The Curse of Econ 101

The thing that the Econ 101 demand curve leaves out of the minimum wage debate is that those extra wages have side effects that impact the curve itself. Workers who are paid more at their primary job may not need to work a second job and/or will have more money to spend at other establishments that also pay their employees minimum wage, driving back up the demand for workers. Also, workers who are less stressed about affording the repairs on their car, or making the rent this month, or paying for their kid's medicine, will be better and more productive employees, generating additional value for their employers. These effects are harder to measure and imagine than a simple demand curve. They encourage more automation, sure, but automation has side effects, too. It keeps prices from growing as fast, which means people can spend the money they save elsewhere on other products or services, which increase demand for workers in that industry/establishment; if designed correctly automation can make workers' lives better even if there don't need to be as many workers; and automation itself is a product that needs to be designed, sold, purchased, installed, maintained, and upgraded, all of which drive up demand for still other jobs.

Sure, there's a point where a minimum wage would be too high and would start causing drag on the economy, and while I personally don't have a handle on the economics required to calculate an estimate of where that point is, it's clear that we aren't anywhere close to that point. Historical minimum wages have been higher, and just basic common sense tells you that so long as inflation is happening, the minimum wage should also be going up to keep pace. The "living wage" standard that a full-time minimum wage job ought to cover basic living expenses for a typical family seems like an entirely reasonable standard that has the benefit of being somewhat possible to be objectively determined.

Supply and demand curves aren't observable: only their intersection, price, is observable. The curves also move about in unpredictable ways. So we've got this situation where unobservable quantities move with unknown dynamics in a noisy, unpredictable fashion.

In order to get anywhere with a statistical estimation problem as hard as that, you have to impose a lot of prior structure. I'd be surprised if, much of the time, you didn't end up effectively assuming the answer you want to see.

I'd go as far as to say supply and demand curves are a pretty useless abstraction. I've watched two economists argue over whether the 2015 oil price slump was a supply shock or a demand shock: if they can't get that right, what hope do the rest of us have?

TL;DR The Econ textbook model of rising prices —> lower demand is incorrect for minimum wage labor. Areas that have raised minimum wages have not reduced employment.

I believe many economists (including this author) are missing the structural changes caused by technology. The labor market is not just humans any more, and if/when human labor prices go up, the economics of automation are even more attractive. Ask any robotics startup.

As the article states, the real dollar value of minimum wage labor hasn’t gone up much in decades. This fact (combined with a relatively low unemployment rate) seems to puzzle many economists. But it makes a lot more sense when you consider the labor “market” to include robots and automation.

Economic models usually assume free movement of labor when actually labor is quite restricted in its movement thanks to immigration laws. Capital is very mobile and can flow easily to where costs are low.

I have a Master's in Economics. In graduate classes, much of the work done in intro-level courses is tossed aside for more capable models that require a fair bit of calculus to understand. The easiest being the Solow growth model.

The intro courses delve into macro static models with at most 2 variables (think Keynsian Cross). In the graduate level, macro theory is taught as micro theory. While the deleterious effects of price controls still hold true at higher levels of economic reasoning, it's a multi-variate problem. With any price control, think of it as damage the rest of the people acting in the economy think of ways to route around. If a minimum wage increase takes place, perks can be cut without affecting overall employment levels as an example of how overall employment may stay the same.

The claim that Econ 101 does not apply to the labour market is a pretty extreme claim. If the studies aren't showing a consensus, I'm betting on Econ 101. Note that below, jasode has pointed out that the 1994 Card & Krueger study was based on phone surveys rather than payroll data.

There is some real and interesting debate on the question of should the cost of minimum living standards be borne by companies (through the minimum wage) or governments (through straight welfare). I'd love to see a compromise where it was still economic to do low-value work that is currently not done due to minimum wages.

The minimum wage argument assumes that corporations choose whether to hire based on an evaluation of the productivity of a worker. In reality, nearly all companies choose a wage by simply looking at their neighbors.

>They concluded, “Contrary to the central prediction of the textbook model ... we find no evidence that the rise in New Jersey’s minimum wage reduced employment at fast-food restaurants in the state.”

This is really the heart of it, people who are libertarian/pro free-market/Austrian economists simply reject any evidence that challenges their assertions about economics: https://en.wikipedia.org/wiki/Praxeology#Austrian_economics

>Austrians argue that that empirical data itself is insufficient to describe economics; that consequently empirical data cannot falsify economic theory.

I've felt for a long time that most of Economics is just political ideology masked as science. That's just me though.

That's a lot of words intended to obscure a very simple subject.

1. Theoretically, a minimum wage will reduce unemployment, and the higher the minimum wage, the higher the impact. However, even in theory it has to be quite high to get meaningful impacts, given how noisy employment data is.

2. The minimum wage in the US is quite low, and has very rarely been increased sharply. As a result, we've never really seen any large spike in unemployment, but we also have no reason to think we wouldn't if we ever did. Nobody has ever tried a minimum wage of, say, 80% of the median wage to see what it would do.

3. Although not entirely conclusive, the data from the largest increases does seem to show spikes in the minimum wage. For example, the University of Washington released their study on the impact of Seattle raising their minimum wage from $11 to $13 per hour, and found that the reduction in hours was 3 times higher than the increase in wages, meaning minimum wage workers lost $125 per month in earnings[1]. By contrast, the best studies which show no effect covered smaller increases or increases to a lower level (as a % of median wages).

For example, one of the most famous of the "minimum wage has no effect" studies was by Card and Krueger, and covered a rise in the minimum wage in NJ from $4.25 to $5.05 in 1992. Adjusting for inflation, $5.05 is around $8.90 in 2017 dollars. It's entirely plausible that $9 might not cause any measurable impact, but $13 would. In fact, there is some data which suggests that around 50% of median hourly wages is an inflection point at which point you start to see real impacts. And median hourly wages in the US is slightly less than $25/hour, which is slightly less than twice $13, so in fact that's entirely unsurprising.

> When the University of Chicago Booth School of Business asked a panel of prominent economists in 2013 whether increasing the minimum wage to $9 would “make it noticeably harder for low-skilled workers to find employment,” the responses were split down the middle.

$9 is a level at which we'd expect to see a minimal effect. It's plausible to think that there might be SOME effect, or that there might be no measurable effect, and some studies have shown such an effect while others have not. It's not surprising that economists would split. But splitting over $9 would have an impact doesn't mean there's any debate over whether an impact exists at some level, and in fact, there is not. And it seems like $13 does have a measurable impact, and there's really no doubt that $30 or $50/hour would have a very major impact.

> The real impact of the minimum wage, however, is much less clear than these talking points might indicate.

No. There is, at most, a debate over how big an increase it would take for the impact to become clear.

[1]: http://www.nber.org/papers/w23532.pdf

I completely disagree with the article's conclusion. But, this is a pretty good article. It actually cites opposing views, including linking to them, and tries to make its case logically. Great writing!

That said, I'm not an economist, but when I see an issue such as this, in which leading economists disagree with some position, I don't think the takeaway of one article should be "I now know what the truth is". I mean as the article itself says, lots of leading economists think a minimum wage is bad. So if after reading the article your position is "well clearly I now know better than people who have spent their whole lives on this topic", you're being arrogant.

Also, from the economists I do hear on a regular basis, it's pretty clear that they're all very anti-minimum-wage. Obviously this is a self-selected sample, but I get the feeling that most economists aren't particularly pro minimum wage - they're either indifferent or at most think that there are some tradeoffs that can be made in favor of the minimum wage. But there are tradeoffs, as in the econ 101 model isn't wrong. It's just worth it sometimes.

On the "empirical" question, I tend to side with Bryan Caplan here - there are studies going both ways, but since in almost every other situation basic supply/demand models are correct, there's got to be really strong evidence against assuming supply/demand models work for labor, too.

This is an interesting article that was cited, pulled out to highlight.

http://davidcard.berkeley.edu/papers/njmin-aer.pdf

I don't think that there is any point to having a minimum wage if the Federal Reserves' mandate of price stability is 2%.

There is so much wrong with this article.

"Many companies can recoup cost increases in the form of higher prices; because most of their customers are not poor, the net effect is to transfer money from higher-income to lower-income families."

Higher income people aren't the primary customer for McDonalds (in the US.) If you raise a Big Mac price by $2, "rich" people might buy the same amount, but poor people will buy less and poor people are a much larger share of the Big Mac market than rich people, thus it would be a net loss.

"Even if a higher minimum wage does cause some people to lose their jobs, that cost has to be balanced against the benefit of greater earnings for other low-income workers."

This is just nuts. Some guy now makes $20k per year, but some other guy now makes $0 per year, instead of both making $15,000 per year. That's a net loss of $10,000 per year in economic activity. It makes one person slightly better off, while making the other person infinitely worse off.

The other interesting thing about this debate is that those arguing for higher minimum wages also tend to advocate for unrestricted immigration and so-called sanctuary city policies which necessarily results in a greater supply of lower-skilled labor. Without a minimum wage, immigration of lower-skilled workers would necessarily regulate itself as wages fall to equilibrium. You don't even need immigration law -- the free market would limit and allow immigration as wages moved. Essentially it would make the market for labor far more liquid.

Despite the article's derisive condemnation of "Econ 101," supply and demand is still a thing. An increased supply of labor means that the equilibrium price is now lower, which with the price floor of minimum wage means that there is an even greater surplus of labor.

The article seems to criticize so-called "Econ 101" while dismissing Nobel economists Friedman and Hayek. I'm pretty sure those guys were far beyond Econ 101. The author talks about how in the 1950s 1 and 3 workers were in a union, with the implication that the higher wages that resulted were good for society. However the author conveniently leaves out that the US had a poverty rate of 22.4% in 1959 and in 2015 that poverty rate was 13.5%.

Perhaps there's an argument to be made for higher minimum wages, but this article certainly hasn't made it.