An electoral revolution: the political evolution of Brazil’s Nordeste

Changes in a region’s voting patterns and overall political mood are often spread out over a long period of time, rarely happening overnight in one election. Possibly, the electoral evolution of the Northeast Region, or Nordeste, of Brazil, is an exception. The traditional image of the Northeast being a backwards region dominated by powerful landlords who oppressed a large and poor peasantry has changed considerably, though that, of course, did not come overnight. The change in voting patterns, however, have changed rather drastically in a short period of time.

The Nordeste was the centre of Portuguese Brazil’s colonial economy and was the colony’s wealthiest region – by far- until at least the mid-nineteenth century. Helped by a favourable location and climate, the (coastal) Nordeste’s economy grew, lived and prospered thanks to sugar cane. The need for manual labour on huge sugar plantations especially after the quasi-extermination of the Tupi natives (who were also inadequate for the task) led to the largest slave trade in the western hemisphere. Most African slaves sent to work in the Americas were sent to Brazil, a fact which contributes to the Nordeste’s cultural and racial diversity. As coffee gradually replaced sugar as Brazil’s main export, the Nordeste suffered a quasi-constant decline and degeneration which has, over time, transformed a region which was once the country’s wealthiest into the country’s poorest.

To understand the region’s politics and societal structure, one must first understand the vegetation and precipitation patterns of the region. Basically, the Northeast is divided into three geographical regions (there are actually four, but nobody cares about the fourth). The first is the zona da mata, a wet and green region which forms a long, thin and narrow strip along the Atlantic coastline. This region actually used to be covered by thick rainforest similar to the Amazon, though it was all wiped out by the Portuguese who established sugar cane plantations all along the coast. Sugar cane was and is produced on large estates called engenhos, who were led by wealthy sugar barons who maintained tremendous political influence. After the green zona da mata, another – inland this time – long and thin strip is formed by the agreste, a kind of transitional region between the wet green coastline and the dry dusty interior. Already in the agreste, rainfall is much scarcer and the landscape is much less green. Small agriculture is dominant here. Further inland lies the vast, semi-arid and dustier sertão, the most well-known of the three geographical regions. Prone to long droughts, the sertão is generally unfertile and most agriculture is based around cattle herding. Most herding used to be done on huge estates, the well-known latifundios, led by landowners who dominated a large, poor peasantry. Drought, poverty and lack of income encouraged a massive migration from the dirt poor sertão to major cities of southeastern Brazil.

Slavery and its legacy, as well as a tradition of hegemonic control, unsurprisingly engendered poverty, oppression and an oligarchic paternalist society. Wielding much political and economic power, oligarchs and landowners controlled the economic and political life of their turf, a system which became known in Brazil (and elsewhere) as coronelismo, whereby a local oligarch (a ‘colonel’) controlled a tightly regimented society and delivered their votes en masse to the highest bidder. This power, which came as a result of their wealth and dynastic status, conferred them with considerable power. Their support for the republic in 1889 played a crucial role in its establishment and survival. The system of coronelismo was quasi-universal in Brazil during the café com leite Republic (1889-1930) and continued unfettered in rural areas throughout the country until the 1980s (in some cases, to this day). However, even in the Northeast, the influence and field of action of these bosses was quickly confined to rural areas (where a great majority of the population still lived) while wealthier urban areas became holdouts of organized urban labour or a progressive educated elite. Recife, after all, was a stronghold of the Brazilian Communist Party (PCB) which polled over 15% in Pernambuco in 1945 and 1947 (industrialist backing of the PCB also explains that in parts).

It is important to note some important things about these ‘colonels’, oligarchs and political bosses which literature on Brazilian history refer to. There is sometimes the image of these men all being the same and favouring the same thing. While it is true, certainly, that most of them largely supported a conservative, elitist (or oligarchic) society which paid little attention to the plight of rural workers or to land reform; some bosses were more progressive than others and some were more nationalistic than others. For example, the elite in Pernambuco was much more progressive than the elite in, say, Bahia or Alagoas. Certain members of the elite had nationalist leanings, and some – I dare say – were somewhat progressive! Not all members of the elite were bad, some did some good for their region, notably by encouraging industrialization. Lastly, a fundamental aspect of the elite was that it was still very much flexible – within the confines of the status-quo – and had no exclusive ties to any political party. During the 1945-1964 era, for example, while the conservative UDN was often noted as the party of the old Northeastern elites, and, true enough, the region was consistently the UDN’s strongest, the local PSD and even PTB were very much dominated by the local elites as well.

It might be surprising to some that even with the advent of the secret ballot and apparent democracy in 1945 that the elites could still hold the power they held. As in much of South America, their power was largely hegemonic, indicating some sort of consent from the base. These political bosses still commanded peasants and through veiled threats of losing their jobs, shelter and source of sparse income they could easily control their voting habits.

Getúlio Vargas’s policies were only aimed at organizing urban labour and using the urban working-class as a base of political power similarly to how the old conservatives used Northeastern peasants as their base of political power. Vargas, himself a landowner (though in a society almost a world apart from that of the Northeast) had no interest in organizing the rural workers or disturbing the conservative status-quo in the vast swathes of the Nordeste. Vargas was never a true left-winger, let alone a radical or a communist. Given the nationalist leanings of certain political bosses in the region, he was able to garner the loyalty of the powerful Northeastern elite; though the Getulist (or PTB) organizing was weaker in the Northeast than anywhere else.

The 1946 Constitution, which restricted franchise to literate voters did much to ensure that the elite would not be disturbed by pesky dirt poor peasants. Encouraging literacy, something promoted both by the church and the government (after 1961), was thus seen as a danger by the elite. João Goulart’s 1963 push for ‘basic reforms’, most notable of which were agrarian and political reform, scared the elite more than a bit. Agrarian reform, of course, would break up the large latifundias which still dominated the sertão and break up the oligarch’s main source of power. Political reform, notably encouraging literacy and expanding franchise to illiterate voters, would transform Brazil into a mass democracy (throughout the 45-64 era, only 15-20% of the population actually voted) and would likely break up the power of the elite; because while the poor (and illiterate) peasantry was tightly controlled by the elites, they were also easy targets for nascent radical movements. What Goulart proposed Vargas would never have done, so the elite was shaken to its base by these reform. They had cared little, in the end, about Vargas’ shenanigans with the urban workers and his flirtation with left-wing nationalism because he didn’t dare disturb them. However, Goulart’s reforms endangered the system altogether. The Northeastern political elite was, therefore, entirely behind the 1964 coup which overthrew Goulart. While the support of the middle-classes and the legalist wing of the military for the coup was crucial, the support of the Northeastern elites undoubtedly played a great role. Conveniently, the 1964 coup and the subsequent decades of military dictatorship killed off the nascent radical political movements and stalled the question of land reform (as a political issue) for a long time.

In this context, the experience of Miguel Arraes in Pernambuco is perhaps of importance in highlighting what the political elite feared would happen. A democratic socialist supported by the PCB, Arraes narrowly won the governorship of Pernambuco in the 1962 elections, in an election which proved to be a major defeat for the ruling conservative elite. Certainly funding and monetary support from José Ermírio de Morais, a big (nationalist PTB) industrialist who was elected Senator the same day helped, but Arraes had been able to win by assembling a coalition uniting urban areas with sugar cane workers who were literate enough to vote. Perhaps Arraes would have been the beginning of the end for the elite if the military hadn’t turned back the clock.

The elite supported the military well until the end, at least until doing so stopped being a good strategy and when the military stopped providing the benefits it had done in the past. Master chameleons, the elite, or at least a good part of it, became, or so they claim, convinced democrats right around the time of the mass protests in favour of direct elections in 1984. Those who didn’t change colours then did so the following year, when the ruling majority’s nomination of Paulo Maluf, an unsavoury man, for the presidency alienated much of the regime’s Northeastern power brokers. These people split from the ruling PDS party and formed the Liberal Front Party (PFL) and promptly supported Tancredo Neves in the indirect 1985 election. For them, it couldn’t have been better given that Tancredo Neves (a fundamentally good man) tragically died right before taking office and José Sarney, one of these master chameleons, became President.

While urbanization, the growth of an independent media and higher literacy disturbed the elite by the 1980s, they still held much sway over the region. In the 1982 election, the pro-military and very conservative PDS swept the Northeast, taking nearly 67% of the vote against 43% nationally. That was a bit before the aforementioned change in colours, and in 1986 the old PDS (which was, by then, a far-right party) collapsed and the main fight in the region was between Sarney’s PMDB (which swept the country) and the PFL. The PFL’s ranks included well-known old conservative oligarchs who had supported the military and now played a key role in the new democratic regime; people such as Antônio Carlos Magalhães (ACM, the strongman of Bahia), Marco Maciel and José Agripino Maia. José Sarney himself, as far as I know, never joined the PFL (but at the same time played a major role in the creation of the party), but he also controlled the PFL organization in his home state of Maranhão. His daughter Roseana, for example, was a longtime member of the PFL (until they kicked her out). That being said, the PFL was not the only party for the old elite, and its support was not even throughout the region (in Ceará, the PSDB was dominant, for example). That being said, the PFL is best thought of as the main party of the elite.

Out of the vacuum on the right in 1989 emerged a political outsider and political novice, the young and flashy governor of Alagoas, Fernando Collor. Collor, whose father killed a fellow Senator and whose grandfather had been one of Vargas’ labour ministers, was the perfect representative for the elite. He was conservative, but also detached from the Sarney government in a way which allowed him to be the candidate of change. Collor was the symbol par excellence of the elite: conservative, dynastic and flexible within the system. After all, he had supported Sarney when it paid to do so, but broke all bridges with him when it was bad PR to do so. The darling of the right-wing media and the elite, Collor was the best candidate to take on the left, led by Lula and Leonel Brizola.

Collor defeated Lula with 53% in the runoff and took 55.7% in the Northeast (which was only his third-best region), but lost to Lula in the state of Pernambuco by a narrow margin of 50.9% to 49.1%. Lula had also done well in Bahia (48.3%), Rio Grande do Norte (47.4%) and Paraíba (45%), and his results certainly were a success for the PT which was extremely weak in the region. It did show that some peasants, formerly strictly regimented into political machines, rebelled in the isolation of the voting booth; but the bottom line was that the political bosses still held considerable sway, especially in rural areas. A better understanding of the nature of the election in the region is provided by a look at results in the context of municipalities.

Nordeste, 1989 election (runoff); source: Eleições em Dados

The map to the left reveals a quasi-perfect urban-rural divide, a divide which was seen throughout the country. Collor won the election almost solely on the power of the rural vote, getting obliterated by varying margins in the quasi-entirety of the country’s major urban centres (he may have won São Paulo narrowly). Maceió, AL is an exception to this rule; but Collor’s ownership of the state (and city) likely explains that. Certainly, however, in cities like Salvador, Recife or Fortaleza; Collor did very badly. Aside from the fact that Northeastern cities, which are far wealthier than the rural areas surrounding them, have been holdouts of urban labour or urban progressives against the ultra-conservative countryside; the urban centres were also the only one in 1989 to have had access to unbiased independent media sources which highlighted Collor’s shady side. His status as an old oligarch and his candidacy as one of the old right was certainly not an advantage in any urban centre, especially the most progressive ones like Recife, Salvador, Fortaleza or Rio.

There are various outspots of red in rural areas, likely explainable by a weak political machine in the specific region or an economy less reliant on agriculture and herding. The best example of such areas are in the São Francisco Valley, in towns such as Juazeiro, Petrolina and Petrolândia, where the economy revolves more around hydroelectricity and industry than around herding. But the main thing here is that the rural areas are varying shades of blue, more often that not darker shades of blue. Certainly Collor dominated in the vast majority of the herding-reliant sertão, which was the base of the political bosses of yesterday, more so than the coastal areas which were slightly less conservative.

Collor’s impeachment in 1992 didn’t end the power of the elite, which showed its flexibility in its relations with Fernando Henrique Cardoso (FHC). As a rather leftie sociologist and intellectual, FHC likely had little love for the elite and they barely hid their contempt for him. The Bahian baron of the PFL, Antônio Carlos Magalhães, criticized FHC’s anti-inflation Plano Real right at the outset. However, when the Plano Real proved to be a resounding success and Cardoso showed his interest in taking the top job in 1994, the elite widely acclaimed Cardoso as the candidate of the right and his running-mate, Marco Maciel (PFL-PE) gave from their ranks. Cardoso certainly wasn’t their natural choice, and probably not their preferred choice. But pragmatically, FHC was the Brazilian right’s only hope against Lula, who had been leading in all opinion polling until at least mid-summer 1994.

FHC defeated Lula on the back of the resoundingly successful Plano Collor by the first round, taking 54.3% nationally by the first round and 57.6% in the Northeast. Lula won 30.3% in the region, slightly more than what he got nationally and did well in Pernambuco (37%), Sergipe (36.9%), Bahia (35.2%) and Piauí (32%). Contrarily to Collor in 1989, FHC also performed very well in urban areas. Of the state capitals, all but one of which had been swept by Lula in 1989, FHC won five of them and Lula’s win in Salvador was not as massive as it had been in 1989. Certainly, FHC had a better image in urban areas than Collor did. Largely the same thing happened in 1998, though FHC did worse in the Northeast than nationally (47.7% vs. 53.1%) though not as much to Lula’s benefit (31.6% regionally vs. 31.7%) but to the benefit of Ceará’s favourite son Ciro Gomes who won 16% regionally and 11% nationally (and narrowly won his home-state, which was practically a 3-way tie).

Some will argue that 2002 represents the turning point in the Nordeste’s voting patterns, but I disagree. Certainly, Lula won big in the Northeast and that is not a small fact. He turned the region red, but at the same time he turned the quasi-entirety of the country red, even the wealthy, white conservative states of the south. Lula’s appeal in 2002 was broad, something which comes both from his campaign style and rhetoric, which was unusually moderate and ‘calm’ (gone, obviously, was the angry bearded Lula); and from FHC’s widespread unpopularity after a bad second term. That explains why the regional results in the runoff were all within 3-4% of the national result (61.3%). With 61.5% in the Northeast, it was hard to say then that the region was the key stronghold of the new left. What’s more, the only state which Lula lost was in the Northeast – Alagoas, the last holdout of the conservative elite; and a look at results by municipality reveals that you still had a fair number of random patches of conservative blue in the traditional rural strongholds of the elite.

Results of state-level results in 2002 also provide ample evidence that the realignment had not occurred by 2002. With the exception of Piauí where Wellington Dias (PT) defeated incumbent PFL governor Hugo Napoleão, the parties representing the old order held on in most states. In Pernambuco, Humberto Costa (PT) was badly defeated by right-winger Jarbas Vasconcelos (PMDB), while in Bahia Paulo Souto (PFL) easily defeated Jaques Wagner (PT). The candidates of the right, generally speaking, also performed well in the senatorial election. Thus, while Lula won big in the region even in 2002, an analysis of results both at a micro level and a downballot level provides evidence to claim that 2002 was not a realignment and the presidential results were merely part of the wave which swept Brazil.

Results of the 2010 presidential election by municipality (own work)

The realignment came in 2006, which was a defining election in terms of defining the effects of the first left-wing government in Brazil since the 1960s on the voting behaviours of Brazilians. In 2002, in the first round, Lula had done about as well in the Northeast than in Brazil (slightly worse, in fact). In 2006, he performed 18% better in the Northeast than in the country as a whole. In the runoff, that figure was 16%. Overall, he won 66.8% in the region by the first round and took a crushing 77.1% of the votes in the runoff. Just by those results, as well as his sheer domination – even by the first round – in most of the backwoods of the sertão, it was clear that a major realignment had taken place.

At the state level, unlike in 2002, the forces of the left did well. Of course in Bahia, Jaques Wagner unexpectedly defeated incumbent PFL Governor Paulo Souto by a ten-point margin in the first round. In Pernambuco, with the quasi-unanimous backing of Humberto Costa’s first round voters, Eduardo Campos (PSB, Arraes’ grandson) easily defeated governor José Mendonça Filho (PFL) in the runoff. In Ceará, Cid Gomes (PSB) defeated governor Lúcio Alcântara by a crushing 28-point margin in the first round. Finally, in Maranhão, governor Roseana Sarney (PFL) was narrowly defeated by Jackson Lago (PDT) in the runoff, a victory which, at the time, was considered a major blow to the Sarney clan, dominant in Maranhense politics.

The results of the first round of the 2010 election provided no indication that the 2006 result were an aberration. Quite to the contrary, it arguably showed, to some extent, an amplification of the 2006 realignment. With the notable exception of Rio Grande do Norte (which was caused by local conditions), big names on the old right went down to defeat badly. Tasso Jereissati, a dominant Cearan politician of the PSDB, was badly defeated running for what was thought to be easy re-election to the Senate. Marco Maciel, a key figure of the old PFL and a key symbol of the old conservative oligarchs who had supported the military, was badly defeated. César Borges, though allied with Lula, lost badly in Bahia; where Jaques Wagner won a landslide reelection over his 2002 and 2006 rival Paulo Souto. In Pernambuco, finally, Eduardo Campos (PSB) was reelected with one of the biggest margins in Brazilian political history, over a well-known politico (Jarbas) no less.

An amusing anecdote showing the new force of the left in the region is how even those on the right are attempting to place more emphasis on their relations with Lula than their relations with their party. A notable example is that of Teo Vilela Jr, the incumbent PSDB governor of Alagoas in a tough fight for reelection. Nothing on his website indicates his party, let alone his party’s presidential nominee, but instead had a nice article on how Lula had praised his work. He himself also had lots of nice things to say about Lula, all part of a strategy to confuse voters into thinking that he was Lula’s candidate, when technically he’s a member of the main opposition party to the President.

An analysis of the results of the 2010 elections by municipality shows the utter dominance of the left in the sertão. This is where the electoral realignment occurred and where it was most dramatic. In large swathes of the massive sertão, which covers most of the region’s interior, Dilma won well over 60% of the votes and in parts of Bahia, Pernambuco, Ceará, Piauí and Maranhão her results in the sertão were even well into the 70-80% range with a few municipalities (mostly small) giving her over 90% of the votes. Only eight years ago, a fair share of those had still given Serra a near majority of the votes. An interesting case study in the voting shifts at work in the sertão is the municipality of Caetés, Pernambuco. Caetés is the birthplace of Lula, and is a typical dirt poor village in the sertão. In 1994 and 1998, Lula lost the village by a big margin (FHC had even taken over 60% there in 1998). Dilma won 81% this year (Serra won 15%). In 2002, Lula had won 54% in the first round. Most of the littoral remains left-wing, though by somewhat smaller margins. Conversely, the left’s strength in the region’s urban centres (which are much more well-off than the rural areas) has abetted somewhat since the highs of 1989, a trend observed in a number of major urban centres throughout Brazil (ironically, Lula’s average voter in 1989 was probably wealthier than Collor’s). Marina Silva did very well in the region’s major cities this year, in line with her strong performances in well-educated and averagely well-off urban areas; though Serra did very poorly, though Marina likely took a lot of votes which otherwise would have gone to him. That being said, most urban centres, especially Salvador, remain staunchly left-wing; with the exception of places like Maceió and Aracaju.

How did such a realignment happen?

While the Northeast was demographically perfect ground for the left in any western democracy, like in most of Latin America the tight control of the population by an oligarchic elite kept the region safely in conservative hands for most of the twentieth century. The slow moves towards agrarian democracy and somewhat equitable land distribution as well as the gradual loss of power by the old elites through increased education, awareness, freedom and wealth undoubtedly played a major role in ending the power of the conservative dominant class. The increased literacy rate and educational level in the region are also exemplified by the massive collapse in invalid votes. Illiterates can vote since 1988 (though, unlike literate voters, they are not legally compelled to do so), and the vote is now done electronically (since 1994). Unsurprisingly, there is a definite correlation between illiteracy and a high percentage of blank votes. The percentage of invalid votes in the region has dwindled from roughly 30% in 1994 to slightly above 10% this year.

The PT was, at its foundation, a largely urban party based in urban, organized labour. While it did have a few bases of more rural support (notably in Acre), the party was especially weak in the rural Nordeste where it had no natural base with rural workers. It was only the experience of the left in power which allowed the party to build a solid base in the region, thanks to the social policies of the Lula government which are of great importance to the region. There is a very high positive correlation between votes for the left in a municipality and the number of Bolsa Família beneficiaries in said municipality. What Lula and the left has achieved is, in a way, the eternal and unrealized goal of most of the Latin American left: creating a durable alliance between urban and rural workers. That was what Goulart wanted to achieve in 1964, but he was far ahead of his day in that regard.

Some of this would probably not have happened without the flexibility and opportunism of the elite. The elite might have been conservative and downright horrible in most senses of the term, but they certainly weren’t inflexible or stubborn. Their flexibility within the confines of the established system allowed them to survive, and will allow some of them to survive in the future. Brazilian politicians, of course, are not known for their moral stature or their ideological consistency. Most don’t see an issue with supporting two ideologically opposed governments, and most choose their allies based on the amount of money they can offer them. A fair share of the formerly conservative elite thus realized that their new interests lay with Lula, and so they choose to support Lula even though they were all diehard opponents of his back in 1989. Notable Northeastern oligarchs who took this route include José Sarney (PMDB), who has become one of the President’s biggest allies; Fernando Collor (PTB); Renan Calheiros (PMDB); Roseana Sarney (PMDB), a former pefelista expelled for supporting Lula in 2006; Henrique Eduardo Alves (PMDB) and César Borges (PR). People like Garibaldi Alves Filho (and his father) can also be counted as allies, albeit less reliable than Sarney, of the left. Those who haven’t abandoned their right-wing roots, haven’t done so well. Marco Maciel and Tasso Jereissati are good examples. As corrupt and distasteful as they might be, the old elite still carry a lot of weight around with their family names and they are still able to bring pork back to their regions. They might not be doing so well as in the past, but the contemporary coronels who’ve understood where the advantages lie aren’t going to kick the bucket just yet.

Evolution of votes for the PFL in elections to the Chamber of Deputies since 1986

A telling story of the decline of the conservative oligarchy in the region is that of the PFL. While not representative of the entire ruling elite, which was spread out over a whole slew of parties, the PFL is the most representative of the old elite and is intimately linked to conservative oligarchs who allied with the military in 1964, supported it until it became the bad thing to do and quickly converted themselves into convinced democrats, or so they claim. For any right-wing candidate, the support of the PFL was crucial. FHC understood that in 1994 and 1998, and thus accepted an alliance with them, even though it kind of went against everything he stood for as a progressive intellectual opponent of the military regime. The PFL controlled a near majority of governorships and senate seats in the Northeast, and their support was thus key to the right (read, the PSDB). The PFL wasn’t strong in all states of the Northeast, for example they always lacked a base in Ceará and were weak in states such as Alagoas or Paraíba; but overall, they were the dominant party in the region. They held the plurality of the region’s seats in the Chamber of Deputies between 1990 and 2010, and between 1990 and 2006 they controlled at least two governorships in the region, often more than that.

However, as the direct representatives of the old order, they stood to lose the most from the realignment described above. And indeed it did. The chart above shows the evolution of the votes for the PFL in elections to the Chamber of Deputies between 1986 and 2010. 1986 is perhaps abnormaly high given the two-party nature of Brazilian politics that year, but between 1990 and 2002 the PFL won over 24% of the votes in each election region-wide. In 2006, for the first time, their vote fell below 20% to reach 17%; but even then they still controlled a narrow plurality of the seats in the region and topped the poll in the region. It was to be their last hurrah. In 2010, their vote collapsed below the 10% line to reach a paltry 9.41%. A spectacular decline of nearly 15% in a period of 8 years. That has significantly reduced the party’s caucus in both chambers, and they are no longer the region’s dominant party (the PT and PMDB, like in the rest of Brazil, have taken up that role).

The story of the Nordeste’s drastic electoral evolution in such a short period of times highlights not only the decline and perhaps upcoming fall of the coronels and caciques of yesterday, but also highlights a fundamental realignment in the voting patterns of Brazilians in the wake of the first left-wing government in the country since the 1960s. The story of one of Latin America’s most fascinating regions is also the story of many regions of Brazil and Latin America as a whole.