Monday, January 26, 2015

"market monetarism passed the test as set up by Keynesians, using the Keynesian ground rules"

See Monetary offset: Reply to my critics by Scott Sumner.
"Not surprisingly, there has been lots of criticism of my claim that the Keynesian test of 2013 failed. Let me respond to some of the points:

1. I was accused of cherry picking dates, as I compared growth in the 4 quarters before and after the onset of austerity on January 1st, 2013 (when all the tax increases kicked in--the sequester came a few months later.) Growth rose from 1.60% in 2012 to 3.13% in 2013 (Q4 to Q4.) But my critics are correct that 2012 was an unusually slow year, so maybe a longer time period would be better. Here are growth rates over:

The previous 2 years: 1.65%
The previous 3 years: 2.04%
The previous 4 years: 1.47%
The previous 5 years: 0.59%
The previous 6 years: 0.81%
The previous 7 years: 1.05%
The previous 8 years: 1.33%
The previous 10 years: 1.91%
The previous 15 years: 2.51%

All lower than in 2013. (I ignored compounding to save computational time; the actual growth rates were slightly less, strengthening my point.)

So cherry-picking data isn't the issue. What if you go forward more than 4 quarters, say 2 years? The winter 2014 quarter was slow because of unusually bad winter; however both the spring and summer of 2014 were red hot. It looks like 2014 will also show decent growth when Q4 data comes in.

Another complaint is that the increase in growth in 2013 was not significant. There are actually two issues here, measurement error, and the problem of ceteris paribus. As far as measurement error, I've always acknowledged that the government probably overestimated the speed up in 2013, as other data like job creation shows a much smaller acceleration. But the point is that even the other data shows faster growth, which refutes the Keynesian prediction that growth would slow.

A better argument is that the speed up was within the normal year-to-year fluctuations reflecting all sorts of factors. To see the problem with this argument, we have to go back and look at the "test," and consider what the Keynesians were trying to show. It might be helpful to first look at a case where the RGDP data went as the Keynesians expected, Britain after the election of the Conservatives in 2010.
The Conservatives were accused of slowing the British recovery with a policy of "austerity." I use the scare quotes because Britain continued to run just about the largest budget deficits in the world during the early years of Cameron. But let's accept the Keynesian method of estimating changes in cyclically-adjusted deficits. One thing I noticed is that Britain had a very odd growth slump:

1. Britain continued to generate more jobs than many other developed countries.
2. Britain experienced relatively high and rising inflation.

Now I'm not arguing that Britain had no AD problem, I think it did have one. But given the jobs growth, surely some of the British slowdown in RGDP growth was due to productivity factors unrelated to low AD. Some have pointed to less North Sea oil output and less earnings from big banks in the City. Perhaps the "big government" policies of the previous Brown government slowed trend productivity growth a little bit. I don't claim to know all the reasons, but Britain would be a textbook case where you might want to question whether it was austerity, or some other factor that explained the RGDP growth slowdown. The ceteris paribus problem.

Nonetheless, the impression I got reading people like Paul Krugman and Simon Wren-Lewis was simply; Austerity ---> RGDP slowdown, case closed.

Suppose that if instead of increasing from 1.60% to 3.13% in 2013, growth in the US had slowed by an equal amount (to near zero). Let's be serious for a moment, and please answer this honestly. Does anyone think the Keynesians would have been saying, "Gee, that pause in the recovery can't be attributed to austerity, because the drop in RGDP growth is not statistically significant?" If any reader answered "yes," I hereby accuse you of intellectual dishonesty.

Now some want to argue that even if Krugman, et al, got this wrong, and also used sloppy techniques for considering UK and eurozone austerity, this doesn't definitively prove market monetarism is correct or Keynesianism is false. Sure, I'd agree with that. Personally, I prefer market tests. I like to look at how market prices respond to new information about monetary policy. And of course this is one reason why the Fed needs to subsidize trading in NGDP (and RGDP) futures markets. And I'd prefer looking at NGDP growth, whereas the Keynesians use RGDP growth.

My point is different; market monetarism passed the test as set up by Keynesians, using the Keynesian ground rules. Their own model failed their own test.

PS. Sometimes Keynesians refer to more systematic studies, but these generally involve lots of observations for regions lacking an independent monetary policy. Numerous researchers have found the correlation goes away if you exclude observations lacking an independent monetary policy. (Mark Sadowski, Kevin Erdmann, Benn Steil & Dinah Walker.)

PPS. I am certainly not dogmatic on this issue:

1. Fiscal stimulus that lowers inflation can be expansionary, even with monetary offset (VAT cuts and employer-side payroll tax cuts are two examples of fiscal stimulus that might work by encouraging monetary stimulus to raise inflation up to target.)
2. Supply-side cuts in capital taxation can boost real GDP growth.
3. Spending on wasteful things like military output can boost RGDP (at the expense of lower living standards) by encouraging people to work harder to try to maintain living standards.
4. If central banks are incompetent in a very specific way then fiscal stimulus might help. But not incompetent in the way the ECB was incompetent when they tightened in 2011 by raising their target rates. More (demand-side) eurozone fiscal stimulus in 2011 would not have helped."

By cherry-picking data points that were favorable to his narrative, Piketty was able to “show” an increase in inequality over the last three decades.

From historian Phillip W. Magness.
"My previous posts on the data problems in Piketty’s Capital in the Twenty First Century have focused almost entirely on errors contained within his data charts and files. But what happens when one tries to reconstruct those files?

To find out, I conducted a simple experiment using Piketty’s Figure 10.5 – the widely cited depiction of wealth inequality in the United States over the past century. I previously deconstructed and critiqued this chart at length, concluding that it is essentially a Frankenstein graph – a clunky assemblage of cherry-picked data points from multiple divergent sources and arranged in an order that seems to confirm Piketty’s historical narrative about a dramatic upturn in inequality since the early 1980s. The purpose was to reconstruct Piketty’s chart using his own source data and techniques, only I would cherry-pick different “representative” numbers than Piketty did from within those sources as needed.

The rules:
1. I had to use the same data sources that Piketty used in constructing the original. These consist of the estate tax study by Kopczuk & Saez (2004) and SCF studies by Wolff (1994, 2010) and Kennickell (2009, 2011).
2. My time series would be constructed the same way that Piketty constructed his: decennial averages of the century long trend, using data points from the aforementioned studies to obtain a “representative” estimate for each decade.
3. I could only use adjustment techniques that Piketty also used in reconciling differences within the sources. This included applying an adjustment ratio to bring Kopczuk-Saez in line with the SCF studies, light averaging to fill in gap years if needed, and a rough estimation of the top 10% from top 1% numbers where such numbers were not available (i.e. copying Piketty’s technique of adding 36 percentage points to the lower trend line).
4. Where source data conflicted I was free to select between preferred data points to form the representative estimate for each decade, in keeping with Piketty who did the same.

The results of the experiment are as follows:

CherryPickingPiketty

The two red lines represent Piketty’s original depiction of U.S. wealth inequality. The two black lines represent my “alternative” depiction, as derived from the exact same data sources. The differences are most evident second half of the chart. Specifically: Piketty tended to cherry-pick data points that suggested increasing inequality after 1980. My alternative does the opposite, cherry-picking data points that suggested flat or decreasing inequality. Please note that I make absolutely no pretenses that my method is more accurate. It simply exists to illustrate that one may obtain dramatically different results by preferentially selecting certain specific data points over others while still retaining Piketty’s own data sources and blending methods.

The effect is readily apparent. By cherry-picking data points that were favorable to his narrative, Piketty was able to “show” an increase in inequality over the last three decades. By cherry-picking data points that ran against his narrative even though they came from the exact same source series, I was able to “show” that inequality is actually decreasing in the same period.

In short, Piketty’s assembly of Figure 10.5 demonstrates nothing more than the ability to get its source data to say whatever one desires with enough discretionary manipulation. The product lacks any scientific rigor or interpretive value, except to illustrate the perils of a data assembly process that succumbs entirely to selection biases."

Friday, January 23, 2015

Two examples of how ‘competition breeds competence’

From Mark Perry.
"When government agencies or heavily regulated industries are insulated from market competition, the incentives to offer better service and lower prices, along with the incentives to innovate, upgrade and improve are either significantly weakened or non-existent. But when faced unexpectedly with some market competition, it’s amazing how the normally sclerotic, anti-consumer and unresponsive government agencies or protected industries can suddenly become responsive and consumer-friendly. Here are two examples:

1. The Kelston Toll Road in the UK. I reported last August on CD that an entrepreneurial UK grandfather built a 400-yard private toll in just ten days that allowed drivers to bypass a 14-mile construction detour. A landslide last February closed a road between the towns of Bristol and Bath and construction was originally scheduled to take until last Christmas to complete. The private owner was therefore expecting toll revenue through December to cover his $500,000 in construction and repair costs, along with the cost of staffing a toll both 24 hours each day, and hopefully generate some profit for his entrepreneurial efforts.

But the local government, possibly unhappy with the competition from the private toll road, suddenly made an emergency decision to spend an extra $1 million to speed up the road construction project, which was completed six weeks ahead of schedule in mid-November. Now the toll road entrepreneur and his wife are upset and have accused the local government of trying to bankrupt him with the early opening of the road five weeks ahead of schedule. And perhaps the road construction would have been completed early even without the private toll road, but it seems pretty likely that the presence of competition from the private toll road may have imposed some additional incentives that changed the normal “we don’t care, we don’t have to” attitude of the local civil servants (who often are neither very “civil nor “servile”).

2. Big Taxi vs. Uber. After being protected from competition for generations by government regulations that restrict the number of traditional taxis in most major cities like New York, Chicago and LA, the “taxi cartel” has recently come under competitive pressure from new ride-sharing services like Uber and Lyft that offer consumers a transportation alternative to taxis at lower prices and with better, faster service. Suddenly, the traditional, sleepy taxi industry is being forced to act and think more competitively in response to the upstart ride-sharing services, which is behavior that is completely alien to an industry that never faced the discipline of market competition before. For example, the LA Times is reporting that:
All taxicab drivers in Los Angeles will be required to use mobile apps similar to Uber and Lyft by this summer, according to a measure passed by the Los Angeles Taxicab Commission this week.
The order, passed on a 5-0 vote, requires every driver and cab to sign onto a city-certified “e-hail” app by Aug. 20 or face a $200-a-day fine. The move is seen as a way to make taxicab companies more competitive with rideshare apps such as Uber and Lyft.
Los Angeles cab companies reported a 21% drop in taxi trips in the first half of 2014 compared with the same period the previous year, the steepest drop on record. Cab companies largely attribute the drop to the popularity of app-based ride services.
William Rouse, general manager of Yellow Cab of Los Angeles, says his company has utilized a mobile app for several years. The app, Curb, allows riders to hail and track a cab, provide payment and rate drivers. “If our industry is ever going to get a chance to move passengers from Uber back to taxis, each one of these companies should have an app,” Rouse told The Times. “It’s a shame that the city had to mandate it in order for this to happen.”
Last summer, ABC News reported that:
Meet the new secret weapon to get a leg up in the cutthroat competition among cabbies — charm school. Taxi drivers in Washington state are getting lessons that they hope will give them an edge against startups such as Lyft and Uber. About 170 taxicab operators paid $60 out of their pockets for a four-hour training session to learn about topics including customer satisfaction and developing relationships with institutional clients.
Pretty amazing how the taxi cartel is suddenly starting to change the way it operates now that its drivers are facing intense market competition/discipline from Uber and Lyft.

Bottom Line: Perry’s Law says that “competition breeds competence.” These two cases above help to illustrate that principle, and provide examples of how direct, ruthless, even cutthroat competition is often the most effective form of regulation, and provides the intense discipline that forces firms to maximize their responsiveness to consumers. To maximize the competence of producers and suppliers, we have to maximize competition, and to maximize competition we usually need to reduce the government barriers to market competition like occupational licensing and artificially restricting the number of taxis that are allowed to operate in a city. In other words, we need to move away from the ubiquitous crony capitalism that protects well-organized, well-funded, concentrated groups of producers like the taxi cartel, barbers, funeral home operators, and sugar farmers from market competition. Government regulation typically reduces competition, which then reduces the competence of producers, and reduces their willingness to serve consumers and the public interest, which make us worse off. I say the more market competition the better, for consumers and for the human race. As Bastiat pointed out in 1850:
Treat all economic questions from the viewpoint of the consumer, for the interests of the consumer are the interests of the human race.""

Economist magazine supports vouchers

See America’s new aristocracy. Excerpt:
"Many schools are in the grip of one of the most anti-meritocratic forces in America: the teachers’ unions, which resist any hint that good teaching should be rewarded or bad teachers fired. To fix this, and the scandal of inequitable funding, the system should become both more and less local. Per-pupil funding should be set at the state level and tilted to favour the poor. Dollars should follow pupils, through a big expansion of voucher schemes or charter schools. In this way, good schools that attract more pupils will grow; bad ones will close or be taken over. Unions and their Democratic Party allies will howl, but experiments in cities such as battered New Orleans have shown that school choice works."

Thursday, January 22, 2015

U.S. Sugar Maple Tree Distribution Expands with Warmer Temperatures

By Craig D. Idso of Cato
"One of the major concerns with forecast CO2-induced global warming is temperatures might rise so rapidly that many plant species will be driven to extinction, unable to migrate fast enough toward cooler regions of the planet to keep pace with the projected warming. The prospect of species demise and potential extinction have served as a rallying cry in calls for restricting CO2 emissions. But how much confidence should be placed in this climate-extinction hypothesis? Do real world data support these projections? Are plants really as fragile as model projections make them out to be?

A new paper published in the research journal Botany investigates this topic as it pertains to sugar maple trees, and the findings do not bode well for climate alarmists. In this work, Hart et al. (2014) analyzed “the population dynamics of sugar maple (Acer saccharum Marsh.) trees through the southern portion of their range in eastern North America,” selecting this particular species for this specific task because its range “has been projected to shift significantly northward in accord with changing climatic conditions” by both Prasad et al. (2007) and Matthews et al. (2011).
The three U.S. researchers
analyzed changes in sugar maple basal area, relative frequency, relative density, relative importance values, diameter distributions, and the ratio of sapling biomass to total sugar maple biomass at three spatial positions near the southern boundary of the species’ range using forest inventory data from the USDA Forest Service’s Forest Inventory and Analysis program over a 20-year observation period (1990-2010),” during which time temperatures increased and summer precipitation declined.  
And what did they discover?

Range expansion (!). Hart et al. write that,
In contrast to a contraction of the sugar maple range, our results corroborate the pattern of increased mesophyte (including sugar maple) density and dominance that has been widely reported throughout the Central Hardwood Forest of the eastern US, including sites near the southern range boundary (e.g., Hart and Grissino-Mayer, 2008; Hart et al., 2008; Schweitzer and Dey (2011).
Or put another way, they say the results of their study indicate that (1) “over the past 20 years, the southern range boundary of sugar maple has neither contracted nor expanded,” and that (2) “when accounting for documented northern range boundary shifts (Woodall et al., 2009), these results indicate an expansion of the geographic distribution for sugar maple at this time attributed to the relatively stable southern range boundary.”

Clearly, the rise in temperature and decline in precipitation observed across the study area has had no negative impact on sugar maple populations, despite model projections to the contrary.  Rather, the observed response has been positive, and largely so as evidenced by increased sugar maple density, dominance and range expansion. To most rational people, these observations represent benefits. To climate alarmists, they are problematic, inconvenient truths which they tend not to  acknowledge."

Spending More Money Won't Fix Our Schools

From Megan McArdle. Excerpts:

"American spending on education is in line with that of our peers in the developed world -- a little higher than some, a little lower than others, but not really remarkable either way:

chart 1
That black bar represents total spending, and as you can see, we spend more on education than most of our peers, not less. To be sure, that is partly driven by our very high spending on tertiary education, aka college. But we spend more than most of our peers at most levels, not just on college.

Of course, we're richer than many of our peers, so maybe we should spend more. If you look at spending as a percentage of gross domestic product, we're no longer the highest, we're just average:

chart 2
We spend more than many of our peers on college and late secondary education, less than a few on primary and early secondary school. Perhaps we should reallocate those resources, diverting more into earlier education. But this is not a problem of inadequate overall investment -- and Japan and Switzerland, which spend less than we do, are hardly Third World hellholes.

What about public investment? Is the problem that we don't put enough public funds into education? I find these sorts of arguments rather unconvincing -- the idea seems to be that we should spend more government money on education not because there's a gap we've identified, but simply for the purpose of spending more government money on education. But at any rate, we spend quite a lot of public funds on education at all levels:

chart 3
And when you look at primary, secondary and post-secondary training, we're on par:

chart 4
You can argue that there's an inequality problem in our schools. In fact, I think there is obviously an inequality problem in our schools, but that the big problem is not at the college level, but rather in the primary and secondary schools that are overwhelmingly government-funded. And those disparities are also not primarily about the dollar amounts going into schools -- Detroit spends well above the U.S. average per pupil, and yet one study found that half the population of the city was "functionally illiterate."

"It is just as likely that improvements will come from changing methods and reallocating resources as that they will require us to pour more money into failing institutions."



Wednesday, January 21, 2015

Did Keynesianism Fail In Haiti?

From Cafe Hayek.
"Maybe Keynesianism doesn’t work if the country you try it in has “H” as its first letter. The Economist reports:
FEW countries have suffered an earthquake so devastating, or have been less prepared for such a calamity. The quake that struck Haiti on January 12th 2010 killed perhaps 200,000 people—no one is sure how many—left 1.5m homeless and caused economic damage equivalent to 120% of the country’s GDP. A cholera epidemic compounded the misery. These disasters called forth the biggest-ever outpouring of humanitarian relief, worth some $9.5 billion in the first three years after the quake. The well-wishers vowed, in the words of Bill Clinton, who helped co-ordinate their early efforts, to “build back better”. Yet five years later, the country is little better off than it was before the disaster—and in some ways it is worse.
Worse? After all those broken windows? And that’s with the money and resources to repair the broken windows coming from outside the economy. Perhaps some other factors hurt Haiti in the meanwhile and those factors explain why the aid didn’t help. But I will continue to maintain that destruction is not good for human beings or the economy they interact in. And that the act of spending money doesn’t generate wealth other than for those who receive the money."