The reapportionment numbers released this week seemed to confirm this fear. To be sure, nationally, census counts came in remarkably close to pre-census predictions. (Those predictions were made using birth and death records, data on international migration and Medicare records.) The total count, nearly 332 million, exceeded pre-census estimates by a mere 0.7 percent. That suggests a high degree of accuracy, especially considering the difficulties of counting under pandemic conditions. But the degree of accuracy was not the same everywhere. The greatest underperformance, relative to projections, occurred in three Sun Belt states — Texas, Florida and Arizona — where head counts fell short of projections by 177,000, 163,000 and 262,000, respectively. These three states also have some of the highest proportions of Hispanics in the country: 40 percent in Texas, 26 percent in Florida and 32 percent in Arizona.
Given that two of those states, Texas and Florida, are red and Arizona is a closely divided purple state, Republicans did not do themselves any favors: Their actions may have suppressed census responses in places where they would benefit from more seats in Congress.
That’s because the consequences for congressional representation largely mirror the regional differences in the census results. Each of those three states received one fewer seat than had been expected from pre-census estimates: Texas gained only two seats, Florida gained one seat, and Arizona gained none. As a result, these three states will send fewer members than expected to the House — and have fewer electoral votes for the presidency.
Other evidence suggests the Trump administration efforts had a suppressing effect: Two other states that are more than one-fifth Hispanic, Colorado and Nevada, were also among the worst seven states when it comes to number of people missed by the census.
There were more factors at play in the undercounts than Trump administration hostility, however. Arizona was undercounted by over 262,000 people, or 3.5 percent. Meanwhile, neighboring New Mexico, which is 49 percent Hispanic, reported nearly 14,000 more people than pre-census estimates, a slight overcount of 0.7 percent, a figure in line with the national overcount average. And California, which is 39 percent Hispanic, likewise reported an 0.5 percent overcount. What accounts for the different outcomes? Those states invested in the census as a matter of self-interest. New Mexico spent nearly $12 million on outreach efforts, and California invested a whopping $187 million. The investment will pay off, since many federal benefits are based on census population. And although California lost a House seat for the first time ever, its high-quality count forestalled a potential loss of an additional congressional seat.
Meanwhile, Arizona and Florida spent no money at all on outreach, and Texas’s efforts were limited to a last-minute ad campaign. Costly outreach is anathema to the small-government attitude of Republicans, an especially unfortunate stance when so many Republican-leaning states have low population density and may be harder to count.
But reapportioned seats are not the end of the story. All states now face the ultimate translation of population growth into power — namely, redistricting, which will start in the fall. Redistricting has even higher stakes than reapportionment, because a gerrymander can assign power in a lopsided manner no matter what the political leanings of the citizens. For example, in the 2010 Census cycle, Texas gained four congressional seats, driven in large part by Hispanic growth, yet no Hispanic members were added to the congressional delegation, because districts were redrawn to diminish their power.
Both Democrats and Republicans have it in their reach to prevent a recurrence of this harm. The Democratic incentive is obvious: Latinos lean toward that party. Democrats in Congress can reintroduce H.R. 4, the John Lewis Voting Rights Act, which passed the last session of the House before dying in the Senate. H.R. 4 would restore the preclearance provisions of the Voting Rights Act — struck down by the Supreme Court in 2013 — that mandate the review of new election laws in states that have a poor history of voting rights. That includes redistricting, and the list of scrutinized states formerly included Texas and Florida. Passage of H.R. 4 would potentially hold back a racial gerrymander that once again disenfranchises Hispanics. The last time the Voting Rights Act was renewed, in 2006, it passed by overwhelming bipartisan majorities; granted, the odds are much stiffer now, but principle demands that it be given a hearing.
It’s also not out of the question that Republicans could revise their thinking about Hispanic voters given that in several parts of the country — including southwestern Texas and southern Florida — substantial numbers of them shifted to President Donald Trump last November. If Republicans want to demonstrate to Hispanic voters that they aren’t interested in disempowering them — and in fact plan to compete for their votes — they could take steps that compensate for the census undercount. State-districting rules allow legislators to draw slightly underpopulated legislative districts, a move that, of course, enhances the power of each voter in such a district. By doing this in districts rich in Hispanics, Republicans would signal that they, too, want to be a multiracial party. Doing so could start to rewrite today’s national narrative linking race and party.
Hispanics appear to have been undercounted in several key states, but the battle isn’t over yet. The redistricting fights ahead may yet give them seats at the congressional table.