Is Homeschooling Actually Growing? An Original Look at NCES’s Own Data

Direct answer: Not in a straight line. NCES’s own data shows the homeschooling rate rose from 1.7% in 1999 to a peak of 3.4% in 2012, then dipped to 2.8% by 2019, before pandemic-era numbers, measured a different way, temporarily spiked far higher. The sharper story in the same dataset: rural households homeschool at nearly double the rate of families in cities.

Most “homeschooling is growing” claims online cite a single before-and-after figure. We went to the actual NCES table behind those figures and pulled every available year, plus the breakdowns most summaries leave out.

The 20-year trend, not the simplified version usually cited

Year Homeschooling rate
1999 1.7%
2003 2.2%
2007 3.0%
2012 3.4%
2016 3.3%
2019 2.8%
U.S. homeschooling rate, 1999 to 2019 Line chart showing the percentage of 5 to 17 year olds homeschooled: 1.7% in 1999, 2.2% in 2003, 3.0% in 2007, 3.4% in 2012, 3.3% in 2016, and 2.8% in 2019. 0% 1% 2% 3% 4% 1999 2003 2007 2012 2016 2019
Chart: Gurukol original analysis of NCES Table 206.10, "Number and percentage of homeschooled students ages 5 through 17...: Selected years, 1999 through 2019."

The rate didn’t climb steadily and it hasn’t kept climbing. It peaked in 2012, actually fell back by 2016, and was lower still in 2019, pre-dating the pandemic entirely. One real caveat on comparing across the full 20 years: NCES itself notes that the 2019 figure uses a stricter enrollment-hours cutoff than earlier years (excluding students enrolled more than 24 hours a week, versus 25 hours in prior years), which makes the 2016-to-2019 dip partly a definitional shift, not purely a behavioral one, a distinction most summaries of this same table skip.

The gap almost no one highlights: rural vs. urban

Homeschooling rate by community type, 2019 Bar chart: rural households homeschool at 4.7%, city households at 2.5%, suburban households at 2.4%, and town households at 2.2%. Rural City Suburban Town 4.7% 2.5% 2.4% 2.2%
Chart: Gurukol original analysis of NCES Table 206.10's 2019 geographic breakdown.

This is the finding most “homeschooling is booming” content leaves out entirely: in the same 2019 NCES data, rural households homeschooled at 4.7%, compared with 2.5% in cities, 2.4% in suburbs, and 2.2% in towns. Rural families homeschool at nearly double the rate of every other community type combined. NCES’s table doesn’t state a cause, and neither will we. But the gap itself is real, sourced directly from the government’s own breakdown, and rarely shown alongside the more commonly cited national average.

Why the pandemic numbers don’t belong on this chart

You may have also seen a much bigger number: the same Census survey found 5.4% of households reporting homeschooling in spring 2020, jumping to roughly 11% by fall 2020. Both figures are real, but they come from a different U.S. Census Bureau survey (the Household Pulse Survey), asked a different way, over a much shorter window, during a period when many families were counting emergency remote learning as “homeschooling.” Splicing those numbers onto the NCES trend line above would visually suggest a single continuous climb that the actual data doesn’t support: two different surveys, measuring two different things, are not the same time series. We’re keeping them separate here for that reason, not omitting the pandemic spike by accident.

What this means if you’re weighing homeschooling

The headline “homeschooling is exploding” framing oversimplifies a rate that has actually leveled off and even declined slightly since its 2012 peak, by the government’s own most consistent long-run measure. If you’re comparing your own decision against a national trend, the more useful comparison is the community-type gap above: homeschooling is a genuinely mainstream, common choice in rural areas specifically, not just a rising national percentage.

Sources: this piece is Gurukol’s own original analysis and visualization, built directly from NCES Table 206.10, “Number and percentage of homeschooled students ages 5 through 17…: Selected years, 1999 through 2019” (U.S. Department of Education, National Center for Education Statistics), not a citation of someone else’s summary of that table. The 2019 methodology note (24-hour vs. 25-hour weekly enrollment cutoff) is disclosed directly from NCES’s own table footnotes. Pandemic-era figures (5.4% spring 2020, 11.1% fall 2020) are drawn from the U.S. Census Bureau’s Household Pulse Survey, a separate survey with a different methodology and timeframe, deliberately not merged into the chart above: see “Why the pandemic numbers don’t belong on this chart.” Published 2026-08-06 as the site’s first Tier 1 original-data piece under its authoritative-sourcing standard.