Hispanic-Majority Districts: Boon or Burden for Democrats?
By NATE SILVERThe intuitive argument is that Congressional redistricting will benefit Republicans, since it will shift the allocation of representatives into G.O.P.-friendly states.
The counter-intuitive argument, which some very smart people are making, is that this isn’t necessarily the case. Instead, the argument goes, redistricting will help the Democrats in some cases because of the creation of new majority-minority districts, which will be required in some states according to most interpretations of the Voting Rights Act. In Texas, for instance, which is subject to the Voting Rights Act’s preclearance process, as many as three of the state’s four new districts could be Hispanic-majority ones, according to former U.S. Representative Martin Frost.
Time for the counter-counter-intuitive argument.
Yes, if three of the four new districts in Texas are majority-Hispanic, those particular districts are likely (although not by any means certain) to vote Democratic. This does not necessarily imply, however, that Democrats will benefit from their creation overall. In fact, there are many circumstances in which the majority-minority requirement can be harmful to Democrats. The key is in thinking about not just what happens with the four new districts, but also how this impacts Texas’ 32 existing ones.
Imagine that you have a state with 100 voters — 50 Democrats and 50 Republicans. This state originally has 4 Congressional districts, so there are 25 voters in each one. In two of these, District A and District C, Democrats have a 14-11 majority. In the other two, District B and District D, Republicans have a 14-11 majority.
But then a new district, District E, is created, and 20 voters must be allocated to it (with the addition of the fifth district, each district will now have 20 voters rather than 25). This new district, for whatever reason — perhaps it is required to be Hispanic-majority — is quite strongly Democratic. In fact, 14 of its 20 voters are Democrats, 5 of whom are taken from each of Districts A and C, and 2 of whom are taken from each of Districts B and D. The other 6 voters are Republicans, 3 of whom are recruited from each of Districts B and D.
So what happens now? Well, the Democrats are indeed huge favorites in District E. But they are now outnumbered, 11 to 9, in the four existing ones. Instead of there being 2 Republican-leaning districts and 2 Democratic-leaning districts, there is now just one district that favors the Democrats and 4 where the Republicans have the advantage.
This stuff is pretty basic, really, but I’m surprised how often analyses of redistricting suffer from this logical flaw. The number of Democrats and Republicans in any given state at any given time is, in essence, fixed. If a new Democratic district is created, those Democrats must be taken from somewhere else. It is quite possible that in the process of creating one new Democratic district, two or more districts will be tipped toward Republicans.
This won’t always be the case, mind you. The chart below, for example, presents an example of another state in which there are 60 Republicans and 40 Democrats, and the creation of a new district helps the Democrats to go from having one member of Congress to two.
Total 46 54 43 57
The key is how efficiently each party’s voters are allocated. What a party would prefer is that, in the districts where it has a majority, that majority is as small as possible, so as not to waste any of its voters. (Although in the real world, it would probably want to build in some tolerance for error since the preferences of the electorate can change slightly from year to year: perhaps it would want a district where it might get 55 or 56 percent of the vote in an average year so that it could still get 50-percent plus 1 vote even in a poor year.) Conversely, in those districts where it didn’t have the majority, it would prefer to lose by as many votes as possible — in fact, it would prefer to have none of its voters there at all.
What a party wants to avoid, meanwhile, is districts where it has say 45 percent of the vote: it’s using up a fair number of its voters, but not enough to give it a majority. It would also like to avoid districts where it has close to 100 percent of the vote, since so many of those votes will be superfluous.
The question, then, is just which kind of districts these new Hispanic-majority seats in Texas are liable to be. If they’re the sorts of districts where the Democrat will win with about 55 percent of the vote in an average year, then they are indeed likely to be good news for Democrats. If, on the other hand, they’re taking 70 or 80 or 90 percent of the vote there, Democrats are probably being harmed on balance statewide.
Districts with African-American majorities are in fact sometimes this lopsided. There were 43 districts in which Barack Obama won at least 75 percent of the vote (conversely, John McCain had just 5 such districts), a little less than half of which are majority-black. Not all majority-black districts are gerrymandered (on the South Side of Chicago, for instance, districts with enormous black populations are an inevitable consequence of the geography), and there may be a couple of case in the rural South (but not more than a couple) where gerrymandering to create majority-black districts works to the Democrats’ benefit since otherwise it would be hard to round up a sufficient number of Democrats at all. In general, however, majority-black districts create an inefficient distribution of Democratic votes and cost them seats in Congress.
Majority-Hispanic districts may be another story, however. (Yes, herein begins the counter-counter-counter-intuitive part of the article.) Some, like New York’s 16th Congressional District in the Bronx, where Puerto Rican and Dominican voters help propel Democrats to 90-point victories every year, are just as lopsided as the majority-black ones. But others are more competitive, and some — like those in Florida where many of the Hispanics are Cuban-Americans — actually tilt somewhat toward Republicans. Among the 28 majority-Hispanic districts that exist now, Barack Obama lost 3 in 2008, and Democrats lost 6 in last month’s elections.
Below are the figures for the seven current Hispanic-majority districts in Texas. Barack Obama won an average of 59 percent of the vote in these districts in 2008, and Democrats won an average of 56 percent of the vote for Congress in November, seeing their incumbents defeated in two of them (the 23rd and 27th districts).
So there aren’t a huge excess of Democrats in these districts. In fact, they’re probably in the range that tends to be helpful to Democrats on balance.
Ironically, Hispanic-majority districts are more helpful to Democrats than black-majority ones precisely because Hispanics are less reliable Democratic voters than African-Americans are, giving perhaps 60 or 65 percent of their votes to Democrats rather than 90 or 95 percent (although there are significant differences in the Hispanic vote based on regional, generational and ancestral patterns). Also, Hispanics vote at much lower rates than either African-Americans or whites (to some extent, because a fair number are not yet citizens): although they make up about 16 percent of the U.S. population, they are closer to 10 percent of the electorate. They will usually suffice to give Democrats a majority, but unless they are complemented by other groups like African-Americans or urban whites, it will not necessarily be a huge one — this is desirous to Democrats from the standpoint of using their voters efficiently.
The counter-intuitive argument, which some very smart people are making, is that this isn’t necessarily the case. Instead, the argument goes, redistricting will help the Democrats in some cases because of the creation of new majority-minority districts, which will be required in some states according to most interpretations of the Voting Rights Act. In Texas, for instance, which is subject to the Voting Rights Act’s preclearance process, as many as three of the state’s four new districts could be Hispanic-majority ones, according to former U.S. Representative Martin Frost.
Time for the counter-counter-intuitive argument.
Yes, if three of the four new districts in Texas are majority-Hispanic, those particular districts are likely (although not by any means certain) to vote Democratic. This does not necessarily imply, however, that Democrats will benefit from their creation overall. In fact, there are many circumstances in which the majority-minority requirement can be harmful to Democrats. The key is in thinking about not just what happens with the four new districts, but also how this impacts Texas’ 32 existing ones.
Imagine that you have a state with 100 voters — 50 Democrats and 50 Republicans. This state originally has 4 Congressional districts, so there are 25 voters in each one. In two of these, District A and District C, Democrats have a 14-11 majority. In the other two, District B and District D, Republicans have a 14-11 majority.
But then a new district, District E, is created, and 20 voters must be allocated to it (with the addition of the fifth district, each district will now have 20 voters rather than 25). This new district, for whatever reason — perhaps it is required to be Hispanic-majority — is quite strongly Democratic. In fact, 14 of its 20 voters are Democrats, 5 of whom are taken from each of Districts A and C, and 2 of whom are taken from each of Districts B and D. The other 6 voters are Republicans, 3 of whom are recruited from each of Districts B and D.
So what happens now? Well, the Democrats are indeed huge favorites in District E. But they are now outnumbered, 11 to 9, in the four existing ones. Instead of there being 2 Republican-leaning districts and 2 Democratic-leaning districts, there is now just one district that favors the Democrats and 4 where the Republicans have the advantage.
This stuff is pretty basic, really, but I’m surprised how often analyses of redistricting suffer from this logical flaw. The number of Democrats and Republicans in any given state at any given time is, in essence, fixed. If a new Democratic district is created, those Democrats must be taken from somewhere else. It is quite possible that in the process of creating one new Democratic district, two or more districts will be tipped toward Republicans.
This won’t always be the case, mind you. The chart below, for example, presents an example of another state in which there are 60 Republicans and 40 Democrats, and the creation of a new district helps the Democrats to go from having one member of Congress to two.
Total 46 54 43 57
The key is how efficiently each party’s voters are allocated. What a party would prefer is that, in the districts where it has a majority, that majority is as small as possible, so as not to waste any of its voters. (Although in the real world, it would probably want to build in some tolerance for error since the preferences of the electorate can change slightly from year to year: perhaps it would want a district where it might get 55 or 56 percent of the vote in an average year so that it could still get 50-percent plus 1 vote even in a poor year.) Conversely, in those districts where it didn’t have the majority, it would prefer to lose by as many votes as possible — in fact, it would prefer to have none of its voters there at all.
What a party wants to avoid, meanwhile, is districts where it has say 45 percent of the vote: it’s using up a fair number of its voters, but not enough to give it a majority. It would also like to avoid districts where it has close to 100 percent of the vote, since so many of those votes will be superfluous.
The question, then, is just which kind of districts these new Hispanic-majority seats in Texas are liable to be. If they’re the sorts of districts where the Democrat will win with about 55 percent of the vote in an average year, then they are indeed likely to be good news for Democrats. If, on the other hand, they’re taking 70 or 80 or 90 percent of the vote there, Democrats are probably being harmed on balance statewide.
Districts with African-American majorities are in fact sometimes this lopsided. There were 43 districts in which Barack Obama won at least 75 percent of the vote (conversely, John McCain had just 5 such districts), a little less than half of which are majority-black. Not all majority-black districts are gerrymandered (on the South Side of Chicago, for instance, districts with enormous black populations are an inevitable consequence of the geography), and there may be a couple of case in the rural South (but not more than a couple) where gerrymandering to create majority-black districts works to the Democrats’ benefit since otherwise it would be hard to round up a sufficient number of Democrats at all. In general, however, majority-black districts create an inefficient distribution of Democratic votes and cost them seats in Congress.
Majority-Hispanic districts may be another story, however. (Yes, herein begins the counter-counter-counter-intuitive part of the article.) Some, like New York’s 16th Congressional District in the Bronx, where Puerto Rican and Dominican voters help propel Democrats to 90-point victories every year, are just as lopsided as the majority-black ones. But others are more competitive, and some — like those in Florida where many of the Hispanics are Cuban-Americans — actually tilt somewhat toward Republicans. Among the 28 majority-Hispanic districts that exist now, Barack Obama lost 3 in 2008, and Democrats lost 6 in last month’s elections.
Below are the figures for the seven current Hispanic-majority districts in Texas. Barack Obama won an average of 59 percent of the vote in these districts in 2008, and Democrats won an average of 56 percent of the vote for Congress in November, seeing their incumbents defeated in two of them (the 23rd and 27th districts).
So there aren’t a huge excess of Democrats in these districts. In fact, they’re probably in the range that tends to be helpful to Democrats on balance.
Ironically, Hispanic-majority districts are more helpful to Democrats than black-majority ones precisely because Hispanics are less reliable Democratic voters than African-Americans are, giving perhaps 60 or 65 percent of their votes to Democrats rather than 90 or 95 percent (although there are significant differences in the Hispanic vote based on regional, generational and ancestral patterns). Also, Hispanics vote at much lower rates than either African-Americans or whites (to some extent, because a fair number are not yet citizens): although they make up about 16 percent of the U.S. population, they are closer to 10 percent of the electorate. They will usually suffice to give Democrats a majority, but unless they are complemented by other groups like African-Americans or urban whites, it will not necessarily be a huge one — this is desirous to Democrats from the standpoint of using their voters efficiently.
No comments:
Post a Comment