1. A GFK poll shows Peru President Pedro Pablo Kuczynski has 32% approval and 58% disapproval. His approval is down from 38% one month ago. The numbers mirror the results of an Ipsos poll I wrote about recently. As with the previous poll, a large number of Peruvians cited concerns about security as a key reason they disapproved of the president.

    Keiko Fujimori has 39% approval, the highest level for any politician polled nationally. Her brother Kenji has 33% approval. Along with Congress's poor relationship with the president, the political battle between the two Fujimori siblings is shaping the political narrative in the country. The Congress has a 19% approval and 71% disapproval.

    70% of the country approves of the imprisonment of former President Humala.
  2. Nearly every week in Argentina has seen an article about the polling war in the province of Buenos Aires. I'm fairly certain there have been more polls in Buenos Aires in the past month than there have been in Mexico, Chile, or Colombia all year.

    This week's article has two polls. Both polls show former President Cristina Fernandez de Kirchner running in first place with about a third of the vote. Where they differ is that the poll from Opolit has Esteban Bullrich, the candidate of President Macri's Cambiemos, running in close second with 32% while Sergio Massa has 20%. The poll from Raul Aragon has Massa in second place with 26% and Bullrich in third with 22%.

    The question of who is in second or third place matters because voters who are anti-Kirchner want to know which candidate is best positioned to win against the former president in a single round election.

    Fortunately for data nerds, Argentina's system has the PASO, a set of open primary elections taking place on 13 August that essentially serve as a vanity race among the candidates. The pollsters are currently trying to predict the outcomes of the PASO, and we'll soon be able to determine the accuracy of their data. This won't stop the polling wars from continuing in September, but it will set one benchmark for who has the most credibility.

    To be clear, I view the PASO as a waste of resources. It was designed to be a vanity election that serves to signal voters about viability. It would be better for democracy to just run two rounds (or, boldly, one round with instant-runoff) and make the winning candidate get a majority of the votes. But as someone who likes data to check polling, the PASO system is a waste of resources that directly benefits me, so I can't complain too much.
  3. Two union leaders from the US and Canada have an op-ed in the NYT about NAFTA renegotiation. They are critical of NAFTA's impact so far, but they also raise issues that can be improved in the current renegotiation effort. Most important, instead of demonizing Mexico for having stolen jobs, they advocate for improving working conditions in Mexico as a way of leveling the playing field and improving the autoworker situation across North America. That starts with this critique:
    Yet Mexican workers have not benefited from this bad deal. Mexican wages have stagnated in real terms since Nafta was enacted in 1994. The average autoworker in Mexico makes $3.00 an hour or less, despite healthy industry profits. Labor standards continue to be dismal, since Mexican workers are prevented from exercising their rights and bargaining for better wages and working conditions. The government, employers and the main Mexican autoworkers union, the Confederation of Mexican Workers, frequently collude to maintain a system of “protection contracts” without workers’ consent.
    I wrote about this issue before and view this as one of the areas where NAFTA absolutely can be improved. Mexico's auto manufacturing and auto parts industries are full of examples where the unions serve corrupt and political interests instead of helping the workers organize and collectively bargain. As these authors point out, that corrupt system hurts workers in Mexico and harms the ability of US and Canadian workers to compete.

    While the US has listed labor rights as part of its priorities in the renegotiation, like the rest of the document, that section is composed of vague general statements and not specific actions that can be taken. I doubt the current administration is eager to push a pro-union agenda, but Congress should take this up as a key area where progress can be made.
  4. 1. Beginning next week I will publish a free weekly Latin America newsletter that will provide analysis from my company Hxagon as well as a roundup of all the blog posts I publish here at Bloggings by Boz. I’ve had a surprising number of readers ask for a regular email service over the 13 years I’ve been writing this blog and I’m finally delivering. You can sign up here.

    2. POLL NUMBERS!!! Hxagon also has a paid weekly newsletter that provides updates on the poll numbers and political situation in Mexico. The cost is $6 per month or $60 per year. If your organization has any operations, investments or interest in Mexico, this is a service that can help you keep an eye on next year’s elections and beyond. The code “Encuesta6” will give readers a free month of the monthly subscription (so you can receive it for a few weeks before you pay) or 10% off the annual subscription. You can learn more about the Mexico poll number report and sign up here.

    Links to both services are at the top of the blog.

    Thanks to everyone who reads this blog and everyone who has supported my new startup.
  5. Brian Winter writes that Jair Bolsonaro appears to be the only Brazilian politician benefiting from the current environment where most voters are tired of politics and the corruption scandals. Though polling second or third in most polls, Bolsonaro is the only candidate drawing large and enthusiastic crowds. Certain wealthy demographics support him and he has the most Facebook followers of the candidates.

    Meanwhile, Bloomberg reports on a Mexican phone app is measuring candidate support and found that nearly 50% of its users support AMLO. The users are younger and more middle class than the average population, but it could also signal support that is missing in polls showing AMLO closer to the 25-30% mark.

    It raises an interesting question of how to measure support that doesn't show up in polls. Candidates often point to supporter enthusiasm, crowd size and social media followers as a proxy for voters that they say polls aren't picking up. In most cases, the polls are correct and the enthusiasm explanation is wrong, but sometimes the crowd size or social media influence does signal support that polls are slow to pick up on. (Also, confirmation/survivor biases mean we tend to remember and perhaps exaggerate the examples where the media narrative "beat" the polls, but not the examples where crowd sizes were just unimportant noise).

    Still, given the growing concerns about the unreliability of polls, driven largely by increasing non-response rates, we should be looking for new data points to capture these trends. From an election modeling perspective, those data points must be measurable. I'm interested in finding ways to know if voters for Bolsonaro or Lopez Obrador or any other candidates are being undercounted in the polls, particularly in countries where polling data is not updated on a daily basis the way it often is in the US during election campaigns.
  6. Yesterday's NYT added to its already extensive coverage HNA, one of several Chinese conglomerates that has been on a buying spree around the world. The key point of the article was to highlight that HNA's subsidiaries and partners are often controlled by family and friends of the co-founders of the organization. The relationship map among the various companies means that the deals are not strictly business. Businesses and investors who have formed financial relationships with HNA and its partners have never had the full picture of its opaque ownership structures and the risks that may exist within.

    This adds to other recent concerns that these conglomerates have taken out too much debt, are not economically sustainable, and may pose a systemic risk to China's economy. At least one major US bank is refusing to do business with HNA due to its murky structure and the debt risks.

    This becomes important for Latin America when considering HNA's investments in the region. They own pieces of infrastructure including key stakes in airlines and airports in Brazil. The risks of the ownership structures and the potential debt timebomb within the conglomerate could impact the logistics of moving people and products throughout the region. The stakes are large enough to impact parts of Brazil's already hurting economy.

    Latin America has eagerly accepted Chinese investment, but Latin American regulators gave very little consideration to the risks that these companies may bring. The hemisphere's economies should seek greater trade and investment with China, but they should also demand a greater level of transparency from the investors and commit to oversight of them.
  7. Mac Margolis covers the split between Ecuador President Lenin Moreno and former President Correa. Moreno has reached out to his political opponents and is moving towards the center, leaving Correa to publicly criticize his former vice president and attempt to defend his legacy.

    Meanwhile, Ecuador announced that it will not hold to its promised cuts in oil production, becoming the first country to publicly split with OPEC. While Ecuador is not the largest actor in OPEC, any split in the organization creates the potential that others will defect from their promises. That would leave OPEC without its main lever of power.

    As Margolis writes, these stories are largely due to the economic reality of Ecuador and the fact the country needs more revenue and/or less spending to avoid economic crisis. Moreno understands that protecting Ecuador's economic gains will be easier if he backs off Correa's ideological battles and finds some middle ground with his opponents that lets him implement moderate reforms. Successful reforms now will give him economic breathing space and political capital for a successful term.
  8. An Ipsos poll shows Peru President Pedro Pablo Kuczynski has 34% approval and 58% disapproval. His approval rating is down five points in the past month.

    There is no single reason that Peru's president has seen his approval ratings decline. There is general discontent with his inability to get things done, which can be blamed on the fact that the opposition controls Congress. Last month the Congress forced the Economy Minister to resign. Additionally, the poll showed about 35% of citizens said high levels of insecurity and violence were their reasons for disapproving of the president.

    PPK's low approval with less than a year in office doesn't bode well for the rest of his term. His predecessors also spent most of their terms with very low approval ratings and were never able to rebound back to above 50%. PPK also suffers from the bad luck of governing during a tough opinion moment for presidents throughout most of South America.
  9. An El Universal poll says 55% of citizens view insecurity as the key concern in their state and 43% believe it is the key issue nationwide.

    Similarly, a GEA poll last month showed 43% of citizens thought the security situation was worse in the country during 2017 compared to only 13% who thought it had improved (the rest said it remained the same). Only 29% approve of the federal government’s security efforts contrasted with 61% who disapprove, a significant drop over the past two years for the Peña Nieto administration.

    The rising violence is a problem across the country, but the specific narrative of why the violence is rising varies state by state and region by region. For example….

    The WSJ covers rising violence in Sinaloa, where homicide rates have doubled this year.

    Alejandro Hope writes about violence in Hidalgo and Veracruz.

    Rolling Stone profiles the rise of El Mencho and the violence of the CJNG in western Mexico including the tourist city of Puerto Vallarta in Jalisco.

    Business Insider covered the rising violence on the northeast border including Tamaulipas and Nuevo Leon.

    The LA Times reported on a massacre in Chihuahua in which 26 people died.

    The NYT covered the 28 dead in a prison riot in Acapulco, Guerrero, earlier this month.

    It’s not clear yet how the violence will impact the election, but if it’s the key issue on voters’ minds and it is rising across the country, all of the major candidates are going to be forced to explain how their policies will improve the situation.

    In the meantime, starting tomorrow, I'll be providing weekly poll number analysis at Hxagon for those who want regular updates.
  10. 1. On Saturday I wrote that I expected between three and five million votes and that anything above five million should be considered a significant victory for the MUD and a major loss for Maduro. After all, the referendum was non-binding, the organization took place in just two weeks and the logistics of voting centers meant people would have to stand in long lines. So the 7.2 million Venezuelan citizens that voted in yesterday's referendum went well beyond expectations. This is a major victory against the government and has the potential to be game changing if the MUD can manage their next moves correctly.

    2. Yesterday's vote was peaceful until it wasn't. The majority of Venezuelan voters who reject the current government certainly deserve credit for their attempt at a peaceful and non-violent revolution through voting. The security forces and pro-government paramilitaries/colectivos were largely absent from polling places as the population peacefully stood in line waiting to vote for most of the day. Later in the afternoon, a collectivo attack on a voting location left at least one dead and several injured, marring what was otherwise a very positive day.

    3. This referendum confirms polls showing the public has rejected the government and its constitutional assembly plans. The government's challenge will be getting more than 7.2 million voters to the assembly election in late July and I doubt they can do it. If a non-binding referendum rejecting an election receives more votes than the actual election, I think it's fair to say the legitimacy of that election is gone.

    4. Some within the MUD leadership want to take this referendum's results further. They claim the positive votes on questions two and three give them the legitimacy to form a new government via the elected National Assembly, one that will serve in parallel with the Maduro government and compete for international recognition. It would be a bold path that could force key actors inside and outside Venezuela to take sides instead of remaining on the sidelines.

    5. Specifically interesting is question two, which asked the military to defend the current constitution and the elected National Assembly. With over seven million voters volunteering their opinion that they want this, it sets up a renewed clash among the leadership of the Armed Forces. Many within the military are already unhappy about the rewriting of the constitution. There is now an arguable legitimacy that the country's population wants the military to side with the National Assembly against the undemocratic actions of the executive. Keeping that institutional clash as peaceful as possible will be a critical mission for the international community and the non-violent majority in the country.