Overview

The 2020 presidential election upended many previously-held political and electoral assumptions. Longtime Republican bastions of suburban strength gave way to Democratic ascendance. On the flipside, Donald Trump’s performance in many working-class minority communities was the best Republican performance in over a century. The political shifts seen between the 2012 and 2016 presidential elections were largely baked in - i.e., 2020 was not a fundamentally realigning election in most areas. That said, the large pro-Republican shift in minority communities was unforeseen. This article does not seek to explain or diagnose the specific causes of the great minority shift - the developments that led to those Republican swings are far too complicated to explain in an article of this scope.

This article centers on New Mexico, a state situated at the crossroads of multiple political trends, with a fascinating plethora of racial, linguistic, and socioeconomic diversity. Particularly as eyes look to the 2022 election season, New Mexico’s status as a competitive state is worth investigating. The state’s partisanship favors Democrats by a spread of roughly 10 percentage points. Democrats’ coalition is most heavily dependent upon a combination of Latino voters, Native American voters of various nations, pueblos, and tribes, as well as liberal white voters based out of the major cities of Albuquerque and Santa Fe. The Republican coalition in the state is dependent upon conservative white voters as well as a minority of Latino voters, particularly concentrated in the southeastern portion of the state, nicknamed “Little Texas,” for its political, geographical, and cultural similarity to West Texas. This article will examine next year’s elections by looking at historical electoral data.

2020 Turnout by Race

We begin by looking at turnout dynamics of the constituent races of the New Mexico electorate. Given the highly diverse nature of the state, it makes sense to break down into regions - of which we have made seven. They will be discussed in further below, alongside graphs of turnout by precinct relative to the white/Latino/Native share of the precinct, as needed.

Albuquerque

Above are the turnout charts for the Albuquerque region. This region comprises the counties of Bernalillo, Sandoval, Torrance, and Valencia. It includes most of the core of the Albuquerque metropolitan area - not the entire media market, as that would cover the entire state and render this exercise pointless. It ranges from the deep-blue urban core to swingy suburbs, both blue-collar (Los Lunas) and white-collar (Rio Rancho), encompassing Native American pueblos and Latino farming communities alike. This region is generally bound together by close ties to Albuquerque..

A pattern we see here, and will likely continue to see, is that whiter precincts tend to punch above their weight in turnout, whereas the opposite is true for Latinos. This is a natural conclusion, given that whiter precincts tend to be less Latino, coupled with many New Mexico Latinos retaining non-citizen status. Among all of this - we can see that Latino voters in the Albuquerque metro hold lower voting propensity than whites. The regression line has a negative slope for the Latino chart - meaning that as a precinct gets more Latino we can expect proportionately lower turnout. From this, we can deduce that Latino voters do not punch as much weight - since they tend to be highly concentrated in specific areas in the predominantly urbanized Albuquerque metro.

Speaking at a macroscopic level, the Albuquerque metro has steadily been gaining vote share. In 2002’s gubernatorial race, the 4 counties of the Albuquerque region cast 43% of the statewide vote. That number rose to 46% in 2006, 45% in 2010, 45% in 2014, and rose to 47% in 2018. I don’t love assuming only raw numbers affects future electoral trends without any hope for persuasion - but rural New Mexico’s voting power has diminished greatly in recent years. This growth is powered by many educated professionals moving to the area - Bernalillo and Sandoval have around 30% college attainment - significantly higher than most counties in New Mexico. Incomes are also fairly high relative to the rest of the state, with Bernalillo and Sandoval’s both above $50,000 for median household income. Valencia and Torrance are both more working-class, both looking by education and income metrics - but their population is dwarfed by the other two counties. Poverty statistics for these counties are in the teens - though higher in Valencia and Torres, as one would expect.

Dinetah

Pictured above are the turnout charts for the Dinetah region. This region encompasses three counties in northwestern New Mexico, namely the counties of San Juan, McKinley, and Cibola. It is a fairly evenly split region that narrowly leans Republican - characterized by intractable racial polarization between highly conservative white voters and Democratic-leaning Native American voters. This region encompasses the overwhelming majority of the Navajo Nation’s presence in New Mexico. As such, it makes sense for us to add a third racial turnout chart for Native voters. This is the only region of the state where Native Americans make up a substantial portion of the electorate, and thus their turnout dynamics must be explored.

This region does not have too many Latino voters, especially compared to the rest of the state - most of the political dynamics here play out between Republican whites in Aztec and Farmington and Democratic-leaning Navajo, Hopi, and Zuni in Gallup and Shiprock. The races are extremely segregated, as evidenced by the preponderance of 0% and 100% precincts in the charts. Analyzing Native voting patterns does not lead to the same issues of noncitizen population as with analyzing Latino voting patterns - after all, Native Americans have had the rights of full citizenship since the Indian Citizenship Act of 1924 and are the original inhabitants of the land that the United States rests upon.

As far as propensity is concerned - Native American voters, particularly in rural northwestern New Mexico, can greatly influence elections if the work is put in. Due to the region’s isolation, it often takes serious effort to organize and register new voters; neighboring Arizona saw unprecedented get-out-the-vote efforts among its Native population with canvassers riding on horseback for days amid a pandemic that hit Native communities especially bad. Republican candidates for office, particularly in the Four Corners region of Arizona/Colorado/Utah/New Mexico, have not given Native communities a lot of reasons to vote red - but Democrats cannot take their votes for granted as the raw turnout these communities provide can often be the difference in a close race. I did some crude math a few months ago that showed that the turnout boost that Native American communities provided Joe Biden with in 2020 netted him roughly 10,700 votes over Hillary Clinton’s 2016 margins in those areas (Biden won the state of Arizona by 10,457 votes - meaning that his efforts to court Native American voters were arguably the decisive factor.) The chart shows that the reservation precincts (those close to 100% Native) have a slight dip in turnout compared to whiter areas - but they tend to be bluer than the whiter areas.

Now that I have finished this digression - politics here is almost all turnout and little to no persuasion. The reservation precincts that are almost 100% Native are positioned in a clump on all the graphs but off the reservation, the same issues of propensity rear their head. Both Democrats and Republicans are fortunate, in a sense, that the political cultures of the reservation and the off-reservation are so different - the parties can afford to put all their eggs in a single basket and have it pay off.

Finally, with respect to education and income - this area is low in educational attainment. White-dominant San Juan has 15% college attainment, and in more Native-heavy McKinley and Cibola this number drops to a mere 13 and 11. With a MHI of over $50,000, San Juan is also distinctly wealthier than the other two counties here, while the other two counties are stuck in the 30,000-40,000 range. San Juan has a 20% poverty rate, compared to 26 for Cibola and 31 for McKinley - some of the highest in the state.

Frontera

The Frontera region (Spanish for “border”) is aptly named, as it is closest to the Mexican border and shares a lot of cultural and political similarities to its sister counties in the adjoining states of Texas and Arizona. It encompasses the counties of Catron, Doña Ana, Grant, Hidalgo, Luna, Sierra, and Socorro. This region is anchored by the city of Las Cruces, the seat of Doña Ana County. In many ways it is an adjoining agglomeration next to El Paso in Texas - by far a more dominating city on the Mexican border. Depending on which census estimate you use, up to 40% of the county’s population is composed of non-college-educated Latinos. Though somewhat surprisingly, Las Cruces and Doña Ana have actually gotten more Democratic. Very minor rightward 2016-2020 swings in the most Latino of precincts were balanced out by strong left swings in localities outside of the urban core - especially in the University Park neighborhood near New Mexico State University.

Outside of Las Cruces, the adjoining border counties are very similar to the famed South Texas border counties that zoomed rightwards by upwards of 40 points - Hidalgo and Luna counties are both Obama-Trump counties. Unlike South Texas however, these counties were not uniformly Democratic in the immediate past - George Bush won them both twice and they have remained competitive. That said, the Latino shift rightward took these counties off the board as Trump won them easily in 2020. Latino-heavy border towns like Lordsburg and Deming do not provide the same blue ballast that they once did. The rest of the region to the north is very sparse and rural - and very conservative as well.

The counties that make up this region tend to be predominantly poorer than areas to the north - the average poverty rate is well over 20% of each county’s residents. This statistic holds true measuring by income as well - the average median household income in these counties is just under $36,000 - significantly less than what it is in Albuquerque. Education rates further cement this story, with college attainment hovering in the 20s.

The Frontera is the Democratic powerhouse of southern New Mexico - and Democrats here are increasingly dependent on non-college educated Latinos as their older Obama-era margins with rural whites slip. That tradeoff comes with its perils - recently legislative and local elections in the area have flipped to Republicans as longtime conservative Democratic incumbents have lost their primaries to progressive Democratic challengers. It suffices to say that most non-college-educated Latinos in this part of New Mexico, like their Texas counterparts, are culturally conservative and maintain a strong independent streak.

Little Texas

Little Texas comprises the five counties of southeastern New Mexico: Chaves, Eddy, Lea, Lincoln, and Otero. This region is so named because it is culturally an extension of West Texas - these areas are hugely rural, dependent on oil, and home to tens of thousands of culturally conservative Protestant Latinos who vote Republican - unlike the more Democratic-leaning Catholic Latinos in the southwest and the rest of the state. Prior to the great party switch of 1964, this was the most Democratic region of the state but of course, it is now the most inhospitable region for Democrats. Its proximity to the Texas border, combined with its cultural conservatism and heavy dependence on oil produced some of the strongest Republican swings in the state between 2016 and 2020. The most oil-heavy areas around Carlsbad in Eddy County and Hobbs in Lea County both swung 11 points to the right. For rural New Mexico, these areas are fairly populated, with over 150,000 residents total.

Compounding this region’s Republican lean is its relatively high income for a rural area - no doubt attributable to the oil and gas industry. Lea and Eddy Counties have incomes upwards of $50,000 and the rest of the counties are above 40,000. Educational attainment hovers in the teens here - and poverty is much lower than it is in southwestern New Mexico - lowest in oil-heavy Eddy, Lincoln, and Lea. Thus Little Texas represents a particularly disfavorable demographic cocktail for Democrats; it is a region brimming with lower-education, culturally-conservative, medium-to-high income voters.

This region is where Republicans draw on their raw turnout - for example, NM-02 Rep. Xochitl Torres Small (D) lost to Yvette Herrell in 2020 largely due to being swamped by conservative voters in Little Texas brought out by Donald Trump’s magnetism. Latino voters here are also lower-propensity than white voters, as evidenced by the negative slope of the regression line. The marginal voters here are likely the most conservative in the state - as Little Texas can be considered the westernmost extension of the cultural South due to its cultural similarities to actual Texas.

Northeast

The Northeast comprises the counties of Curry, De Baca, Guadalupe, Harding, Quay, Roosevelt, and Union. The largest city here is Clovis in Roosevelt County - but overall this region does not pack quite the punch as the cities of Little Texas do. Latino percentage is probably weakest in this portion of the state though it still exists - it is fairly similar to the Texas/Oklahoma Panhandles, or southeasternmost Colorado. The relative lack of Latino voters is most evidenced by the wide variance from the regression line. It is fairly low education, lower-income, and has poverty rates consistent with most of rural New Mexico as well. This is by far the smallest region of the state with only a few tens of thousands of people.

Novomexico

The Novomexico region comprises the counties of Colfax, Mora, Rio Arriba, San Miguel, and Taos. This is by far the region I understand the most about, given my prior work studying and reporting on the San Luis Valley in Colorado. The San Luis Valley is the northernmost extension of the Greater Novomexico region, and both sides of the border share many cultural similarities. This region encompasses touristy ski-towns in Taos, farming and ranching communities in Las Vegas and Mora, and mining communities near Raton. It is an overwhelmingly Latino and Catholic region - white political power is almost nonexistent here. In short, it is what Colorado’s fabled 3rd district would look like if Latinos were an overwhelming majority of the population. This region is small but by no means unimportant - in recent history these counties have been the most Democratic in the state, even amid Republican routs.

In contrast to Latinos elsewhere in New Mexico, this region is home to a predominantly Spaniard/Hispano Latino composition - from which this region takes its name. The Hispanos are descendants of the original Spanish settlers and conquistadors who first arrived in the area in the 16th century to settle the upper Rio Grande watershed. Many have assimilated, either voluntarily or forcibly at the hands of Anglo settlers - but Hispanos retain a unique culture with a unique variety of Spanish spoken too.

They are politically unique too in one regard - they live in the only region where more-heavily Latino precincts do not suffer from a decrease in propensity relative to less Latino precincts. One potential confounding issue in divining the true source of the uncertainty around Hispano voting patterns lies in the fact that many Hispanos identify as white Hispanics - meaning that there could be an issue of collinearity not present with other New Mexico Latinos. One possible hypothesis for Hispano voting behavior to supplement this potential issue of collinearity is that due to a longer presence in the United States relative to other Latinos, Hispanos have assimilated more clearly into white voting patterns than have other Latinos. Still, like other rural Latinos, they have swung right recently - Taos is an exception due to its growing skiing/tourism presence but Mora/Rio Arriba/San Miguel have all moved between 5 and 8 points right between 2016 and 2020. This is an area which stands to fall rightwards due to educational polarization, as it is full of less-educated Latinos with lower incomes. Democrats are trading their old dependence on these rural Latinos for new suburban college-educated whites and Latinos in the Albuquerque metro.

Santa Fe

Our seventh and final region is the Santa Fe region - made up of the counties of Santa Fe and Los Alamos. Home to many college-educated whites, as well as staunchly Democratic Latinos and Native American voters, this region is cobalt blue, save for a few southwestern precincts around Edgewood that function as Albuquerque exurbs. Santa Fe is likely another Democratic card to play, in addition to suburban gains in Albuquerque, to stem the tide of rural losses. Los Alamos is famously where the atomic bomb was tested - and is now home to the fastest Democratic-swinging community in the state, made up of highly-educated professionals. Los Alamos and Santa Fe are the two most educated counties in the state - which is a huge boon to Democrats riding a wave of education polarization.

In short, this region is insanely Democratic, and will likely remain insanely Democratic.

Now that we have examined all seven regions and their relative propensities, let us look at the next important factor of propensity: presidential vs. midterm turnout.

Midterm vs. Presidential Turnout Dynamics

2022 is the big question on everyone’s mind - how is New Mexico going to vote? One of the biggest factors to think about is the fact that the electorate that shows up in presidential years is different than the one that shows up in midterm years. This factor is arguably one of the key reasons that Xochitl Torres-Small won the southern-based 2nd district - she didn’t have to contend with tens of thousands of Republican voters because they simply did not care to vote. As my good friend Lakshya says, many of these folks probably didn’t know there even was an election. Jokes aside though - the midterm electorate is a unique beast - particularly in a state with a lot of moving parts from a demographic perspective, like New Mexico.

Lakshya has also conducted excellent analysis on the importance of minority turnout to Democratic hopes in a midterm. Much has been made of the “global realignment” of college-educated whites and college-educated voters in general. Namely, there is an idea floating around that the advantage in propensity that Democrats get by trading non-college voters for more educated voters would be enough to withstand any other unfair dynamics. I even wrote about this in the past with regards to the Georgia Senate runoffs. But anecdotes aside, Democratic strength in a midterm depends on energizing their minority voter base. And in New Mexico, that base comprises two main groups - Latinos and Native Americans.

The above graph shows the turnout dropoff between the 2016 presidential and 2018 midterm elections in the Dinetah region. Interestingly, the whiter areas had a dropoff ranging between 0 and 5% of the 2020 total population (the 2016 and 2018 raw population numbers are smaller so the percentage dropoff is likely higher). Native areas had virtually no dropoff - and a few precincts actually increased their vote turnout - so Native voters appear to punch above their weight. This is one solid card that New Mexico Democrats have to play - their organizational strength in the northwest is solid. One caveat of this is that 2018 was a bluer-leaning year - Republican-leaning white voters might have simply been less excited to vote. 2022 is shaping up to be a strongly Republican-leaning year - so these voters with high levels of cultural conservatism and racial anxiety could be more motivated to vote. One reasonable comparison to make would be to look at 2010 or 2014 precinct data - but the boundary lines have changed too much to make meaningful comparisons. In short, 2016-2018 is the best empirical dataset we have for analyzing the turnout dropoffs and we can still learn some lessons.

Looking at the Frontera region’s Latino population - we find more evidence to support the thesis that Latino voters tend to be lower propensity than whites. There’s a roughly 4-5% differential between the most and least Latino precincts in terms of the turnout delta. We saw similar turnout dynamics in culturally similar regions of other states - notably in Texas, where weak turnout and reduced Democratic margins in the border region and the Rio Grande Valley contributed to dashed Democratic hopes. Similar turnout dropoffs occur in Little Texas and Novomexico as well.

This article does not diagnose the causes of lower voting enthusiasm among Latino voters - there are people far smarter than me who can opine on that subject. But - given the renewed focus on New Mexico’s status as a competitive state, Latino voters will be a make-or-break piece of a winning coalition. Democrats will need to keep Latinos energized - a task which may be running into some headwinds. It is our view that one of the prime reasons New Mexico Latinos are still broadly more Democratic than more Republican-leaning Latinos in Texas and Florida, is because the New Mexico GOP’s Latino outreach is not as central to its outreach as it is in those other states. As a result, New Mexico Democrats enjoy certain state-party effects with Latino voters that may not hold up long-term. Unlike Texas and Florida though, New Mexico Democrats have a strong record of running candidates with strong Latino appeal. (Florida Democrats have nominated a Latino statewide candidate only once in history). The net result of the electoral forces that are meeting in New Mexico’s Latino electorate cannot be simply stated yet. At this stage, this far out from the election, a qualitative judgement is easier to make than a quantitative judgement.

Looking at the broad Latino trends from 2018-2020 in terms of raw turnout - the increases are strongest in the whitest precincts everywhere, as expected. For whites, the presidential turnout increase is positively correlated with white share everywhere, but is strongest in the regions of the Northeast, Frontera, and Little Texas. This is expected as Trump brought out a lot of low-propensity culturally conservative whites. Looking at Latino voters, there is somewhat of a negative correlation between turnout increase and the Hispanic share of a precinct. As usual, Novomexico stands out - once again proving that Latinos in this part of the state are unique and have a political voting culture more similar to white voters. The Albuquerque metro is also interesting here too, as it is the largest part of the state and shows a Novomexico-esque resistance to propensity drops. The Frontera, Little Texas, and the Northeast all see drops.

This divergence between the voting patterns of urbanized and more rural Latino voters has been growing lately - and has been especially visible in post-Trump elections such as the California recall election. Urbanized Latino-heavy precincts in Los Angeles and San Diego saw swings towards Clinton high-points or stabilizations, while more rural Latino precincts in the Central Valley and in Imperial County (Mexican border region rurals) saw further slippage to the right. Excepting the unique case of Novomexico, Albuquerque has a different Latino propensity pattern than more rural regions of the state.

Finally, looking at the turnout differential in Dinetah, the dropoff was minimal - reservation precincts had a turnout increase on par with whiter precincts in San Juan County.

In short, these regressions confirm that Native American voters punch above their weight for Democrats, and that Latino voters outside of the Albuquerque metro tend to suffer a turnout drop. Since most of these voters tend to benefit Democrats, it’s fair to say that it could cost Democrats up to a point or two of statewide vote margin if the drop is severe enough. And once again, my good friend Lakshya’s work is confirmed that energizing minorities is the key to retaining large Democratic margins.

Regional Breakdowns, 2016-2018-2020

Below, we have used regressions and ecological inference to break down approximations of the partisan vote share by race for each of our seven regions. These do come with a little bit of error, as all statistical estimation does - but we believe these numbers are a rather accurate assessment of the shifts that have gone on in New Mexico between 2016 and 2020. Note - there is no easily available source of educational attainment data that one can gather at the precinct level - and to artificially construct a dataset from Census block data would be a project beyond the scope of this article.

As an aside - the 2016 presidential numbers are highly influenced by the presence of Libertarian Party candidate Gary Johnson, who previously served as Republican governor of New Mexico - the partisan splits incorporate the Johnson factor and thus for 2016 the vote is not two-party.

Albuquerque

The two biggest stories of the Albuquerque metro are the Latino shift as well as the inexorable march of whites in the area to the Democratic Party. White voters in the metro are predominantly college-educated and represent an increasing share of the electorate. Hillary Clinton and Michelle Lujan Grisham lost this demographic handily - but Biden held Trump to only 51% and only lost whites by 5 here. The recent New Mexico special election for the 1st congressional district, based in the Albuquerque metro saw continued gains with this group - dampening support for any idea of white reversion post-Trump. These areas include the communities of Rio Rancho and Corrales in Sandoval County, and the neighborhoods near North Albuquerque acres. Republican DNA still persists here - many precincts in this area voted for Biden and for Mark Ronchetti for Senate.

Looking at Latino voters, we find the fabled Latino shift to the right once again. It is repeated in the Albuquerque urban core, in suburban neighborhoods like Atrisco and Arenal, and in rural farming communities further south in Belen and Los Lunas. The shift from 2016 to 2018 shows that Michelle Lujan Grisham recouped most of Gary Johnson’s vote share - but those gains were wiped clean in 2020 when Democratic Latino vote share decreased and Republican Latino vote share increased. Biden only won Albuquerque-area Latinos by 27, compared to Clinton’s 34-point margin. Latino voters are not a monolith - and the reasons for and the mechanics of each individual Latino subgroup’s shift will vary - but this 7-point shift was among the smaller shifts seen.

Speaking of Native voters, it is clear Hillary Clinton’s campaign had an anemic organization that allowed many votes to bleed to Gary Johnson. Michelle Lujan Grisham dramatically improved over Clinton and Biden held serve suffering minimal defection.

Dinetah

Once again, we find our fabled Latino shift here - but Latinos do not make up as large a share of the electorate in these three counties. The Latino voters here tend to be more Republican leaning, perhaps as a consequence of their more rural makeup. The real battle royale is in this region is always between Republican-leaning whites and Democratic-leaning Native voters. It did not disappoint.

Democrats managed 1 percent of the vote, plus or minus a few points (basically no whites here vote for Democrats), while Republicans only suffered minor defections to Johnson in 2016 - Trump and Pearce both managed mid to high 90s in their white support. Native voters were Clinton +52, then MLG +55 and Biden +53 - these shifts are small but are not significant - any increase in raw turnout would negate their rightward shift. To the extent that persuasion-based campaigns exist here, they are in the capacity of persuading lower-propensity individuals in one’s base to vote rather than persuading people to cross party lines and switch. As mentioned before - similar turnout patterns on other side of the border in Arizona were crucial to winning that state - so there is a lot of cross-border political pollination that occurs here. The Navajo nation spans three states and numerous other tribes inhabit the area - so there is fertile opportunity for impacting many states’ elections with future political pursuits here.

Frontera

2016 data for Native voting is not included in this region because the error estimates are too high due to a malformed configuration of the dataset used and the relatively low Native population in the area. That said - this area bears some interesting lessons. Whites in this region have gotten successively less Republican - even taking into account Gary Johnson in 2016. This is likely a consequence of a few whites in Las Cruces, but also the presence of the smaller tourist town Truth or Consequences - with multiple hot springs and related wellness centers.

The big red flag here is the Latino shift. Hillary Clinton won Latinos here by 53, 72 to 19. That mostly held up in 2018, wih MLG winning this demographics by 54. This Democratic fortress came crashing down - with Biden winning these Latino voters by 41 points. That represents a 13-point shift from 2018 and a 12-point shift from 2016. This is consistent with other shifts we have seen along the Mexican border - ranging from Nogales in Arizona to Imperial County in California. It is unknown whether or not these shifts will continue - but given that Las Cruces, Lordsburg and Deming are more rural and less connected to large educated metros, it would not be ridiculous to hypothesize that they would continue to shift right. This Latino shift is a flashing-red warning sign that Democratic strength is in trouble with voters in southwestern New Mexico.

Little Texas

This area, culturally similar to West Texas, votes exactly like West Texas. Republicans vote roughly 87% white or so, and Democrats seem to get 12% of the vote. Native voters are fairly minimal in this area but are likely Republican-leaning due to their rural distribution.

Once again, the Latino shift is extremely stark here. Clinton tied Trump with Latinos here in this strongly conservative region - and Michelle Lujan Grisham held her own here, only losing by 3. Joe Biden, to put it frankly, tanked. Trump won Latinos by 24 in this region of Texas. It makes sense when you factor in that southeastern New Mexico is heavily dependent on oil, in addition to being predominantly rural and close-ish to the Mexican border. Lea and Eddy Counties shifted 11 points to Trump - almost certainly influenced by a 21-point rightward shift.

This shift likely doomed downballot Democrats here, especially Xochitl Torres Small. Her overtures to the oil and gas community in southern New Mexico were to no avail - as the surge in rural white turnout plus the Latino shift destroyed any chance of survival. Former Democratic U.S. Rep. Harry Teague, who once was an oilman from Hobbs in Lea County, actually endorsed her Republican opponent, as a sign of how conservative this area is. In short, an unforgiving region for Democrats has become utterly desolate.

Northeast

White voters in the northeast got collectively more Republican - Democrats are stuck in the low single-digits with whites here. 2020 does not have Native data because of data configuration issues and due to an extremely small sample size to conduct regressions. Suffice to say though, that Native voters in northeastern New Mexico are Democratic-leaning.

The interesting story here is that Latino voters in the northeast did not shift to the right dramatically, as Biden mostly matched Clinton’s numbers here. Johnson took about equally from Democrats and Republicans here. There are Democratic-leaning Latinos in Clovis, as well as in Guadalupe County - and some more Republican-leaning Latinos to the far rural northeast. Still, due to the region’s size, it does not have a particularly outsized impact on state politics outside of the closest races.

Novomexico

Novomexico is one of the few rural regions of the state where Democrats do not get stomped on. Democrats hold their own with whites here - which could be attributed to three things: 1) Close ties to the Santa Fe metro mean that some precincts very close to the city are influenced by education/urbanization factors. 2) Many whites are also cross-identified as Hispano and have residual Democratic loyalty. 3) Non-Hispanic whites in Democratic-leaning minority-heavy areas have shown to be been more Democratic than non-Hispanic whites in whiter areas (this pattern is observable from a historical standpoint, all across the South). With respect to Native voters - Biden exceeded MLG’s margins here - evidently, his campaign did something right with Native Americans.

As far as Latinos are concerned - Biden mostly matched Clinton’s vote share - but Republicans have steadily improved here. The 2016-2020 shift is rather small, at Biden +38 vs. Clinton +42. But from 2018 - Biden did do noticeably worse here than MLG. The Hispano shift is indeed real but this is another data point that confirms that up until the present it has been smaller than that of the Tejanos in Texas and of other border Latinos elsewhere.

Santa Fe

We are at the crazy point, now, that Santa Fe whites are more Democratic than Santa Fe Latinos. One of the only other places I can think of that this is true, is the Bay Area, where whites are more Democratic than Asian voters, the second-largest racial bloc in the area. Native voter data isn’t easily convertible into a regression and thus the data is not fully available. One would guess that they are Democratic-leaning and either stabilized or swung left over 2016-2020.

Santa Fe whites’ Democratic trend is monotonous, much like the whites of Albuquerque - these are many college-educated whites that are moving to the area for the lifestyle and for the jobs. The Los Alamos National Laboratory employs almost 13,000 people and has an entire community ecosystem to support it. The Latino shift here isn’t all that large either - Biden only did 2 points worse than Clinton (Biden +48 vs. Clinton +50). To make a long story short, this is the opposite of Little Texas - instead of a highly Republican-leaning demographic profile that is getting redder, the Santa Fe area has the potential to get even bluer. The two cannot wholly counteract each other - as Little Texas is about 70% larger than the Santa Fe region - but there is indeed potential for the areas to offer some neutralization in terms of statewide margin.

To conclude this section, let’s finish by stating that there are good and bad things for Democrats and Republicans alike. Democrats have a growing white suburban base in the Albuquerque metro as well as a ride-or-die Native American voting bloc, and Republicans have a rock-solid rural white base that has the potential to become a rural white and Latino base. There is nothing demographically in this state that should doom Republicans or Democrats to the wilderness.

As a sanity check, let’s look at actual exit polls to see how certain groups voted statewide. According to Fox, Biden won whites, 49-48, while Biden won Latinos 61-38. Native Americans are not specifically identified in this poll, but the category of Other, under which Native Americans must no doubt be included, voted for Biden 53-44. I would say that the numbers we have identified generally comport with reason, even if they do not perfectly align with exit polls.

Now that we have looked at the past - let us look at the future: to 2022.

2022 preview; electorate screen

With some quick and dirty math using the CNN exit poll electorate screen from 2016, we can calculate the 2018 midterm electorate. For the racial categories of white, Latino, Native, and everything else (black/Asian), the percentages are 49/40/9/2. In 2018, using very crude calculations we can estimate those same numbers respectively to be 57/34/8/1.

This would indicate that the groups that punch above their weight the most are whites and Native Americans. Splitting the white vote by education is not possible - since the precinct-level data cannot be disaggregated by education - but a wealth of electoral studies by Lakshya, Kraz Greinetz, and other analysts have shown that college-educated voters (especially college-educated whites) punch above their weight in midterm/lower-turnout environments.

At the same time, Latino voteshare went down from 40 to 34 percent. Given that Latinos are still a broadly Democratic-leaning group, this probably cost Democrats a point or two of statewide margin in the pre-Trump era when lower-propensity Latinos were accepted as an untapped pool of Democratic strength. Now, that picture is less clear - but given the recent stabilizations we have seen with urbanized Latinos, particularly urbanized Mexicanos, Democrats should want as many Latinos to turn out in their urban areas. Obviously turning out voters elsewhere is the most important, but the most shift-resistant Latino groups are likely concentrated in Albuquerque and Santa Fe - the Frontera and Novomexico are less certain long-term as they are more rural and working-class in character. High turnout in an area like Little Texas probably is one ingredient of many in Democratic doom - as those voters do not show many signs of swinging back to Democrats as they are extremely conservative.

Native American voters barely lose any ground at all, going from 9% of the electorate to 8%. This demographic is somewhat analogous to black voters in Georgia - who punch above their weight in lower-turnout scenarios. Democrats fearful of statewide slippage should try to get that Native share as high as possible - especially to offset any further slide with Latino voters.

Up until now we have not acknowledged the presence of black and Asian voters in New Mexico - as overall they are a fairly small concentration. But the ones that do cast ballots tend to be solidly Democratic-leaning demographics.

In 2022, what can we expect? Native Americans will likely hold their ground in the electorate and the propensity deltas with whites/Latinos will exist still. In an unfavorable national environment, the partisanship of every demographic probably will swing uniformly to the out-party, excepting specialized campaigning on a particular set of issues. We saw this pattern in New Jersey and Virginia in 2021 - Glenn Youngkin and Jack Ciattarelli both broadly outperformed Donald Trump everywhere without specific appeal to individual demographics. The only logical conclusion is that people of all demographics were equally affected by negative macroscopic factors (economy, Covid, gas prices, etc.) and chose to send a message to the incumbent. It is with these factors in mind that we can now look at the Governor’s race in 2022.

2022 Preview: Governor

New Mexico is like and unlike those states - the wealthier, more college-educated suburbs around Albuquerque and Santa Fe still have the potential for ticket-splitting. Some of the Democratic margin in the state in 2020 absolutely can be attributable to an “anti-Trump” factor rather than more permanent Democratic persuasions. Unlike those states, however, it has a large demographic that has much potential to swing to the right: conservative and lower-education Latino voters.

To win as a Republican in New Mexico, you need a few things to happen: 1) a good national environment, 2) substantial ticket-splitting in the Albuquerque metro, 3) rural turnout among whites and Republican Latinos that matches or exceeds Trump, and 4) a Democratic Party that is weak and/or divided. New Mexico is a Democratic state - the days of Pete Domenici winning by double digits are long gone. But the state did have a Republican governor as recently as three years ago, so the possibility of a blue-to-red flip in Santa Fe definitely merits discussion.

Step 1, the national environment, seems to be well on its way. Republicans are favored to regain a majority in the House, and are favored to win the Senate assuming that they do not nominate the Barnum and Bailey Circus as their slate of candidates.

Step 2, generating significant ticket splitting, seems to be achievable with their presumptive nominee for Governor, weatherman and 2020 Senate candidate Mark Ronchetti. Ronchetti outperformed Trump and came within 5 points of winning the Senate race in 2020 - largely by improving everywhere in the Albuquerque media market which his weather station serves. By all accounts, Ronchetti has solid conservative viewpoints but is not a fire-breather that alienates the weak-partisan suburbanites of Albuquerque. In other words, Ronchetti has the potential to be the it candidate that can achieve step 3: generate staunch rural turnout while holding your own in the metro.

It also bears repeating that there is an increasing ideological disconnect between the Latino communities in the border and the direction of the Democratic Party coming out of more urbanized areas.

The above map from analyst MoreThanPol shows the cessation of the 35th Senate District, based in Lordsburg, Deming, and Truth or Consequences. Conservative Democratic incumbent John Arthur Smith lost his primary to Neomi Martinez-Parra, a more mainstream left-leaning Democrat, in 2020. Martinez-Parra had none of the crossover appeal of Smith and lost the seat amid the border Latino shift. Similarly - in the Silver City-based 28th Senate District just to the north, incumbent conservative Democratic Sen. Gabe Ramos easily lost his primary to Siah Correa Hemphill - a more progressive Democrat. Correa Hemphill survived the general by the skin of her teeth. It will be interesting to see how southwestern New Mexico’s Democratic proclivities hold up over the next few election cycles.

The final step is the fourth - run against a weak or divided Democratic Party. Michelle Lujan Grisham won comfortably last time - but her tenure as Governor has run into a few snags. Most importantly, she has had to deal with a harassment scandal which has hurt her image a bit. She also has some ethical issues with regards to healthcare, especially pertaining to accusations that she profited off New Mexico’s high-risk insurance pool. Other allegations include violating her own stay-at-home orders, and ignoring abuse complaints in long-term care facilities. Given the age of some of these issues, it is unlikely that they directly affect the race. But considering Ronchetti’s rather fresh face against a career politician who has a few battle scars, the potential is absolutely present for an upset.

Taking into account all of these four factors, combined with the demographic information from the past analyses, Split Ticket justifies its rating for New Mexico’s gubernatorial race as Lean Democratic.

Let’s now look at the congressional races that are up in New Mexico.

2022 Preview: House

We now examine the New Mexico’s 3 new congressional districts.

The maps are clearly drawn with partisan intent to eliminate the Republican lean of Rep. Yvette Herrell’s (NM-02) seat to replace it with a Democrat. As such, the Albuquerque metro and its many Democratic voters are cracked between the the 1st and 2nd district and deeply-Republican Little Texas is spread between all three seats. Both existing blue seats are dramatically weakened and the 2nd is turned into a seat that Biden won by 6 points.

The 1st takes in the eastern half of Bernalillo as well as suburban Sandoval, and stretches southeast as far as Roswell. The 1st’s portion of Bernalillo is whiter and more Democratic-trending, which is designed to cancel out any Republican shifts in the more rural areas. The incumbent here is Melanie Stansbury, and she would likely be re-elected with ease in this seat that voted for Biden by 14 points. Due to ticket-splitting, we anticipate that 2022 may be somewhat close - especially given the national environment and the GOP’s persistent presence in the Albuquerque suburbs. The logical rating we have here is Likely Democratic - as an incumbent, Stansbury should not have a problem unless the things go really far south for Democrats.

The 2nd is now blue-collar Latino central - stretching from the southeasternmost corner of the state all the way to downtown Albuquerque, taking in Las Cruces and the Frontera region in the middle. Yvette Herrell is from Alamogordo, in Otero County, and has her seat moved from Trump +10 to Biden +6. Her return to Washington is significantly complicated by this new map - especially because of the addition of 200,000 new Albuquerque-area voters, who lean very solidly Democratic. That said, we already know that Latino share of the electorate dips in midterms and while there is a right-swing, they are still a blue voting bloc. Democrats are thus disadvantaged by the midterm turnout dynamics here - this factor, plus Herrell’s incumbency and retention of enough of a base in southeastern New Mexico make the new 2nd a complicated case. In a neutral year, a Democrat like Xochitl Torres Small could make a strong case for a comeback, but we believe there is no clear favorite here. Our rating is a Tossup.

The 3rd is the most interesting district - it stretches from most of the way from Durango to Odessa - a huge distance to traverse. Since the 2nd had to shed much of Little Texas, the 3rd had to absorb much of it. Most of the city limits of Clovis, Hobbs, Roswell, and Artesia are all in the new 3rd - hundreds of miles from the old base of the 3rd. In the northwest, the Dinetah region is broadly contained here - except for a few Puebloan precincts in the south of Cibola County. Santa Fe is at the heart of this district and is the big Democratic anchor here. Rating this seat is arguably the toughest one here - it is not a tossup, but what is it?

One can be bullish on Democrats due to the propensity edge they enjoy with Native voters, particularly in the northwest. There is also Santa Fe, the blue jewel in New Mexico Democrats’ crown that is only getting bluer. Democrats also have held up fairly well in Novomexico with Hispanos up until now - and Teresa Leger Fernandez has a very strong appeal to these voters. Part of this is due to the fact that as a Latina candidate, she enjoyed a “solidarity boost” among Latino voters - particularly when the New Mexico GOP tends to run Anglo candidates, or non-visibly-Latino candidates for office. She outperformed Biden in most of the Novomexico region, sometimes by double digits.

In favor of the Republicans, here are a few arguments. There is a ton of new territory in this district that Teresa Leger Fernandez has no connection to. The interests of the Rio Grande watershed and the interests of oil country require different approaches to politics and the voters in Hobbs and Roswell will likely be angry at their new representative. Furthermore, much of this new territory has Republican-leaning Protestant Latinos, with a different approach to politics from the more Catholic-heavy areas in northern New Mexico. Speaking of Catholic-heavy areas in northern New Mexico, we don’t know when Democratic luck with Hispanos will finally run out - the now-commonly-accepted theory of educational polarization says that at some point it will. Finally, the best argument for Republicans is the raw partisanship of the seat. Instead of the Biden +17 (57-40) that is the current seat, the seat is now Biden +10, 54-44. A good rule of thumb is that in large wave elections (which 2022 is shaping up to be), any seat under 55% for the incumbent party is fair game. NM-03 clocks in at 54% Biden, just under the wire. A good Republican candidate could possibly overcome that with a solid environment and some persuasion.

We settled on Likely Democratic for the 3rd, since Leger Fernandez is an incumbent and still enjoys popular support among Latino voters in her existing base in Novomexico. This is not enough to merit a safe rating due to 1) the national environment and 2) the new 150,000 or so Republican voters added to the district in the southeast. Another Latino shift could take Leger Fernandez out, but based on all available evidence at present, the macroscopic factors necessary for such a sea change are not present. We will just have to wait and see whether it happens.

Conclusion

Longer-term, we think New Mexico is still going to be a Democratic-leaning state, but one that incorporates more and more of a liberal white coalition based around Albuquerque, rather than latent Democratic strength with Latino voters as has traditionally been the case. Currently, Republicans grasp at the governorship, mostly on a ladder of white suburban ticket splitting - but in future, it is perfectly reasonable to hypothesize more blue-collar Latino voters joining the GOP as they have done in Texas. In terms of blue-leaning but not immensely-so states with large blue-collar Latino populations, there is one other southwestern US state that is a more immediate danger for Democrats - but there is enough of a demographically explosive cocktail that Democrats should keep a long-term watch on the state’s competitiveness.

This report details the political shifts that have passed in New Mexico in the past three electoral cycles. There are good and bad things for Democrats, and for Republicans. Here are just a few. Democrats should appeal to Native voters more than they do - and work tirelessly to protect their margins among Latino voters - which are a lot better than they should be in many rural cases. Furthermore - it bears repeating that Biden won whites - this is unprecedented. If New Mexico Democrats can hold their own with Latino voters and continue to win large majorities - New Mexico Republicans are left with limited options if they cannot even win whites - the state is increasingly becoming more Latino. For Republicans - New Mexico is a state where educational polarization could make the state much more competitive. Issues that activate voters and make them pick a side on the culture war are the principal drivers of these trends (I can write about this at a later date) - and the New Mexico GOP has plenty to pick from in a state filled with blue-collar cultural conservatives. Oil, immigration, education perhaps as well.

We do not hold a palantir or a mirror to look in the future. But we can say with confidence that New Mexico’s political shifts have the potential to greatly reorder the existing balance of power in the state. In cracking Albuquerque and diluting Little Texas, New Mexico Democrats are increasingly reliant on blue-collar Latinos to hold down their majorities. In light of the finds of this report, we shall see if it holds up.

Sources and Special Thanks

Dave Bradlee, Alec Ramsay (Dave’s Redistricting App)
Ethan Chen
MoreThanPol
Stephen Wolf
Lakshya Jain
Kavi Gupta
Clare Considine
alexanderao