Quando o crescimento não chega

Este blogue tem andado parado, em parte por excesso de trabalho e em parte porque alguns segmentos do que fazia aqui foram deslocalizados para outras paragens (análise no Radar Económico e clipping de investigação na minha página do Scoop.it). Mas há coisas que merecem estar em mais do que uma plataforma. Como o discurso do Ben Bernanke no último Fórum de banqueiros centrais em Sintra: When growth is not enough

Regarding the United States, let me start with the positive. The nation’s cyclical recovery is entering its ninth year this month and appears to have room to run. Although the Great Recession was exceptionally deep and the recovery was slower than we would have liked, real GDP is now up about 12.5 percent from its pre-crisis peak, and real disposable income is up more than 13 percent (…)

And yet, despite the sustained cyclical upswing and the country’s fundamental strengths, Americans seem exceptionally dissatisfied with the economy, and indeed have been for some time. For example, those who tell pollsters that the country is “on the wrong track” consistently outnumber those who believe that America is moving “in the right direction” by about two to one. And, of course, last November Americans elected president a candidate with a dystopian view of the economy, who claimed that the “true” U.S. unemployment rate was 42 percent (…)

So why, despite the undoubted positives, are Americans so dissatisfied? The reasons are complex and not entirely economic. Without trying to be comprehensive, I’ll highlight here four worrying trends that help to explain the sour mood.

First, stagnant earnings for the median worker. Since 1979, real output per capita in the United States has expanded by a cumulative 80 percent, and yet during that time, median weekly earnings of full-time workers have grown by only about 7 percent in real terms. Moreover, what gains have occurred are attributable to higher wages and working hours for women. For male workers, real median weekly earnings have actually declined since 1979. In short, despite economic growth, the middle class is struggling to maintain its standard of living.

Second, declining economic and social mobility. One of the pillars of America’s selfimage is the idea of the American Dream, that anyone can rise to the top based on determination and hard work. However, upward economic mobility in the United States appears to have declined notably over the postwar period. For example, in a paper aptly entitled “The Fading American Dream,” Raj Chetty and coauthors studied one metric of upward mobility, the probability that a child would grow up to earn more than his or her parents. Using Census data, they found that 90 percent of Americans born in the 1940s would go on to earn more as adults than their parents did, but that only about 50 percent of those born in the 1980s would do so. Other research finds that the United States now has one of the lowest rates of intergenerational mobility among advanced economies, measured for example by the correlation between the earnings of parents and their children. For a supposedly classless society, the U.S. is doing a good job of rigidifying its class structure through means that include residential and educational segregation, social networking, and assortative mating.

The third adverse trend is the increasing social dysfunction associated with economically distressed areas and demographic groups. For example, other former Princeton colleagues of mine, Anne Case and Angus Deaton, have done important work on morbidity and mortality among white working-class Americans (more precisely, people with only a high school degree). They find that midlife mortality rates among white working-class Americans have sharply worsened, relative to other U.S. demographic groups and working-class Europeans. Case and Deaton refer to the excess mortality among the white working class as “deaths of despair,” because of the associated declines in indicators of economic and social well-being and the important role played by factors like opioid addiction, alcoholism, and suicide.  Indeed, in 2015, more Americans died of drug overdoses — about 60 percent of which involved opioids — than died from auto accidents and firearms-related accidents and crimes combined (…)

The fourth and final factor I’ll highlight, closely tied to the others, is political alienation and distrust of institutions, both public and private. In particular, Americans generally have little confidence in the ability of government, especially the federal government, to fairly represent their interests, let alone solve their problems. In a recent poll, only 20 percent of Americans said they trusted the government in Washington to do what is right “just about always” or “most of the time” (…).

I’m hardly the first to observe that Trump’s election sends an important message, which I’ve summarized this evening as: sometimes, growth is not enough. Healthy aggregate figures can disguise unhealthy underlying trends. Indeed, the dynamism of growing economies can involve the destruction of human and social capital as well as the creation of new markets, products, and processes. Unaided, well-functioning markets can of course play a crucial role in facilitating economic adjustment and redeploying resources, but in a world of imperfect capital markets and public goods problems there is no guarantee that investment in skills acquisition, immigration, or regional redevelopment will be optimal or equitable. Tax and transfer policies can help support those who are displaced, but the limits on such policies include not only traditional concerns like the disincentive effects of income-based transfers but also conflicts with social norms. Notably, people can accept temporary help but transfers that look like “handouts” are often viewed with extreme suspicion or resentment. Some active interventions thus seem a necessary part of a responsive policy mix.

Providing effective help to people and communities that have been displaced by economic change is essential, but, on the other hand, we should not understate how difficult it will be. Addressing problems like the declining prime-age participation rate or the opioid epidemic will require the careful and persistent application of evidence-based policies which populist politicians, with their impatience and distrust of experts, may have little ability to carry through. Moreover, to be both effective and politically legitimate, such policies need to involve considerable local input and cooperation across different levels of government as well as cooperation of the public and private sectors. The credibility of economists has been damaged by our insufficient attention, over the years, to the problems of economic adjustment and by our proclivity toward top-down, rather than bottom-up, policies. Nevertheless, as a profession we have expertise that can help make the policy response more effective, and I think we have a responsibility to contribute wherever we can.

Há ainda outra parte muito interessante, acerca de um tema a que tenho dado destaque neste blogue: a sensibilidade dos efeitos finais de reformas estruturais às circunstâncias macroeconómicas em que aquelas são implementadas. Esta questão devia ser mais do que apenas uma nota de rodapé no debate político europeu, onde há um combate silencioso em torno dos efeitos destas reformas e do melhor timing para as implementar. Que seja um americano a ter a coragem de discutir isto abertamente não deixa de ser uma curiosa ironia:

A small literature has argued that structural reforms can be counterproductive when interest rates are at the zero lower bound, because of disinflationary effects. I tend to agree that those ZLB effects are probably quantitatively modest. However, whether rates are at zero or not, it seems quite likely that policies that have the effect of releasing redundant labor resources could have adverse short-run effects if insufficient aggregate demand exists to re-employ those resources in a reasonable time. It’s consequently important for the content and sequencing of reforms to take into account the macroeconomic situation, as has been pointed out by the International Monetary Fund and others. Likewise, reforms can complement, but should not be viewed as a substitute for, appropriate macroeconomic policies. In particular, labor market reforms should not by themselves be expected to solve national competitiveness problems, at least not in the short term. Also needed are appropriate macroeconomic policies, especially fiscal policies, to help ensure adequate demand and remedy the underlying source of trade imbalances.

Advertisement

Teoria ou dados?

Já falei ali em baixo sobre o soul-searching a que muitos macroeconomistas se têm dedicado nos últimos anos, depois de constatarem fertilidade limitada do uso de DSGE’s como programa de investigação. Mas há debates menos herméticos do que esse a correr em paralelo. Como este, acerca da interacção entre teoria e investigação empírica: How should theory and evidence relate to each other? (Noah Smith)

Without a structural model, empirical results are only locally valid. And you don’t really know how local “local” is. If you find that raising the minimum wage from $10 to $12 doesn’t reduce employment much in Seattle, what does that really tell you about what would happen if you raised it from $10 to $15 in Baltimore?

That’s a good reason to want a good structural model. With a good structural model, you can predict the effects of policies far away from the current state of the world.

In lots of sciences, it seems like that’s exactly how structural models get used. If you want to predict how the climate will respond to an increase in CO2, you use a structural, microfounded climate model based on physics, not a simple linear model based on some quasi-experiment like a volcanic eruption. If you want to predict how fish populations will respond to an increase in pollutants, you use a structural, microfounded model based on ecology, biology, and chemistry, not a simple linear model based on some quasi-experiment like a past pollution episode.

That doesn’t mean you don’t do the quasi-experimental studies, of course. You do them in order to check to make sure your structural models are good. If the structural climate model gets a volcanic eruption wrong, you know you have to go back and reexamine the model. If the structural ecological model gets a pollution episode wrong, you know you have to rethink the model’s assumptions. And so on.

(…)

Economics could, in principle, do the exact same thing. Suppose you want to predict the effects of labor policies like minimum wages, liberalization of migration, overtime rules, etc. You could make structural models, with things like search, general equilibrium, on-the-job learning, job ladders, consumption-leisure complementarities, wage bargaining, or whatever you like. Then you could check to make sure that the models agreed with the results of quasi-experimental studies – in other words, that they correctly predicted the results of minimum wage hikes, new overtime rules, or surges of immigration. Those structural models that failed to get the natural experiments wrong would be considered unfit for use, while those that got the natural experiments right would stay on the list of usable models. As time goes on, more and more natural experiments will shrink the set of usable models, while methodological innovations enlarges the set.

But in practice, I think what often happens in econ is more like the following:

1. Some papers make structural models, observe that these models can fit (or sort-of fit) a couple of stylized facts, and call it a day. Economists who like these theories (based on intuition, plausibility, or the fact that their dissertation adviser made the model) then use them for policy predictions forever after, without ever checking them rigorously against empirical evidence.

2. Other papers do purely empirical work, using simple linear models. Economists then use these linear models to make policy predictions (“Minimum wages don’t have significant disemployment effects”).

3. A third group of papers do empirical work, observe the results, and then make one structural model per paper to “explain” the empirical result they just found. These models are generally never used or seen again.

A lot of young, smart economists trying to make it in the academic world these days seem to write papers that fall into Group 3. This seems true in macro, at least, as Ricardo Reis shows in a recent essay. Reis worries that many of the theory sections that young smart economists are tacking on to the end of fundamentally empirical papers are actually pointless

(…)

It’s easy to see this pro-forma model-making as a sort of conformity signaling – young, empirically-minded economists going the extra mile to prove that they don’t think the work of the older “theory generation” (who are now their advisers, reviewers, editors and senior colleagues) was for naught.

But what is the result of all this pro-forma model-making? To some degree it’s just a waste of time and effort, generating models that will never actually be used for anything. It might also contribute to the “chameleon” problem, by giving policy advisers an effectively infinite set of models to pick and choose from.

And most worryingly, it might block smart young empirically-minded economists from using structural models the way other scientists do – i.e., from trying to make models with consistently good out-of-sample predictive power. If model-making becomes a pro-forma exercise you do at the end of your empirical paper, models eventually become a joke. Ironically, old folks’ insistence on constant use of theory could end up devaluing it.

(…)

In other words, econ seems too focused on “theory vs. evidence” instead of using the two in conjunction. And when they do get used in conjunction, it’s often in a tacked-on, pro-forma sort of way, without a real meaningful interplay between the two. Of course, this is just my own limited experience, and there are whole fields – industrial organization, environmental economics, trade – that I have relatively limited contact with. So I could be over-generalizing. Nevertheless, I see very few economists explicitly calling for the kind of “combined approach” to modeling that exists in other sciences – i.e., using evidence to continuously restrict the set of usable models.

O que está mal com a macroeconomia?

Há mais ou menos uma década, Olivier Blanchard escreveu um paper acerca das longas  e violentas discussões entre macroeconomistas nos anos 70 e 80. Blanchard começava por lembrar algumas das controvérsias desse tempo – desde a contenda entre keynesianos e monetaristas, até à discussão em torno das expectativas racionais e microfundações -, para concluir que, pela altura em que escrevia, as grandes batalhas tinham acabado numa trégua perpétua. Havia então um largo consenso em torno da metodologia a aplicar (as “regras do jogo”) e dos factos a que qualquer teoria tinha de se conformar, que justificavam a conclusão do autor: «The state of macro is good».

As palavras de Blanchard podem ter sido precipitadas, porque pouco depois as discussões voltaram à baila. No New York Times, Paul Krugman escreveu o famoso How did economists get it so wrong?, seguido de uma célebre série de posts acerca da dark age of macroeconomics. John Conchrane respondeu em How did Krugman get it so wrong?, houve alguma “troca de correspondência” e a coisa acabou por azedar.

O curioso disto tudo é que aquilo que começou por ser uma discussão técnica sobre uma questão muito específica da macroeconomia – a eficácia da política orçamental para controlar o ciclo económico – rapidamente se transformou num debate mais vasto sobre a natureza do conhecimento macroeconómico. Ou, como costuma dizer um amigo meu…

Robôs, produtividade e pontas soltas

Num post anterior acerca dos robôs-que-nos-roubam-empregos notei os sinais contraditórios que recebemos de fontes diferentes. Por um lado, os media (e a experiência pessoal, convenhamos) sugerem que vivemos numa época de inovação tecnológica extraordinária. Por outro lado, as estatísticas agregadas mostram que a produtividade está pelas ruas da amargura.

Será que uma impressão está correcta e a outra está errada? Ou há alguma coisa a escapar-nos, e a contradição é mais aparente do que real? Eu diria que há pelo menos três explicações possíveis.

A explicação mais trivial é que há um delay considerável entre o momento em que as inovações são descobertas e o momento em que são incorporadas nos processos produtivos. Há inúmeros exemplos retirados da Revolução Industrial, mas o meu favorito é a afirmação de Robert Solow, de que “podemos encontrar computadores em todo o lado, excepto nas estatísticas da produtividade”. Poucos anos após pôr meio mundo a discutir o verdadeiro impacto das tecnologias de informação (1987), o alerta revelou-se extemporâneo. A produtividade disparou nos anos 90 e os estudos subsequentes mostraram que isto se devia em parte… às tecnologias de informação.

Se o passado serve para iluminar o futuro, então talvez os robôs sejam mesmo para levar a sério. Talvez seja só uma questão de tempo até que os protótipos passem das capas da Wired, onde fazem manchetes mas (ainda) não produzem, para os lares das famílias e linhas de montagem das empresas. Se for este o caso, então talvez o desemprego tecnológico – transitório, mas real – seja uma possibilidade séria nos próximos 10 ou 20 anos.

Uma segunda explicação é que podemos estar apenas a medir mal o crescimento do PIB – e, consequentemente, temos uma ideia incorrecta da produtividade.

Robôs a roubar empregos

Este é um daqueles post guardados como draft há algum tempo. A ideia era publicar alguma coisa durante a Web Summit, mas acabei por não ter tempo. Felizmente, o Observador fez-me o favor de manter o tema vivo:

Os robôs vão ajudar-nos a mudar o mundo. E vão roubar os nossos empregos: O Fórum Económico Mundial prevê que, até 2020, desapareçam cinco milhões de empregos nos quinze países mais desenvolvidos do mundo por causa da evolução da robótica e da inteligência artificial. Segundo o estudo, divulgado no início deste ano na Conferência de Davos, os setores da saúde, energético e financeiro serão os mais afetados, mas também haverá perdas de trabalho consideráveis na construção, na extração de recursos e no setor das artes e do entretenimento.

Vamos supor que há de facto uma legião de robôs capazes de fazer tão bem ou melhor o trabalho que hoje é feito por seres humanos. Este é um grande ‘se’, como veremos daqui a pouco. Mas mesmo assumindo a premissa como verdadeira não é óbvio em que é que isto difere dos processos de mecanização e automatização que estão em curso há… enfim, há vários séculos. Trocar mão-de-obra por maquinaria é o que tem acontecido nas economias desenvolvidas pelo menos desde a Revolução Industrial.

Confusões sobre a Estagnação Secular

Nos últimos tempos tem-se falado cada vez mais da Estagnação Secular, um conceito cunhado há quase 80 anos e ressuscitado em 2013 por Larry Summers. Nas suas linhas gerais, a ideia anuncia um futuro distópico para as economias desenvolvidas: pouco (ou nenhum) crescimento, níveis de vida estagnados e crises económicas recorrentes.

Summers argumenta que este é, ou pode provavelmente ser, o futuro da maior parte dos países ricos. Em parte porque é mais ou menos isto que vemos quando olhamos à volta – e em parte, suspeito, porque o nome da coisa se presta bem a manobras de marketing – a ideia cravou os dentes no debate público e agora aparece recorrentemente na comunicação social. Mas a forma como o tema é abordado, quer na sua formulação, quer nas suas implicações, deixa muito a desejar.

Em particular, tornou-se habitual dizer que a Estagnação Secular é uma teoria acerca do ‘fim do crescimento’, um facto da nossa vida económica ao qual temos de nos resignar. Na verdade, é precisamente o contrário.

A Estagnação Secular explica por que é que o amadurecimento das economias – tomado um facto exógeno – pode conduzir a falhas recorrentes e persistentes na procura global. De acordo com a teoria, os mecanismos de mercado que durante mais de dois séculos foram suficientes para estabelecer o pleno emprego podem tornar-se cada vez mais ineficazes, exigindo o apoio de outras forças para tapar essa lacuna. E isso tem um remédio.

Mas sobre esta questão o melhor que posso fazer é reencaminhar para o excelente A tale of two stagnations, de Noah Smith.  Vão lá ler tudo, porque aqui só incluí alguns trechos.

The term “secular stagnation” has become a catch-all description for long-term economic pessimism. But it’s gotten confused with a very different idea — the technological stagnation hypothesis, proposed by economist Robert Gordon (and by Bloomberg View’s Tyler Cowen). These are two very different ideas. Both would lead to slow growth in the long term, but they imply different causes and different remedies.

Summers’ secular stagnation is all about aggregate demand. Normally, economists think of demand as something that falls temporarily in a recession and then bounces back. But the failure of many economies to return to their previous trends after big slowdowns has made some economists worry if demand shortfalls could be very persistent.

Demand gaps usually emerge when everyone tries to save money at the same time. This could happen because people become more pessimistic about the future, for example, or because they suddenly decide they need more liquid assets. But when everyone tries to hold onto cash, they don’t spend, and so companies don’t produce things. Companies that don’t produce things lay off workers, and pretty soon there’s a recession.

Usually this process ends naturally. Eventually people need to replace their old cars and fix up their houses, or their temporary bout of pessimism ends, or some other force acts to restore demand. But under certain conditions, in some models, it’s possible for an economy to trap itself, so that low demand and slow growth become a self-reinforcing, self-perpetuating cycle.

(…)

Technological stagnation is a different beast. According to Gordon and others, humanity has simply picked most of the low-hanging fruit of science and technology. Airplanes and cars travel no faster today than they did 50 years ago. Electricity, air conditioning and household appliances have made our homes about as pleasant as they’re likely to get, and so on. That doesn’t mean advances stop, but it means that each one is less game-changing than the last.

A key piece of the tech stagnation hypothesis is that production of the things we want isn’t going to get much cheaper. Gordon points to slowing productivity as evidence that our economy is getting worse at finding new ways to do more with less. This trend is worldwide, which makes sense, since a decline in science and technology should be global in nature.

So technological stagnation is all about supply, while secular stagnation is about demand. The two are related — slower productivity growth tends to reduce interest rates, putting the economy closer to the zero lower bound that drives demand shortages. But the two types of stagnation are very different things, requiring very different policy responses.

If we’re in secular stagnation, the economy is wasting its potential. Workers are staying home — not counted as officially unemployed, but out of the labor force completely — playing video games while offices sit empty and unused. In that case, we need something like fiscal stimulus to raise demand and lift us back to full employment.

But if we’re in technological stagnation, there’s not much we can do. Yes, there are some things government can do to boost innovation at the margin, like reforming patent laws, lifting onerous regulations, and investing in research and development. But in the long term, the forces of progress are difficult to predict and control. If we’ve already exploited the biggest innovations, we need to reconcile ourselves to living lives not much better than those of our parents. That would be a disappointing outcome, but it might be the best we can do.

Como combater a próxima crise

As recessões são fenómenos cíclicos: nas economias desenvolvidas costumamos ver uma em cada década. E, talvez porque a última começou precisamente há nove anos, o FMI publicou uma espécie de manual com as principais instruções para os policymakers: Macroeconomic Management When Policy Space Is Constrained: A Comprehensive, Consistent, and  Coordinated Approach to Economic Policy.

Ao todo são 43 páginas de análise, propostas e simulações. A novidade não está tanto nas principais recomendações, que o Fundo já tem publicado de forma dispersa aqui ou ali. O que é novo é o facto de aquilo que era apenas ‘investigação académica’ feita pelos geeks do Departamento de Investigação ganhar agora forma de doutrina, ao ser publicado numa Staff Discussion Note (e assinado por três pesos pesados do Fundo).

E que novidades são estas? Correndo o risco de simplificar em demasia, parece-me que são cinco:

Aprender com os erros

O que é que é preciso acontecer para convencer alguém a mudar de opinião? Este foi o tema de um Seminário acerca das narrativas da crise, em Coimbra Lisboa. E é uma óptima discussão. Nos últimos anos não faltaram previsões para a evolução da economia global e da Zona Euro. Salvo honrosas excepções, saíram quase todas ao lado – e, nalguns casos, flagrantemente ao lado.

Quando as expectativas são frustradas de uma forma tão gritante, o mínimo que se exige a quem trabalha no ramo é que seja capaz de olhar para trás e, à luz dos acontecimentos recentes, reavaliar as ideias em que acreditava e mudar os modelos a que recorre. Se depois de sete anos continuam a acreditar na mesma coisa, das duas uma: ou acertaram em tudo, o que é improvável, ou não aprenderam nada, o que é deplorável. No meu caso, isto foi o que eu aprendi:

A Estagnação Secular está aí?

Por que estão as taxas de juro tão baixas? Por que é que alguns países europeus estão na prática a receber dinheiro sempre que pedem um empréstimo nos mercados? Por que é que a remuneração dos depósitos mal chega para cobrir a inflação?

Se lhe perguntarem isto, a sua resposta acabará provavelmente por fazer referência ao BCE. É possível que invoque a política monetária dos últimos anos, o Quantitative Easing, o facto de a taxa de remuneração de depósito estar em níveis negativos ou alguma acção deste género. Mas o argumento, com toda a certeza, girará em torno de uma determinada opção tomada pela autoridade monetária.

Num certo sentido, isto é verdade: o BCE fixa as taxas de juro, e essas taxas de juro são transmitidas ao resto da economia. Por outro lado – e este é um ponto crucial a que quase ninguém costuma fazer referência – a moderna teoria económica assenta na ideia de que a política monetária tem efeitos exclusivamente no curto prazo. O Banco Central pode aumentar ou reduzir a quantidade de dinheiro em circulação e, assim, influenciar as taxas de juro. Mas assim que os preços se ajustam ao novo volume de massa monetária, os efeitos reais desaparecem.

Como ensinar economia

O ensino de economia é um tema quente desde a célebre debandada de alunos de uma aula do célebre Gregory Mankiw. A questão levanta tantos problemas – mais científicas do que propriamente pedagógicas, parece-me – que qualquer discussão acaba sempre por ser longa e demorada. Para quem tem tem tempo e paciência, aqui fica o contributo de Karl Whelan (o melhor que li até agora): Thoughts on teaching economics after the crash.