[Matticus_Rex]

I'm fairly well acquainted with the literature on this topic, and I think the opposite; the fact that most people don't understand why these "nitpicks" actually seriously undermine the premise and totally butcher the prescriptive suggestions makes it easy for readers to dismiss the criticism, especially in a general-interest publication. While there's a contrary effect for the skimming reader who just accepts that P&S' work is flawed without understanding it, I think that tends to be a much lower risk with the Economist audience.

[CharlesColeman]

> I'm fairly well acquainted with the literature on this topic

How so, exactly?

> and I think the opposite; the fact that most people don't understand why these "nitpicks" actually seriously undermine the premise and totally butcher the prescriptive suggestions

Personally, I'm inclined to withhold judgement until the dust settles, yay or nay. The topic is technically difficult, and there are a lot of actors with political interests in the outcome of the discussion that muddy the waters.

[rayiner]

My biggest problem with Piketty & Saez's work is that it made pretty major economic policy prescriptions based on a model that was so simplistic that the result was borderline click-bait. And the end-result is that assertions have entered the political discourse ("the middle class's standard of living has been stagnant since the 1970s") that just aren't true.

Let's unpack that. If you want to quantify the middle class's standard of living, you have to look at:

1) Post-tax income. 2) Prices/product quality/ 3) Benefits/services. 4) The characteristics of the middle class (age, race, immigration status, household size, etc).

Piketty & Saez look at pre-tax income, apparently only because it's easy to assess via tax receipts. That overlooks data that's highly relevant to addressing deep and substantial policy questions. For example, consider what would happen if the U.S. and Mexico merged. Immediately afterward, median income in the U.S. would drop significantly. But nobody's standard of living would change. I.e. looking at the "median" fails to account for changes in who comprises the median. In the U.S. the immigrant share of the population has almost tripled from 1970 to 2017. Even if you've got a great economy that's good at integrating people and equalizing their incomes over time, it's probably not realistic to expect first generation immigrants as a group to have the same income as native born people. Looking at just pre-tax incomes completely masks that effect.

Similarly, marriage rates and household sizes are down significantly since 1970. Married couples make significantly more money than non-married individuals. The median marired couple in Michigan has an income of $80,000, which is 45% higher than the combined income of the median single woman and median single man: https://www.mlive.com/news/2017/09/michigans_median_income_i.... Looking at pre-tax household income overlooks that effect entirely.

There's a story in here somewhere. But Piketty and Saez's analysis doesn't actually tell us anything.

[mensetmanusman]

Okay, but it’s not quite true that ‘nobody’s standard of living would change.’

Cartels would be free to cross the border and operate in the U.S. (no way the U.S. could handle a sudden spike in 100,000+ Member sophisticated criminal operations), so their quality of life would increase.

The lower class in the U.S. would have even more wage competition, so theirs might decrease, etc.

[dangerlibrary]

You are mixing up correlation and causation re: married couples.

[TheOtherHobbes]

Over the time period, married couples often had a single income.

Now working class married couples are just as likely to have a total of two to six jobs between them, but their real median income is still lower than it was when only one person was working.

There are no "nit picks" to debate. Refusing to acknowledge the destruction of financial stability for the working class and lower middle class in the US and certain other Western countries is simple denialism.

[xyzzyz]

> Now working class married couples are just as likely to have a total of two to six jobs between them, but their real median income is still lower than it was when only one person was working

This is very much false. Look at the data. Real median income of all households had risen by a relatively small amount, because there are many more households today with one adult than in the past. However, real median income of households with two adults had rose very substantially. Similarly, real median household income of households with one adult also rose substantially. The relative stagnation of real median household income is mostly an artifact of changes in the composition of the index.

[fullstackchris]

> because there are many more households today with one adult than in the past

Not only that, but we should remember that there are simply more _people_ globally. Using the U.S. as an example - when my dad was born in 1954, the U.S. population was HALF of what it is now. Should we assume that all markets and wage/labor share increase linearly AND directly (in time) with population? Certainly not.

Somewhat tangential, this population factor also plays a major role (though not the only one) for the reason city or near-city housing prices have skyrocketed - building construction just didn't keep up with the sheer population growth in many advanced economies, even more so in cities of advanced economies. This is why my buddy in Palo Alto is, for all intents and purposes, forced to live with two retirees and 4 other Millenial devs to afford rent.

[darawk]

The relationship is causal. Married couples are a household. Married couples consist of two people. Two people is more than one person. Two incomes is more than one income, etc. Even though all married households are not dual income, those that are shift up the mean for the rest. This is an artifact of measurement via 'household'.

[likpok]

You've missed rayiner's argument, which is that the household income is more than the sum of the median man and the median woman. That is: two married people make more than two single people.

There are a number of arguments you could make as to why (for example: married people tend older and so are farther in their careers), but it's not just "married households have two incomes and single households have one".

[darawk]

> There are a number of arguments you could make as to why (for example: married people tend older and so are farther in their careers), but it's not just "married households have two incomes and single households have one".

Sure I didn't say it was the only explanation. But it is a sufficient one to prove the relationship is causal.

[dennisgorelik]

People who make more money are more likely to marry and to stay married.

[mensetmanusman]

True, but even poor people who decide to get married see a benefit relative to cohabitation.