Thursday, February 02, 2012

Thursday Roundup (2/2/2012)

Posts you may have missed in the blogosphere this week...Ride em econ cowboys!

1. Matt Yglesias reminds us that it is never wise to give our money to the experts, no matter how much money they make for themselves.

2. Tyler Cowen argues that inequality, at least recently, has been mostly about stagnating incomes and the explosion of the income of the top 0.1%. Very true. (Tyler's post also contains some insults in an addendum at the bottom, which I do not endorse.)

3. John Cochrane demonstrates the perils of over-differencing time series data. This is a specific case of a very deep problem with time series econometrics: rigid lag structures.

4. Steve Keen has a great essay on the problems facing the field of macroeconomics.

5. James Kwak explains how private equity works, how it is is supposed to benefit the economy, and how it might not actually benefit the economy.

7. Brad DeLong uses an equation to explain his thinking on fiscal policy. Unlike a lot of people in the blogosphere, I love it when people do this. And here, he explains his thinking on monetary policy. Read!

8. Scott Sumner points out that Britain talks a lot about austerity but has actually been running big budget deficits.

9. Intra-blog civil war at Modeled Behavior, as Karl Smith doubts that government workers can be overpaid, while Adam Ozimek thinks they most definitely can. I score this one for Adam (to see why, consider the marginal product of Chairman Mao and compare it to his total compensation package). In fact, Adam is so obviously right that I hereby award Karl a bat boy! ;-)


  1. I agree that government workers can be overpaid - it doesn't really make sense to say otherwise. Wouldn't it be more reasonable to argue that they *aren't* overpaid, even where in some cases it would appear that they are? After all, we're not paying letter carriers millions of dollars; we're paying federal administrative workers about twice their private sector counterparts, which is still not a huge salary. Maybe the private sector is underpaying its less exceptionally educated workers. But maybe I'm asking a different question - what should people be paid as a matter of policy, rather than, how much is their work worth in value.

  2. It's also worth noting that we tend to pay lower-level and blue collar workers relative to the private sector average, but underpay (sometimes vastly underpay) public sector workers with postgraduate degrees or with specialized knowledge.

    The military has especially been having problems with this when it comes to doctors, who are offered all kinds of bonuses and incentive pay depending on their specialities. This can (sometimes often) mean that MDs in uniform get paid more that people several paygrades above them. And yet we still have trouble retaining doctors, because private sector health care market compensation is just impossible for Congress to keep up with.

  3. Absolutely true and very problematic. I'm a government attorney who gets paid literally about half of what my private sector counterparts get paid (and I'm well paid for a government attorney)! For me and for many others in this position, the payoff is in quality of life and benefits (among other more intangible reasons it would take much longer to explain). Those working in the private sector have less job security and work many more hours than I do, in addition to having far less flexibility.

    So, for me, it's absolutely worth it, and my sense from my colleagues is that they agree. Of course, that doesn't apply in the military context, and I'm not sure how it would apply in other specialized professions. But, my question is, how do we deal with that/incentivize highly trained and specialized professionals to work in the public sector (as I've been incentivized), whether through salary or by other means? Not, why are we paying administrative staff and blue collar workers so much when specialized workers get paid (relatively) so little?

    I'm no longer sure I actually made any kind of coherent point there.

  4. What are differencing and over-differencing in plain English? thanks,

  5. Anonymous7:01 PM

    "If not for Ellis Island I'd be 'Noah Swartz Kuznets'"...

    And likely the esteemed creator of the "Swartz Kuznets Swirly"

  6. @Luke Lea:

    Differencing the data is when, instead of regressing Y on X, you regress the change in Y on the change in X.


    I wouldn't say I invented it...

  7. This Mike Woodford guest post is potentially very important and valuable:

    I say potentially, because I haven't been able to really go over it well yet. But it's a rare statement of positions from the very important Woodford.