored enforcement but still have children delayed from the earlier cohort. Treating the causes as a competition in which only one can be true would impose a simplicity that the evidence does not provide.
There is also a difference between a contributor and a complete mechanism. Saying that access may matter does not tell us how many missed vaccinations were caused by access, which families were affected, or whether the barrier persisted. Saying that attitudes may matter does not show that attitudes changed in the same direction or at the same rate in every state. The study’s own discussion keeps enforcement, delivery systems, attitudes, and other contextual differences in view. That is not evasive wording; it is the appropriate boundary for aggregate observational evidence.
A stronger causal conclusion would require more than the coincidence of a pandemic and a coverage decline. It would need evidence that compares changes in relevant exposures and outcomes over time, accounts for differences among jurisdictions, and distinguishes delayed vaccination from permanent non-vaccination. The supplied study was designed to describe trends and examine a later state-level association, not to assign causal shares to the pandemic, policy, access, or attitudes. That design choice does not erase the trend. It tells us how far the inference can travel.
Why a national recovery story still needs a map
A national time series can answer “when did the average move?” It cannot answer “where did protection weaken most?” The 1–2 percentage-point pandemic-era decline was an aggregate description, and the later 92.7% and 92.5% estimates are also aggregate descriptions. They do not show whether the loss was spread evenly, concentrated in particular jurisdictions, or offset by high coverage in some places and deeper declines in others.
This matters because a national plateau can contain several different local stories. Some jurisdictions may have recovered close to their earlier level while others remained far below it. Some may have strong catch-up programs but rising exemptions. Others may report high coverage overall while having a school or district where susceptible children are concentrated. The national line cannot choose among these possibilities. It is a signal that tells us the next layer of analysis is needed.
The distinction also changes how “recovery” should be defined. A return to the old national average would be encouraging, but it would not prove that every community had regained the same protection. Conversely, a national figure that remains near 92.5% does not prove that every locality is equally vulnerable. Recovery has at least two dimensions: the level of the combined estimate and the distribution of that level across jurisdictions and local settings. The first is visible in the time series; the second requires disaggregation.
That is why the post-pandemic plateau connects directly to the exemption and policy question. A persistent national shortfall could reflect several interacting weaknesses, including incomplete catch-up and changing exemption patterns. The study reports that exemption prevalence increased over the same broad period, but temporal alignment alone does not establish how much each factor contributed. To understand the relationship, we need to examine how exemptions are defined, how state rules differ, and how administrative enforcement turns policy into a recorded coverage pattern.
The conclusion the trend can support
The national evidence supports three disciplined statements. First, kindergarten MMR coverage fell from a high pre-pandemic level to the low-92% range. Second, the decline was not described as a fully temporary dip; coverage remained near that lower level in the updated estimate. Third, the persistence makes recovery systems, enforcement, access, outreach, and attitudes relevant questions, but it does not identify one proven cause.
The trend therefore changes the meaning of the national percentage without completing the explanation. It tells us that the country is no longer operating with its earlier reported margin, and that a later year’s figure may carry forward unresolved gaps from earlier cohorts. It does not tell us whether a particular school is protected, whether a family’s decision was primarily logistical or ideological, or whether tightening one policy would restore coverage by a known amount.
A responsible reader should hold two ideas at once: the plateau is real enough to demand attention, and its causes are open enough to require investigation. The next layer is not a search for a villain hidden inside the time series. It is an examination of the exemption pathway and state school-entry rules that vary across jurisdictions. Those rules help explain why a common national trend can produce very different coverage levels in different places.