Friday, January 16, 2015

Increase Minimum Wage, Kill Jobs


Raising the minimum wage as a way to reduce poverty has become all the rage lately, with lots of states doing what many think the federal government should do. Claims about the effect of the minimum wage in popular discourse have largely split along predictable ideological lines, with conservatives saying that minimum wage increases are job killers and liberals arguing that the wage increases don't kill jobs and help alleviate poverty.

Oddly enough, economists seem equally split on an issue that one might expect to be resolved easily with data. Certainly, standard labor economics argues that making labor more expensive should reduce demand for labor (what they call a negative wage elasticity). The scholarly consensus strongly supported this idea up until the 1990's, when new approaches began to muddy and shift the picture, starting with the famous study by Card and Krueger (1994) that found no disemployment effect from a minimum wage increase for fast-food restaurants in New Jersey. Overall, studies still seem to favor a small but negative effect, though there are claims that even these effects are due to publication selection bias, as illustrated in this funnel graph:


This past year provided a perfect opportunity to test these ideas, with 13 states increasing their minimum wage on January 1, 2014. Of these, nine made small cost-of-living adjustments (and have been for a number of years) but four--New Jersey, New York, Connecticut, and Rhode Island--made substantial one-time increases. The two largest, $1.00 in New Jersey and $0.75 in New York, are large enough that they should show significant disemployment effects if there are any. Indeed, in the early summer of 2014, many in the media trumpeted the results that the 13 states with minimum wage hikes had better growth rates than those that didn't. Even President Obama got into the act.

So is there any observable effect? The graph at the top shows not job growth rates, but percent change in the job growth rates from fiscal year 2013 to calendar year 2014 (Jan. - Nov.) as a function of the percent change in the minimum wage. What everyone ignored in previous analyses is that states with the largest increases have done the worst in maintaining job growth, leading to a downward slope in the best fit line. Indeed, the fit indicates that a 10% increase in the minimum wage correlates with 1.5% decrease in the number of jobs. This number is especially interesting given that the percent of workers directly affected by these raises is roughly 10% of the workforce. These data clearly support the claim that minimum wage increases are job killers.

Friday, January 9, 2015

So NOW the Republicans Want to be Dynamic?

Now that the Republicans are in control of Congress, they are looking to transition from a defensive (some would say obstructionist) posture to an active one. Their first step is to implement so-called 'dynamic scoring' on legislation with budget impacts as large as 0.25% of annual GDP or higher. On its face dynamic scoring seems to make a lot of sense--it takes into account the effect of the legislation itself on economic growth and the resulting federal tax revenue. For example, tax cuts would be scored not only for the loss of revenue but also for the (likely) smaller increases in revenue due to increased growth. The Republicans like this idea because it may make tax cuts look better and therefore easier to pass.

Not surprisingly, Democrats have taken a dim view of this approach, with the White House budget director, Shaun Donovan, being particularly vocal:

“Dynamic scoring can create a bias favoring tax cuts over investments in infrastructure, education, and other priorities.”

The main objection appears to be a political one, namely that dynamic scoring will make Republican priorities easier to pass at the expense of Democratic ones. To an extent this is true, since dynamic scoring as proposed by the House would only apply to tax changes and not to the effects of changes in infrastructure spending.

Still, the overall effect of this change is likely to be  relatively small. Only a few bills each year would qualify for the 0.25% threshold, and those that do wouldn't see a dramatically different scoring. A recent example was the tax reform bill offered last year by Rep. Dave Camp (R., Mich.). With dynamic scoring, the plan was estimated to add only $5 billion to $70 billion per year to federal revenue.

The biggest problem seems to be the uncertainty in the estimates, which depend on various untestable assumptions. In particular, the computer models can't calculate an outcome without assuming that debt to GDP eventually reaches a constant percentage, and guesses about that percentage and how it will be achieved can give a broad range of outcomes, as observed in the Camp plan.

Ultimately, dynamic scoring looks like it could be the right way to go, but it would benefit from a couple of improvements. First, the uncertainty inherent in the estimates should be acknowledged by requiring that the full range of possible outcomes be included in the official scoring. Ideally, this would be presented as a best estimate and uncertainty, as is done in any scientific endeavor. Second, Congress should consider applying dynamic scoring to more than just tax legislation, for example infrastructure spending. At the very least, these improvements would help remove the appearance that dynamic scoring is primarily a tool to gain a political advantage.

Thursday, January 8, 2015

Is 'Extreme Conservative' an Oxymoron?

In the self-replicating world of grievance-mongering known as the Internet, it is hardly surprising that everything seems extreme. In a sea of a billion websites, the easiest way to get noticed is to be extreme, or to accuse others of being extreme. Nowhere is this more true than with political ideology, where conservatives and liberals battle to cripple the other side.

But which side is more extreme? Certainly, there's no shortage of claims in the media that conservatives are. Governor Andrew Cuomo is famous for claiming that 'extreme conservatives' are so extreme that they 'have no place in New York.' Even a left-leaning outlet like MSNBC is extremely conservative, according to Russell Brand. Google and Yahoo agree. A Google search for 'extreme conservative' returns 115 million hits compared with 78 million for 'extreme liberal.' Yahoo gives a similar ratio, with 4 million versus 3 million.

The odd part of this consensus is that the term conservative would seem to be the opposite of extreme. Liberals are known as being creative, radical, and willing to push the envelope; conservatives, not so much.

One way to approach this conundrum is to find an operational definition of extreme and then test how conservatives and liberals behave. Since talk is cheap but action is more meaningful, making beliefs operational is a way of ensuring honesty.

A non-partisan political analysis firm, Crowdpac, has done this by scoring the ideology of political donors based on the politicians to whom they have given money. For example, a donor who gives solely to highly liberal (Democrat?) politicians would be rated most liberal (10L) while one who gives solely to highly conservative (Republican?) politicians would be rated most conservative (10C).

Not surprisingly, Crowdpac found that some professions skew heavily one way or the other, with the entertainment industry and academia being the most liberal and mining, agriculture, and oil industries being the most conservative.

What is interesting, however, is the shape of each distribution. Consider newspapers and print media, which skew heavily liberal.

It's not simply that liberal donors outweigh conservative ones. Rather, note that the modal response of those donors is most liberal (10L). That is, of all donors that could be rated liberal, being as-liberal-as-possible was the most popular choice. Conservative donors, by contrast, don't prefer to give to highly conservative politicians, with a modal response in the middle (7C).

The obvious claim in response to this data is to suggest that extreme conservatives have been weeded out by the left-leaning profession, and that truly extreme conservative donors go into other professions. If so, we should see the opposite skew in conservative professions, like oil, gas, & coal.

Here we see that conservatives strongly outweigh liberals, as expected. But note that the modal responses are the same as before. For the much smaller group of liberals, donors still prefer to give to the most liberal politicians over any other outcome. The dominant conservative group, on the other hand, is still relatively diverse and centrist, with a mode of 7C.

A thorough inspection of the Crowdpac data shows that these tendencies hold for all professions that skew either right or left: liberals have a modal rating of 10L while conservatives have a modal rating of 5C - 7C. Only in highly mixed professions, like hedge funds, lobbyists, and banking does the liberal mode shift to non-extreme values, like 7L, though the 10L response is never as small as 10C for any profession (except mining).

The data seem pretty clear. Extreme is as extreme does, and conservatives don't do it well.

Monday, January 5, 2015

On the Significance of True Baboo

Since this is my first blog post, I've decided to briefly explain the meaning behind the name. Baboo can mean a variety of things, such as a term of respect for adult males, an English-speaking clerk in India, or as a term of endearment made famous by Sally's Sweet Babboo in Peanuts.

For me, baboo began as a term of endearment. As a graduate student in the 1990's, my university's e-mail format was the first letter of the first name followed by the first three letters of the surname. Mine was bbue. My girlfriend at the time, now wife, thought this tag was hilarious and jokingly referred to me as baboo. Neither the joke nor use of the term lasted long, but I would sometimes use it on-line when I wanted to maintain anonymity. Over time, I became baboo.

The meaning for this blog, however, connects more closely with its Indian use. As an English-speaking clerk, a baboo would have played the role of bridging the gap between the English-speaking British, who controlled India, and the native Indian population. The goal of this blog is to bridge the gap between what is known formally through science (broadly understood) and everyday society. The blog will specifically focus on analyzing current events and hot issues through the lens of rational thought and the tools of science. Given how poorly science and scientific thought is understood by the general population, it is my job as an educator and scientist to bridge the gap and bring clarity to the relationship between science and society. My goal, then, is to serve as a true baboo, a genuine go-between and translator in the context of science of the issues that perplex society.