Econ bloggers have really missed the point about Landsburg’s free trade screed. The estimable Dani notwithstanding, the issue isn’t ultimately ethics or even procedural fairness. The problem is that doctrinaire economists understand less about trade than the average person with no academic training in the subject.
Ordinary people in many parts of the world, and not just in the US, worry about trade because they are afraid that jobs lost to imports will not be counterbalanced by jobs gained through exports. They worry that there will be fewer economic opportunities for them and their children. They worry that their wages or working conditions will be pushed downward through competition with even more vulnerable, desperate workers in other countries. They are right to worry about these things. Such miseries are not destined to happen, but they cannot be ruled out either.
Except in standard economic models which begin with the assumption that increases in imports automatically call forth equally valued increases in exports. If trade balances on the margin we live in the happy world of comparative advantage, and it is indeed true, as Landsburg says, that “when American jobs are outsourced, Americans as a group are net winners.” But the assumption that trade balances at the margin is simply a modeling convenience, something that enables Landsburg to regale his students with blackboards full of elegant diagrams and equations. It is not grounded in real experience, and especially not the experience of the US economy since the 1970s.
You have to be very well trained in economics and have high-level skills to make such a brain-dead assumption and not even know you’ve made it. Then you don’t have to give serious consideration to counterarguments because, hey, why pay any attention to the fallacies of economic illiterates and mathphobes?
But let’s get specific. Here’s how Landsburg illustrates his claim that international trade makes us better off: “I doubt there’s a human being on earth who hasn’t benefited from the opportunity to trade freely with his neighbors. Imagine what your life would be like if you had to grow your own food, make your own clothes and rely on your grandmother’s home remedies for health care.”
Notice a problem here? Landsburg assumes that there is no difference between trade within an economy and international trade. (To be more precise, the only difference is that governments interfere more often with trade across national borders.) Worse, he accuses anyone who recognizes the difference of woolly thinking, based of course on his assumption that there is no difference.
This is why the French students complained about autistic economics.
For what it’s worth, my view is that we as a society ought to provide opportunities for as many of us as possible to have a satisfying livelihood. If a community is down and out we should step in and do what we can whether or not trade played a role in creating the problem. We should create rules for international trade that minimize downward pressure on wage, environmental and social standards and that limit dangerous imbalances. These ideas are fairly widespread among the general population, and if economists think really creatively they might just be able to rise to the same level.
Ay, no.
First, you’re right, the assumption that trade is balanced is a modeling convenience. But it is just that. A convenience. I can relax the assumption that trade balances in each period, bring in some dynamics, perhaps impose some kind of transversality condition, re do the whole thing and get the same answer.
Second, that third paragraph doesn’t have much to do with the first (the second and fourth just being some empty assertions; my view is that we as a society ought to provide opportunities for as many of us as possible to have a satisfying livelihood. – who disagrees with that? Certainly not free trade folks). Third paragraph is about whether there is a difference between trade within an economy and international trade.. But there really isn’t except for the fact that you’re talking about trade between two people named Bob, or two people named Bob and Zhou. Yes, yes, yes, there are differences between US and China in terms of things like environmental standards and labor regulations. But there are differences between environmental standards between Montana and California and the state set minimum wages of Nebraska and New York. While institutional details matter and should be included in the analysis, the same model DOES apply, generally, to whether you’re talking about trade between countries or states. The difference being of course that the US Constitution prohibits the levying of tariffs and other protectionist measures (tariffs being a very small part of the story anyway) between states.
So if you want to bring specific institutional detail in the analysis, by all means do so. But this details should be specified out whatever kind of trade one is talking about. In that sense Landsburg is right.
When your livelihood depends upon producing a continual stream of arguments as to why greed is good you leave rigor behind.
The idea is not to produce an unimpeachable argument, but just to keep up the barrage of misinformation so that the real facts get drowned out. It helps that those providing the funding for the libertarian pundits are among the wealthiest people in the world.
There are two sides every story, they say. Some say the earth is flat, while others say it’s round. Let’s discusss.
Why is it that small farmers, workers, and the average person in developing states fail to see the benefits of “free” trade, while the highly paid pundits find it so obvious? Follow the money.
YNS,
Both links are PDFs, not just the second, and, having spent enough time in different parts of Mexico among different social classes from 1970s on, I put some trust not only in data but in ‘my lying eyes’ – neoliberalism has been a disaster and not only there. Why continue trying to justify evident failures or is a ‘save the theory’ sort of thing.
However, the numbers I posted above on poverty rates in Mexico are about as good as you’re gonna get so if your numbers disagree the problem is with them.
No problems then with an income rather than basic needs approach? No problem using PPP which aggregates prices of all commodities?
No problem with reliance on limited often questionable national data sets?
No problem relying on data series (WDI) provided by an institution which has been in neoliberalism’s vanguard?
Gee, next I’ll hear that tens of millions in Lat Am have been anti-IMF, anti-WB because these institutions have helped them so much.
Does X. Sala i Martin still believe that world poverty has been essentially eliminated or has he come to some senses beyond the ideological?
The comment trail seems to have led to Mexico and its NAFTA experience. That’s certainly worth debating, but all sides should be aware that this is not fundamentally about “free trade”. NAFTA differs from pure trade liberalization in many respects, and Mexico’s policy shifts are about more than just trade.
And one more reply to YNS: by trade balancing at the margin, I mean that a perturbation that, say, increases exports by a certain amount would also increase imports by the same amount. The use of marginal analysis here is the same as in economics generally — changes in some magnitude rather than the magnitude itself. Hence, in calling for models in which trade does not balance at the margin, I am looking past models in which the aggregate trade balance is fixed (for instance by global portfolio preferences), and only the composition changes in response to micro factors. There would be no Keynesian feedback in such a world from trade to the volume of employment or national income.
I apologize for using economic jargon. It’s easier for me to write but perhaps harder for many of you to read.
NAFTA is not ‘free trade’ in that it represents a set of international and domestic trading relationships – enforced by law – that protect the strong over the weak. Take away all the subsidies and protections for multi-national corporations and what do we have?
Free trade cannot be found where there is:
(i) heavy agricultural subsidies in the US that put downward pressure on Mexican (and other) farm prices.
(ii) Lack of recognition of humanity’s ‘ownership’ the environment that sustains them. Pollution occurs because governments collude with big business and allow the latter to poison the environment of other people – without charge. without recognition that there has been no ownership rights to the biosphere given or traded away and therefore no right has been given to pollute or degrade it.
(iii) Free trade cannot exist where different participants in trade are taxed at differing rates.
And so on.
As for YNS’ definition of ‘poverty’. Hey! The loss of 3 million hectares of native forest and jungle per year in Mexico is not just Mexico’s loss. It’s our loss.
It’s poverty. No fraudulent accounting system can cover that up.
If you want to live on $2/day ‘younotsneaky’ go ahead. Your commitment to such a level of income must surely also see you advocating the complete openness of the border between the United States and Mexico. After all, if $2/day is good for those down South it must surely be good for everyone!
““the most serious consequence of NAFTA has been its failure to protect the rights of workers as promised by its supporters. To attract investment to the maquiladoras, Mexican government authorities cooperated with investors and compliant official unions in maintaining low wages, reinforced with a system of labor control. According to Martha Ojeda, director of the Coalition for Justice in the Maquiladoras, the government-mandated minimum wage for workers on the border is about $4.20 a day, the same as 10 years ago. Ojeda estimates that a majority of maquiladora workers earn close to this wage. A study by the Center for Reflection, Education and Action, a religious research group, found that at the minimum wage, it took a maquiladora worker in Juarez almost an hour to earn enough money to buy a kilo (2.2 pounds) of rice. A gallon of milk, which costs $3 in a Tijuana supermarket, requires five to six hours of labor. To enforce this system, maquiladora workers are required to belong to unions that have no intention of raising low wages or ending dangerous working conditions. Throughout NAFTA’s 10-year history, workers have organized independent unions, willing to fight for a larger share of the enormous wealth the factories produce. But these efforts have been met with firings, plant closures and even physical violence. Ten years of hearings held under NAFTA’s labor side agreement have documented extensive violations of labor rights. In those few instances in which workers have successfully formed independent unions, as they did at Tijuana’s Han Young plant in 1998-9, their strikes were broken, despite guarantees under Mexico’s Constitution and federal labor law…”
NAFTA’s Legacy — Profits and Poverty, by David Bacon
Published on Wednesday, January 14, 2004 by the San Francisco Chronicle
http://www.commondreams.org/views04/0114-04.htm
The figures from Sala-i-Martin & the WDI are worthless:
http://www.columbia.edu/~sr793/LettertoEconomist.pdf
“a self-described “hardcore libertarian”
Propertarian would be a better description — they do, after all, have little in common with genuine libertarians like Emma Goldman.
Nor, for that matter, do they have any issues with the restrictions of freedom associated with capitalism and wage labour… Which is a strange position for someone who claims to favour liberty to take…
Iain
An Anarchist FAQ
THis is a long, productive and interesting lilst of comments on a significant economic issue that seems to have been “measured” to death. Anon’s citation, just above, brings everything back into the clarity of the light of day. Error of measurement x error of measurement x disputable assumptions = totally useless conclusions. That’s not the worst part of it all. Apparently a great many economists don’t find this phenomenon too disturbing and continue to base their arguments on such “research,” if that’s what it can be called. I’d call it ideological screed.
YNS, What’s your opinion?
“
If you want to live on $2/day ‘younotsneaky’ go ahead. Your commitment to such a level of income must surely also see you advocating the complete openness of the border between the United States and Mexico. After all, if $2/day is good for those down South it must surely be good for everyone!”
Don’t be dumb. No one’s coming even remotely to saying that living at 2$ is a good thing. I don’t know what else to say. Just don’t be dumb.
“The figures from Sala-i-Martin & the WDI are worthless:”
This looks like a potentially interesting criticism of the Sala-i-Martin numbers, it’s not clear however how it relates to the WDI numbers since the revisions was produced by WB researchers as well.
“YNS, What’s your opinion?”
My opinion is that economists generally try to use the best numbers out there, which can often still be pretty bad. Measuring people’s incomes, not to mention various intagibles, is a difficult business. But rather than seeing this as a economists’ ideological bias, this is usually just doing research in an imperfect world. If better, more convincing (i.e. constructed with better methodology) numbers become available usually people start using those (there’s of course the usual dissemination and diffusion lags).
The ideological bias here is more with people who a priori reject any numbers that don’t fit in with their preconceptions.
I plan on looking at the Chen and Ravillon numbers in more detail later this week. This doesn’t appear like something you can just make up your mind about in a few minutes of internet browsing.
There is a difference between developing a theoretical construct then generating the testable hypothsis that extend from that theory and then actually testing that(those) hypothesis in an emperical fashion and, otherwise, going through steps one and two and then being satisfied to rest on the laurels of ones brilliance in the field of theoretical economics. Less than veridical data is insufficient to uphold hypothetical constructs in the social sciences. Theoretical economics is fine as an intellectual exercise. It doesn’t cut the mustard when it comes to predictive validity.
In response to:
BR: “.. After all, if $2/day is good for those down South it must surely be good for everyone!”
YouNotSneaky said:
“Don’t be dumb. No one’s coming even remotely to saying that living at 2$ is a good thing. I don’t know what else to say. Just don’t be dumb.“
Well, please clarify what you meant YNS. After all you did say:
“Poverty rate for Mexico (% of people living at less than 2$ per day in PPP terms):
1984 — 40%
1992 — 22%…”
Is ‘poverty rate’ and ‘poverty’ two different things for you??
jack,
But there’s no ‘theory’ here. It’s all empirics – estimating income distributions across countries and time. I don’t know what you’re talking about.
Brenda
“After all, if $2/day is good for those down South it must surely be good for everyone!”
This is you saying something like this, and insinuating/implying that I or someone else has said that. It’s crap.
“Well, please clarify what you meant YNS. After all you did say:
“Poverty rate for Mexico (% of people living at less than 2$ per day in PPP terms):
1984 — 40%
1992 — 22%…””
Uhhhh… I don’t know what is there to clarify here. In nineteen eighty four, forty percent of all Mexicans were living on less than two dollars per day per day in purchasing power parity terms. In
nineteen ninety two, twenty two percent of Mexicans were living on less than two dollars per day per day in purchasing power parity terms.
How in heck do you get the implication from that, that there is something good about people living on 2$ a day, or that that is enough?
“Is ‘poverty rate’ and ‘poverty’ two different things for you??”
Poverty is being poor. Poverty rate is number of people who are poor divided by total population.
Again, what is there to clarify?
YNS,
“Its all empirics – estimating income distributions…”
Estimating something is akin to educated guessing. The fact that those guesses are based on data that are subject to significant errors of measurement only makes the endeavor even less empirical. Putting a ruler to an object doesn’t give us a measure of that object if the gradations on the ruler are inaccurate, except if you’re a social scientist who can’t recognize the limits of his data collection techniques or the inadequacy of the type of data to begin with. Simply wishing that your science is as accurate as physics doesn’t make it so. Physicists know when they are stepping out over the line of empiricism and identify themselves then as theoretical physicists. They state with relative clarity that what they are engaged in is formulating reasonable hunches about the nature of the universe. There’s nothing wrong with such a branch in economics so long as the participants know which branch they’re hanging off of.
Btw,
Here’s (one of) the Chan and Ravallion papers that the letter to the editor provided by anonymous (thanks for linking it) references:
http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2007/04/16/000016406_20070416104010/Rendered/PDF/wps4211.pdf
(yes, I just got my pdf capable computer back)
It should be noted that even with these numbers, the incidence of poverty has fallen although the absolute numbers have stayed constant or gone up somewhat in a world of positive population growth.
I’m not 100% clear on what accounts for the difference between these estimates and those of other researchers as the general methodology is actually in an older paper (2004) which I haven’t been able to find online yet.
However, in general there’s 3 issues, or 3 classes of issues here, when trying to come up with poverty measures (or for that matter, inequality, or any moment of the income distribution):
1. Estimating the general distribution of nominal income.
1a. What source to use? Household surveys or national income accounts, or yet some other source? Or is there some way to utilize all information. Both of these have their disadvantages and advantages. No free lunches here.
1b. Given that you’ve settled on a particular data source (or mix of them) what method to use to infer an overall distribution from the few, possibly non-random observations you got? Kernel estimation or nonlinear least square? What size bins to pick? Baysian methods? Etc. All methods have their disadvantages and advantages. No free lunches here.
2. Converting it all to numbers that are comparable across countries and time
2a. PPP adjustments
2b. What price index to use? It seems like at least part of the possible problem with the SiM #s is that he used a general price level (though, contra that letter I believe it was the CPI not the GDP deflator) rather than a price index specific to poor households. However it is notoriously hard to get both the prices of individual, disaggregated goods, as well as to figure out what a typical consumption basket of poor people in various countries actually is.
3. Conceptual issues – ‘what is poverty’
3a. Jack (or was it juan) above mentioned basic needs. So a poverty line should reflect the inability to acquire these. And 2$/day may be too low for that. Actually I think the standard 1$ and 2$ benchmarks are generally used because they’re round numbers and they underscore the seriousness of the problem. As long as the same consistent benchmark is used and it’s not, I dunno, 100$ per day or something, then it should be fine.
3b. More problematically, should poverty be measured using consumption or income data? From a cursory reading this appears to be another source of divergence between the two numbers. I think you can make some arguments both ways here.
Ok that’s it for now.
YNS,
You see, that wasn’t too difficult. A well organized review of the deficiencies inherent in the data collection and summarization processes and the difficulty of the analysis that follows. It would be helpful in any discussion of any social phenomenon, of which economics is one, if there were a more wide spread recognition of these short comings. I’ve not been saying not to do the research as it is. The preface in such studies is all too often woefully lacking in disclosure. Simply put, confession is good for the soul, or the argument, as the case may be.