Letter to the NYT, etc.
My posting will be relatively infrequent for the next couple of weeks. I just returned from DC, where I finally got a chance to meet John Cochrane. He’s taken a position at the Hoover Institute. Although we disagree on sticky wages, we have uncannily similar views on a wide range of other issues, such as immigration, health care reform and financial system reform. I also got a chance to meet quite a few of the Mercatus scholars, and came away very impressed with the organization.
[BTW, the acronym for the new monetary program is POMP (Program on Monetary Policy.) My enemies are thinking, “I always knew Sumner was a pompous ass.” Or perhaps “pomp” indicates that it will eventually be regarded as the queen of the monetary policy programs.]
Over at Econlog I have a post on Krugman’s recent minimum wage column. Caroline Baum (you’ve probably seen her columns at Bloomberg, and elsewhere) responded to the column with a letter to the NYT. They didn’t print it, so I thought it would be a good idea to post it here. The rest of the column is her letter. I request that commenters be more polite than usual. I don’t mind obnoxious comments, but let’s please treat her as a guest—if you disagree, do so respectfully:
To the Editor:
In his July 15 op-ed, “Liberals and Wages,” Paul Krugman states definitively: “There’s just no evidence that raising the minimum wage costs jobs, at least when the starting point is as low as it is in modern America.” In support of his no-evidence conclusion, he cites a widely discredited 1994 study by economists David Card and Alan B. Krueger.
So flawed was the study – it relied on telephone surveys of fast food restaurants in neighboring counties in New Jersey and Pennsylvania after New Jersey raised its minimum wage – that Card and Krueger were forced to redo it. Using official employment data the second time around instead of a telephone survey, they re-published their findings in 2000, claiming similar results to the first study.
Economists who have reviewed the body of literature on the effect of an increase in the minimum wage have criticized both the methodology and the results of the second Card/Krueger study. David Neumark and William Wascher, both widely respected for their work in the field, cite the vastly different patterns of teenage employment in the two states that pre-dated the study, disqualifying Pennsylvania as a good “control” group. They also find that the depressing effect of a minimum wage hike on employment occurs with a lag, not within Card/Krueger’s short-term time frame. (Neumark and Wascher’s study can be found here: http://www.nber.org/papers/w12663.pdf.)
What’s more, unlike a randomized controlled trial for a new drug, Card and Krueger have no way of measuring what would have happened to fast-food employment in New Jersey absent a minimum wage increase.
It is disingenuous for Mr. Krugman to ignore the wide body of evidence demonstrating that an increase in the minimum wage deprives entry-level workers of an opportunity to enter the workforce. Instead he relies on the findings of an outlier study that contradicts basic economic theory. An increase in the price of any good or service, including labor, results in a decrease in demand for it.
No one will argue with Mr. Krugman’s point that paying workers a higher wage and providing good benefits increases employee loyalty. Businesses choose to do it all the time. Henry Ford didn’t double his workers wages to $5 a day in 1914 because he wanted them to buy Model T’s. He paid his workers more because he wanted to reduce turnover on the assembly line, which proved to be a hard, unappealing line of work.
When the government mandates a floor on wages, many low-margin businesses can’t absorb the higher costs and raise their prices. Even high-margin businesses pass the cost along to their customers.
The New York Times does a disservice to its readers when it allows a Nobel prize winning economist to dissemble to make a political point. Progressive economists may argue in favor of a minimum wage on compassionate grounds, but they all understand the economics of supply and demand. The non-partisan Congressional Budget Office reported last year that raising the federal minimum wage to $10.10 an hour from the current $7.25 would eliminate 500,000 jobs nationwide. (Currently 29 states have minimum wages higher than $7.25.)
And yes, a higher minimum wage is great for those who keep their jobs. But it’s an impediment to those starting out in the workforce.
Mr. Krugman is entitled to his own opinion; after all he writes opinion pieces. But he is not entitled to his own facts. As an opinion writer myself for three decades, my work is always fact-checked for accuracy. Perhaps the Times should make accuracy in support of opinions a priority.
Caroline Baum
West Tisbury, Mass.
Tags:
23. July 2015 at 17:10
Cochrane disagrees with Sumner on sticky wages? I’m with Cochrane’s position.
BTW, OT, a Sumner supporter and recent PhD recently tried to argue using ‘net migration’ that the EU labour market was not that much different from the USA’s (but on gross migration, the proper measure, it clearly is, yet he refused to concede this). But Vox contradicts this person: http://www.vox.com/2015/7/16/8978275/euro-chart-disaster-eurozone
(“Europe has very limited labor mobility compared with, say, the United States. If the economy is strong in the Netherlands but weak in Spain, it’s difficult for Spanish people to simply move to Amsterdam, as they don’t speak Dutch. European countries maintain separate welfare states,and have very different average living standards.”)
Who to believe? the reputable Vox or this just-minted PhD? Who to believe, the eminent Hoover scholar Cochrane, or the “Rush Limbaugh” of economics, Scott Sumner. My money is on quality.
23. July 2015 at 17:13
“What’s more, unlike a randomized controlled trial for a new drug, Card and Krueger have no way of measuring what would have happened to fast-food employment in New Jersey absent a minimum wage increase.”
Has there ever been a randomized controlled experiment with the minimum wage? If this is the standard, then there is no data on what happens when the minimum wage changes and no claims either way should be made. Even *theoretical* claims would be theory unsupported by data.
The basic assumption that allows you to employ supply and demand logic is that workers have sufficient market power to negotiate a large, i.e. o(1), fraction of their marginal product. That’s the implicit model and it does not appear to be remotely true.
Which is better? Game theory or supply and demand?
23. July 2015 at 17:40
I’m not in favor of the minimum wage (at the federal level) but I do find it disingenuous that the same people who cry crocodile tears about minimum wage hikes causing more unemployment will, in a different argument, maintain that so few people actually earn minimum wage that it hardly matters at all.
23. July 2015 at 17:42
Ray,
right, whom to believe.
An actual life long scholar who’s been patiently explaining complex monetary issues on a prolific blog for over half a decade, and who replies to nearly ALL comments point by point.
Or some random dude barking from the sidelines of semi retirement on a beach in the Philippines?
23. July 2015 at 17:56
@mbka – so you concede my Cochrane vs Sumner point then? As for Sadowski, he deleted my comment (or somebody did) at Historinhas (sic) where he guest blogs. Nuff said.
23. July 2015 at 17:57
“The basic assumption that allows you to employ supply and demand logic is that workers have sufficient market power to negotiate a large, i.e. o(1), fraction of their marginal product. That’s the implicit model and it does not appear to be remotely true.”
What would the average return on invested capital be if employees were able to negotiate almost all of their MP? In modern US, it’s tended to run about 7-8% per year.
23. July 2015 at 18:00
“I’m not in favor of the minimum wage (at the federal level) but I do find it disingenuous that the same people who cry crocodile tears about minimum wage hikes causing more unemployment will, in a different argument, maintain that so few people actually earn minimum wage that it hardly matters at all.”
-As Tyler Cowen recently pointed out, both are true:
http://www.bls.gov/opub/ted/2014/ted_20140403.htm
http://www.bls.gov/oes/current/oes_nat.htm
http://www.bls.gov/oes/current/oes_ms.htm
23. July 2015 at 18:37
Jason, You said:
“The basic assumption that allows you to employ supply and demand logic is that workers have sufficient market power”
No, the S&D model assumes no market power, it assumes they are price takers, as are firms.
23. July 2015 at 19:15
Anybody who thinks a study is needed to answer the question of whether a minimum wage costs jobs, is an idiot. I get the political hacksterism behind the debate, but really the only response needed to PK is “Liar!”
23. July 2015 at 20:20
OT – Sumner should take this Chicago Booth quiz, which other economists have answered, on important economic issues. http://www.igmchicago.org/igm-economic-experts-panel/participant-bio-2?id=24
23. July 2015 at 21:02
min wage hikes r bad for a society/species whose last mass-distributed, labor-intensive works will b forms of care and attention to the people least able to afford a hi wage:
pensioners who need diapers changed
children of working folks who need diapers changed
children of unlearned folks who need tutors
24. July 2015 at 00:47
My position is this: if there is a theory that contradicts what seems to be one of the most basic principles in the economics, then there has to be far better argument in favor of such theory to be accepted.
Waving hands and saying that it is because of morals or something else is not sufficient for me. There needs to be more in-depth research to this. Maybe there is some gold hidden there. Maybe government should mandate that all chickens should have more comfort not on a moral grounds but based on economic theory that this increases their morale and they lay more eggs – a fact that farmers cannot realize for any reason.
24. July 2015 at 03:54
Seems to me minimum wage is a politcal issue. Everybody agrees that having a very high minmum wage would cause unemployment. Everybody agree that a small increase would have a very small effect. This is probebly why its so hard to make sense of this.
The Minimum Wage is really a of minor importance compared to immigration and montary policy. Its a shame so much time is wasted on it.
24. July 2015 at 04:29
“There’s just no evidence that raising the minimum wage costs jobs, at least when the starting point is as low as it is in modern America.
How can this be?”
To the extent this is true, isn’t there an obvious Econ 101 answer? The current minimum wage is mercifully low enough that raising it a small amount will have little disemployment effect.
No ‘intellectual revolution’ needed Professor K.
The sad thing is, even $15 an hour won’t have much disemployment effect in high cost-of-living areas, but in the hinterland, ouch.
Puerto Rico would be clobbered.
24. July 2015 at 05:09
I am against minimum wage laws. But why is it never a topic that push-cart vending or sidewalk retailing is outlawed in almost every city in the United States and robust housing construction is criminalized?
24. July 2015 at 05:36
Is Krugman right that there’s NO evidence that the minimum wage costs jobs at current levels? I had always thought there were some studies that found an effect but that pro-minimum wage advocates liked the fast food studies better?
24. July 2015 at 05:39
Good luck in your new position Scott.
And I like Caroline Baum. See no reason to be impolite.
24. July 2015 at 06:39
What about: the evidence is mixed, so let’s look at other ways to help the poor.
24. July 2015 at 06:59
Prof. Cochrane has a post describing his testimony on monetary issues. It seems congress is willing (at least on paper) to include NGDP targeting as an alternative policy regime. Congrats Prof. Sumner
http://johnhcochrane.blogspot.com.br/2015/07/monetary-testimony.html
24. July 2015 at 08:24
I do agree that articles should be fact checked and economists should actively try not to fool themselves with confirmation bias, and perhaps Krugman is guilty of that. However, when I read this:
“Instead he relies on the findings of an outlier study that contradicts basic economic theory.”
I have to ask, does anything in the field of economics at this point in history deserve the title “theory?” Other fields have falsifiable hypothesis and theories, and there are many examples of falsified hypothesis and theories to demonstrate that “falsified” is not just window dressing. But is this the case in economics? Does economics ever give a failing grade to one of their so-called “theories?” Is there a discarded economic theory that is recognized by at least 99% of all current economists to have been falsified? What’s an example?
Also, regarding minimum wages in particular, Kenneth Duda liked this recent post on the subject.
24. July 2015 at 09:40
The minimum wage is going up, in some places, and not going up in others. We’re all going to see the results. Why write letters to the NYT (or columns) implying you already know what’s going to happen?
Is this experiment too disastrous to even consider just letting it evolve?
24. July 2015 at 09:44
I agree with Bill Gates on this issue…I believe he recently said that this will greatly accelerate automation (robotics). So yes I think it will reduce jobs. We are at a beginning of a robotics revolution that is going to transform the world. I am sure Krugman will then move on to his “robotics are creating robber barons” posts
24. July 2015 at 09:57
Good letter from Caroline. If there’s anything we economists should be able to agree on, it’s supply and demand. If humanitarian benefits outweigh the cost of lost labor hours, fine, but to ignore those lost labor hours is simply dishonest.
24. July 2015 at 10:00
Minimum wage is a political issue. I can’t contribute much to the economic arguments about whether it has a large effect on unemployment or a small one, but, as I’ve said before, why do you ignore the elephant in the room?
The effect on unemployment is preconditioned on no reaction from the Federal Reserve. The Fed targets unemployment as one of its mandates. Unemployment should not move under this assumption when raising minimum wages. This is especially true when inflation is below target.
The political question, once you recognize this, is whether you want to redistribute real earnings to the low income earners with a slightly higher inflation as the cost, not whether you want more unemployment. Some of us say yes, and we’re wiling to live with slightly higher inflation to do this.
24. July 2015 at 10:35
The question of whether a rise in unemployment is due to a higher minimum wage is meaningless. It’s similar to your admonition to never argue from a price change.
low wage employers could reduce employment for many reasons. A rise in wages would normally be accompanied by a rise in their prices. Only if that causes a move by consumers to substitute will employers be forced to invest in capital to reduce labor costs.
But what are these consumer chosen substitutes? It’s just as likely they are low wage employers. Their prices would in general have risen as well. And now with more customers the number of employees will rise as well.
So, unless demand falls generally (due to a prioritization of inflation reduction over unemployment reduction by the Fed) the decision to raise wages causes an adjustment (including investment in new capital) that is complex and, I believe, ultimately beneficial.
It’s very simplistic to just throw demand/supply at this while not pointing out the complexity due to the reaction of the important players. Only real world measurements can shed light on the impact and even that does not give the whole answer which should include the counterfactual that assumes the Fed would respond to prop up demand.
24. July 2015 at 10:36
I agree with Bill Gates on this issue…I believe he recently said that this will greatly accelerate automation (robotics). So yes I think it will reduce jobs.
You don’t need robots.
A good bunch of years ago I was the chair of a non-profit that owned a small commercial building in Manhattan as an endowment. We employed typically three or four minimum wage youths (mostly students from NYU) overnight to answer the door, take messages, watch the place, sweep and clean up for the morning, etc. Then when NYC upped its minimum wage and added a bunch of regulatory requirements we cut back to one or two plus a vacuum cleaner, burglar alarm and an answering machine added to a phone line.
Perhaps also worth noting about our real-world experience amid all the noise and argument of the current moment…
* We didn’t fire anybody, just waited for turnover. So all surveys of all involved would have accurately reported “Nobody fired due to minimum wage increase”.
* The minimum wage for us was a “learning wage”, anybody who learned on the job and showed competence quickly got a raise to well above it, those who weren’t competent and didn’t learn fairly soon became turnover.
* These kids *needed to learn*. Even though they were almost all NYU college students and some very good academically, most had no clue how to work on a job. They’d decide on their own to show up late figuring they could make it up on their own later, disappear without explanation, were condescending to visitors and customers … but because they were minimum wage we *taught* them instead of firing them, and most of them did learn.
The greatest example was one guy who rented out rooms to Law & Order to film in and pocketed all the money for himself. We’d never have known except he made a “gift” to us of 10% of it. To our “WTF?” he replied that he’d earned all the money because he had answered the door when L&O knocked, he had made the deal with them, the rooms would have been unused if he hadn’t found this use for them, etc., so the money was all rightfully his. That earned him a good dose of ‘splaining on the concepts of “employee” and “other people’s property”. But we kept him on because we liked him, he did learn, and eventually he became our head of staff who ran and taught all the other new hires.
Although if we’d been paying him twice as much we’d have fired him in five seconds.
24. July 2015 at 10:51
Elwailly has it right. Just as legislative decisions that impact AD (fiscal policy) is offset by the Fed, aren’t legislative decisions that impact unemployment offset by the Fed as well? At least under a dual mandate regime, one would expect that if a policy such as higher (nominal) minimum wage law were passed, then the Fed would start to see unemployment rise, and would cause higher inflation to push it back down again.
Certainly we can all agree that at a higher price level commensurate with no change to the real minimum wage, there would be no effect on unemployment, and it is tough to see why the Fed wouldn’t do just that.
24. July 2015 at 12:01
Following up on Jim Glass’s story;
http://bloooo.com/2015/07/thomas-sowell-confessions-of-an-exploited-teen/
As Sowell concludes, Today–with a minimum wage high enough to get the attention of employers–you don’t see many ‘exploited’ teens in his old neighborhood. Instead they’re standing around on street corners, often getting into trouble. But the humanitarians feel better.
24. July 2015 at 12:05
And McDonalds, concerned about lower profits, is experimenting with kiosks at which customers can bypass cashiers and order cheeseburgers themselves. Just by pushing buttons.
But Paul Krugman probably doesn’t eat at McDonalds.
24. July 2015 at 12:47
@Elwailly, Ninnja,
What you are describing is a roundabout way of increasing nominal wages at the low-end without increasing real wages. This will not increase the purchasing power of minimum wage workers. So… what are you trying to accomplish here?
24. July 2015 at 14:56
Now here is one tough broad;
http://foreignpolicy.com/2015/07/24/venezuelas-marked-woman/?utm_source=Sailthru&utm_medium=email&utm_term=*Editors%20Picks&utm_campaign=2014_EditorsPicks%20SO%20July%2024
———–quote———-
Congresswoman Maria Corina Machado…is used to taking punches as part of her job “” literally. Two years ago, a pro-government legislator broke her nose in the floor of the legislature. The assault went unpunished.
After meeting then-President George W. Bush in the Oval Office “” the only member of Venezuela’s opposition ever to have such an audience “” Machado was charged with treason. (She was subsequently pardoned.) More recently, she has been charged with attempting to kill President Nicolás Maduro, and a judge has barred her from leaving the country.
Her emails have been hacked and, according to independent experts, altered. Her phones are tapped, and her private conversations are frequently aired on state-run TV. Her family is constantly hounded, and mobs of government activists attack her public appearances. After being elected with the most votes of any of her colleagues, she was unceremoniously kicked out of parliament. It is fair to say that the government is obsessed with Machado.
———-endquote———-
And she talks like Maggie Thatcher;
———-quote———
The Venezuelan state, according to Machado, has to downsize “” both in terms of GDP and in terms of the number of people it employs. “The state,” she told me, “oppresses Venezuelans. It needs to be a provider of public goods, and it needs to stop interfering with Venezuelans’ desire to live freely.” In a country where the state monopoly in the oil industry is taken by many as a dogma, she defends private participation in the industry.
…. Machado is clear: [price] controls must be lifted as soon as possible while implementing measures to protect vulnerable families during the transition. She also believes the price of gasoline “” the lowest in the world “” has to be raised, although she would oppose the move if it were implemented by this government. Machado claims it would be immoral for the current government to raise the price of gas when it gives away tens of thousands of barrels of oil per day to Cuba for free, a policy everyone in the opposition rejects.
————endquote———
She’d eat Hillary for breakfast.
24. July 2015 at 18:02
Caroline Baum wrote: “Perhaps the Times should make accuracy in support of opinions a priority.”
The NYT doesn’t bother fact-checking news stories. Why would it waste resources doing the same on opinion pieces?
24. July 2015 at 20:44
@ Brian Donohue
“What you are describing is a roundabout way of increasing nominal wages at the low-end without increasing real wages.”
Not true. Since everyone else’s income would not change, the adjustment in inflation to keep demand level would be much less than that required to keep real low wages from rising.
Raising minimum wages is a straight forward transfer of wealth to the low end. It’s a much healthier way to do it than taxing the high end and giving money directly to those the government thinks are needy enough. Raising the minimum wage rewards hard work and has the right incentives built-in.
24. July 2015 at 21:52
“An increase in the price of any good or service, including labor, results in a decrease in demand for it.”
Ah jeez…
24. July 2015 at 23:53
Caroline Baum wrote: “Perhaps the Times should make accuracy in support of opinions a priority.”
This is nothing new. Daniel Okrent, the Times’ first Public Editor, had the same suggestion 10 years ago when he wrote his final column in that role. (As an aside, if the Public Editor is truly independent, one wonders why the hard hitting stuff was left for the last column).
Krugman was one of the usual suspects:
“2. Op-Ed columnist Paul Krugman has the disturbing habit of shaping, slicing and selectively citing numbers in a fashion that pleases his acolytes but leaves him open to substantive assaults. Maureen Dowd was still writing that Alberto R. Gonzales “called the Geneva Conventions ‘quaint’ ” nearly two months after a correction in the news pages noted that Gonzales had specifically applied the term to Geneva provisions about commissary privileges, athletic uniforms and scientific instruments. Before his retirement in January, William Safire vexed me with his chronic assertion of clear links between Al Qaeda and Saddam Hussein, based on evidence only he seemed to possess.
No one deserves the personal vituperation that regularly comes Dowd’s way, and some of Krugman’s enemies are every bit as ideological (and consequently unfair) as he is. But that doesn’t mean that their boss, publisher Arthur O. Sulzberger Jr., shouldn’t hold his columnists to higher standards.”
http://www.nytimes.com/2005/05/22/weekinreview/13-things-i-meant-to-write-about-but-never-did.html
Krugman didn’t like that one bit.
Subsequent public editors have shown even less courage in confronting the issue. Don’t expect anything to change. Exasperated with the lack of factual accuracy and objectivity, particularly, but not solely on the editorial side, I cancelled my subscription to that paper about a year ago. The only thing I miss is the fun of dissecting the logical and rhetorical sleights of hand that are routinely used to support their ideological and political biases. Of course, other papers do it too, which is why most of them function as organs supporting the politics of one party or the other.
25. July 2015 at 06:31
Vivian Darkbloom
+1
Last line of public editor comment was best:
“I didn’t give Krugman, Dowd or Safire the chance to respond before writing the last two paragraphs. I decided to impersonate an opinion columnist.“
25. July 2015 at 06:35
As for Scott, it’s way better to be a member of the
Program On Monetary Policy
than a member of the
Aggressive Stimulus Spending Expansion Society
Although if you want to join both, be my guest 😉
25. July 2015 at 07:42
Yes, Vivian, I remember that well. Back then amid all that dispute (largely Krugman disputes) the Times editors made it very clear that they had a policy of no fact checking at all for their in-house OpEd writers, would stick to it. Unlike the WaPo.
(And so utterly unlike the way they treated their ‘guest’ OpEd contributors at the same time, as the great Boris Johnson very amusingly related )
But to give the Times’ editors their due, Krugman later recalled that they did instruct him to stop using the word “liar” so much.
My personal favorite example of this policy in action (because it was such a whopper, and because I had the fun of initiating the following brouhaha) was when Krugman wrote that the revenue lost to the Bush tax cuts was, quote: “more than enough to ‘top up’ Social Security and Medicare, allowing them to operate without benefit cuts for the next 75 years.”
And of course he couldn’t stop there, but went on the denounce the lazy press for pretending to be “balanced” and “non-shrill” in repeating all the false stories from the right that had been around for so long to the contrary, that entitlement funding was on some sort of unsustainable path.
Ponder that for a moment … Krugman was totally ignorant of all the official analyses of the subject by by CBO, GAO, the Medicare Trustees, the SS trustees, & Treasury, so he assumed the entire national press corps was too. There is psychology in action.
Well, in dollar terms this was a mere 14-digit error at present value. In very much appears that what PK had done was read the abstract of a paper then rush out before reading the body to use it to bash Bush thusly, without realizing the body referred only to the payroll tax funding component of Medicare Part A.
I pointed this out in comments on DeLong’s blog, and a few other blogs picked it up, and rather print a correction PK, well, just didn’t like it at all. He threw up a response on his web site that denied any error, then said it was the study’s error not his, then … well, enjoy the weaseling yourself … concluding that apologies were owed *to him*, and with words “Dishonesty, dishonesty everywhere“, among those who wanted a correction — or even a mere clarification.
Geeze, however did Okrent form his opinion of PK? 🙂
Anyhow, that’s as close as NY Times OpEders are required to come to admitting 14-digit errors of fact.
25. July 2015 at 09:17
What is the plan for showing the market prediction for 2016 NGDP? It seems like that prediction is also meaningful now too. Thanks.
25. July 2015 at 14:32
Ray Lopez,
“Who to believe? the reputable Vox or this just-minted PhD?”
Only Ray Lopez could believe that a popular news article written by a reporter working for a internet media company would totally trump doctoral training in the subject.
“As for Sadowski, he deleted my comment (or somebody did) at Historinhas (sic) where he guest blogs. Nuff said.”
I don’t have the ability to delete comments at Historinhas, I just guest blog at Marcus Nunes’ request. If Marcus deleted your comment (I haven’t bothered to asked him if he did) then I’m sure he had every reason to do so.
Moreover, I’m absolutely sure that nobody would be devastated that they might be unable to read a deleted Ray Lopez comment, as even Major Freedom has taken time out to point out that some of your comments are logically incoherent.
25. July 2015 at 16:53
“Instead he relies on the findings of an outlier study that contradicts basic economic theory.”
An incredible combo of two poorly informed things in one sentence.
1. Card-Krueger is hardly an “outlier study.” Numerous other studies have found near-zero after near-zero for the impact of minimum wage on employment. Take a look at the funnel plot in http://econbrowser.com/archives/2014/03/faith-and-econometrics-minimum-wage-edition for example.
2. It contradicts “basic theory” if your only exposure to “basic theory” comes from a shoddy GMU education. There is plenty of theory that explains such results. Daniel Kuehn gave some examples in a blog post here: http://factsandotherstubbornthings.blogspot.com/2013/02/some-thoughts-on-minimum-wage-theory.html
Scott, why is it that you and the rest of the Econlog crowd continue to completely suspend any form of questioning logic when it comes to something against your political views? Do you really think everyone is as partisan as you? Do you really think that all those other economists just saw one study from 1994 and are taking it on blind faith? These days Card refuses to discuss the literature because he hated the politics of it.
25. July 2015 at 18:12
The usual suspect, I’m always puzzled when someone provides a quote that I did not say, and then uses it to attack my views.
I don’t think Krugman is as partisan as I. He’s far more partisan. Now he almost always supports the Dems (unlike in the 1990s when he was more nonpartisan.). In contrast, I despise the two parties equally.
As far as my political views, do you mean utilitarianism? Yes, I oppose any public policy that goes against my political views.
I did not get my education at GMU, I got it from what was at the time the number one ranked economics program in the world. However today GMU would probably be my first choice,
If you want to criticize me, why not respond to my Econlog post?
25. July 2015 at 18:52
@Mark A. Sadowski – thanks for taking timeout from your obviously busy life to respond to me. Vox’s editor is M. Yglesias, and the piece in question was written by T. Lee. While both are indeed journalists, Lee, Yglesias > Sadowski, sorry to deflate your ego. They fact check their stuff, don’t have an ax to grind, you clearly do. That you refuse to concede that two countries / regions that exchange 10M (or less say 100M to be more extreme) people every year, looking for new jobs, is more dynamic in job creation than two countries/regions that exchange ten people, is clear evidence of your bias, ideology and sheer stupidity. Thanks for clarifying that Nunes perhaps was the censor of my post. My post btw was innocent, along the lines that money neutrality makes monetarism a farce. And finally I find it amusing that you side with Major Freedom. Politics make for strange bedfellows, not that there’s anything wrong with that. Just be honest with yourself and come out of the closet, concede you are a polemicist advocating something. You are not practicing the dismal science but political economy, the most apt description of your sordid field.
@ssumner – interesting that to avoid ‘The usual suspect”s piquant argument, you say: “I got it [my econ PhD] from what was at the time the number one ranked economics program in the world [Chicago]. However today GMU would probably be my first choice” – wow! Since when is GMU > Chicago? The things you will say to please your new employer! Reminds me of the joke of the economist job interviewee who tells the interviewer, ‘2 + 2 = ? what do you want it to equal? I’ll endorse that answer’. No integrity.
25. July 2015 at 20:20
The usual suspect: Meta-regressions like this amount to quantified groupthink and tend to incorporate gross methodological errors that may have a systemic bias. Engineers and physicists roll their eyes at any significance attached to this kind of analysis; science demands correct answers, not statistically popular ones, and they’re often very difficult to get even from randomized, controlled trials.
There are some good arguments for minimum wage hikes (such as the high MTRs for the poor created by welfare) but if your political ideology requires you to give up supply and demand you have to realize how you’ve prioritized your beliefs.
25. July 2015 at 21:40
I think I’ve almost converted on the euro as being a useful exercise, maybe its the only way to save the welfare states — at some point Syriza must have realized there was simply no money left and capitulated rather than accept that. I predicted socialists couldn’t leave money on the table but I do wonder if they will come to the next negotiating table with printing presses behind them.
But I have to admit there’s a real possibility they’ll actually enact some useful reforms and hew their way to long-term solvency.
Now if only the euro came with better ECB policy…
26. July 2015 at 07:13
Want to thank the usual suspect for reminding me of one of favorite econbrowser posts.
http://econbrowser.com/archives/2014/03/faith-and-econometrics-minimum-wage-edition
Fun times in a comments section, indeed. Say, Rick Stryker’s riposte;
———quote———
I’m pretty surprised that you would post a funnel plot as evidence that the elasticity is near zero.
The funnel plot is not a model selection tool. It’s a diagnostic to detect publication bias. But you’ve used it to make the argument that estimates should be selected with the smallest standard error.
As you well know, that’s wrong, since the underlying models in the studies might be misspecified in some way. For example, if you omit a relevant regressor you bias the other estimates. If you have measurement error in the regressors that is correlated with the error term, the estimates are biased, even asymptotically. Small standard errors aren’t necessarily good if the estimator is biased. And uncorrected heteroskedasticity and serial correlation invalidates inference.
I could go and and on. But the point is that you have to look at each study and judge it. You can’t just pull out standard errors and make a plot.
No one has really done a serious meta study of the minimum wage literature that way it is supposed to be done. I already commented before on how the meta study you cited above does not meet the standard. Maybe somebody will do that one of these days. Meanwhile, we do have expert narrative review that indicates that elasticities are typically between -0.1 and -0.3, as I mentioned.
———endquote——–
Which was followed by even more specific refutations of Menzie’s disingenuous methods. You did read those comments?
26. July 2015 at 07:45
Caroline Baum (and Scott of course),
So #Uber4Welfare https://medium.com/@morganwarstler/guaranteed-income-choose-your-boss-1d068ac5a205
Is my wage subsidy plan which uses non-medical welfare spending and search matching to increase consumption by 30%+ for the poor (without new taxes).
This is the part of the minimum wage discussion that I think Economists miss and is far better than hammer Krugman’s feet to the floor.
1. Obviously, a MW set about the ROI that a laborer can deliver means that he doesn’t get to work.
2. But the more compelling flip side is that his labor doesn’t get to be consumed.
And when Econos band away on 1, they don’t recognize that the mental response of liberals and guys like Tyler is first between “good for him!” and “he’s not worth much anyway, maybe even negative for the firm.”
Tyler is wrong of course. There’s virtually no able minded, able bodied human being at $40 per week (with a $240 wage subsidy) that I as an end buyer can’t find a use for. Perhaps in a firm, Tyler might be right, other at one workers might be affected BUT if instead hundreds of different self interested buyers are looking at this person thinking: can he weed my garden, pick up dog sh*t, spin a sign on the street, clean up my garage – once greed for super cheap subsidized labor takes over, WHOEVER finds a ROI use for this guy gets to keep the cream, but overtime, that cream is proven and the price of it rises.
The other part that I think Econos miss when dealing with a Krugman, is that POOR SERVICE POOR.
They live and work amongst each other… this has two really harmful effects, liberals do not notice:
1. The price level in poor areas is kept over-inflated. The poor are forced to consume less.
2. Minority entrepreneurs, the natural hustlers born into deep poverty, are denied their birthright of CHEAP LABOR as a real advantage to go out in the world and save their race. Other nations compete on cheap labor. Historically new minority groups arriving in the US and made their gains when the smartest among them, round them up and put them to work and use that advantage to grow rich – which serves the group.
Taken together, I’d encourage you guys to take a new tack on Krugman et al, and make the first clear point:
Someone who can cut hair profitably at $3 per hour but not $4, being told to not work, reduces the total number of haircuts sold in poor neighborhoods, and reduces the consumption of the poor. And the guy who could run the cheapest hair Salon in town, isn’t able to steal business away from a white guy, luring his own customers to the black side of won.
YEs, I’m aware that there would EVENTUALLY be an equilibrium found, but:
1. during that time, wealth flows to minorities.
2. if you believe that Americans are racists, like Krugman et al does, than irrational RESISTANCE to going to the cheap barber on the bad side of town – prolongs the natural labor price advantage of minorities.
The punch line is this: we often times as Libertarians argue that racism is irrational, and naturally greed and self-interest for consumption will overwhelm bias in people… and we’re right!
BUT Minimum wage laws priced even marginally above the natural clearing rate, serve to REWARD the racist, bc he doesn’t feel the loss of consumption relative tot he non-racists.
26. July 2015 at 09:14
Ray Lopez,
“Vox’s editor is M. Yglesias, and the piece in question was written by T. Lee. While both are indeed journalists, Lee, Yglesias > Sadowski, sorry to deflate your ego. They fact check their stuff, don’t have an ax to grind, you clearly do.”
While I read and enjoy Vox, they occasionally get things wrong. Take for example this article by Matt Yglesias where he claimed that Iceland had rejected fiscal austerity:
http://www.vox.com/2015/6/9/8751267/iceland-capital-controls
I happened to know this was way off base in part because of some things I was researching for Scott in my capacity as Primary Research Assistant for POMP at the Mercatus Center. Hence I wrote the following post:
https://thefaintofheart.wordpress.com/2015/06/12/is-iceland-krugmans-inadvertent-case-for-the-monetary-policy-offset-of-fiscal-policy/
(continued)
26. July 2015 at 09:18
(continued)
That post was picked up by Scott:
http://www.themoneyillusion.com/?p=29606
And then by Tyler Cowen:
http://marginalrevolution.com/marginalrevolution/2015/06/did-iceland-reject-fiscal-austerity.html
(continued)
26. July 2015 at 09:25
(continued)
And then by Robert Murphy:
https://mises.ca/posts/blog/krugman-and-yglesias-get-tough-with-the-bankers-after-it-doesnt-matter/
And then by Russ Roberts:
http://cafehayek.com/2015/06/krugman-and-other-stuff.html
(continued)
26. July 2015 at 09:29
(continued)
And by Newsweek Europe:
http://europe.newsweek.com/syriza-split-looms-left-wing-propose-grexit-icelandic-option-328732
And by the Washington Post blog:
http://www.washingtonpost.com/blogs/wonkblog/wp/2015/06/17/the-miraculous-story-of-iceland/
(continued)
26. July 2015 at 10:10
Ray Lopez,
“Vox’s editor is M. Yglesias, and the piece in question was written by T. Lee. While both are indeed journalists … They fact check their stuff, don’t have an ax to grind…
Yeah, that’s right!.
26. July 2015 at 10:15
(continued)
[There’s comment held up because of one too many links to a few of the European blogs that linked to that post.]
The point is that occasionally bloggers either misrepresent the facts, or they simply get their facts spectacularly wrong. And when that happens, it opens up easy opportunities for me.
So you have this precisely backwards (as usual). It’s bloggers like Krugman and Yglesias that have an axe to grind. If and when I get widely read (quality counts more than quantity), it’s usually when I catch people misrepresenting the facts.
Now Ray, if you don’t mind, I have bigger fish to fry. I can assure you that nobody is linking to your comments, so it’s not really worth my time to check your “facts”.
26. July 2015 at 13:27
“Vox’s editor is M. Yglesias…”
Not really. Vox has a number of “editors”, but the editor-in-chief is Ezra Klein. M. Yglesias is *an* editor, but not the only one, and it’s far from certain if he had any role in “fact checking” the column in question.
http://www.vox.com/masthead
26. July 2015 at 15:54
I’d distance yourself…you should already know.
Good to see your ever expanding influence. Money matters. 7/22/15 played out. Contraction, because of a payment inversion, is now a little deeper.
26. July 2015 at 16:19
@Sadowski – thanks I’ll check these links later. I will grant you that you are persistent, that Vox gets things wrong and you get things right, but on the issue of labor mobility and sclerotic labor markets in the EU vs the more dynamic USA, you are clearly wrong. As for my post upstream, it was a brilliant albeit nasty piece of esoteric writing that was more rhetoric than fact. And I get read as much as the main article. Everybody knows the comments are often read more than the article.
And please don’t hate me because I believe in money neutrality; even the great Fisher Black did.
26. July 2015 at 16:24
Patrick Sullivan-
I always wonder why no one looks at minimum wage at the firm level. If profits are high per employee-hour, it’s probably not a big deal, but if profits are low per employee-hour, something will change: shorter hours or lower service quality.
The pricing power of the firm (or the industry, since minimum wage affects all players) matters too. Since the minimum wage is an indirect tax on the consumer, it matters if consumers willing pay more, or swap into less valuable services, e.g., grocery or junk food instead of fast food.
26. July 2015 at 17:39
Mark, I’m afraid that trying to explain economic concepts to Ray is like trying to teach quantum mechanics to a cow. It’s pointless (unless, like me, you enjoy making fun of him.)
26. July 2015 at 19:53
Mark Sadowski, yeoman work in a lost cause. We can hope that at least someone learned something in there.
Vox is hit or miss, they write some nice articles but other times do crazy things like complain about Israel restricting traffic on an imaginary bridge from Gaza to the West Bank.
Funny, those Hayek-Keynes rap videos from Russ Roberts are still classics, though now of course I realize they promulgate some shortcomings of the Austrian perspective on monetary policy. Haha, maybe someday we can throw you and Sumner into a third episode.
26. July 2015 at 20:15
Adam Ozimek: “What Liberals Get Right About Poverty”
http://www.forbes.com/sites/modeledbehavior/2015/07/26/liberals-poverty
26. July 2015 at 23:03
Ray, a neutral money is a contradiction in terms.
You believe in a fallacy.
You don’t have to depend on the doctrine of neutral money in order to reject the other doctrine that inflation grows economies.
Read this essay:
https://mises.org/library/fallacy-superneutrality-money
27. July 2015 at 05:34
“I’m not in favor of the minimum wage (at the federal level) but I do find it disingenuous that the same people who cry crocodile tears about minimum wage hikes causing more unemployment will, in a different argument, maintain that so few people actually earn minimum wage that it hardly matters at all.”
The lack of large numbers of people working at the minimum wage suggests that jobs that pay less than the minimum are in fact often destroyed, for it they were not, we’d expect all those people who would otherwise be making less to be earning exactly the minimum.
27. July 2015 at 05:42
Lars Christensen: “The market’s message to Yellen: You have become too hawkish”
http://marketmonetarist.com/2015/07/27/the-markets-message-to-yellen-you-have-become-too-hawkish
27. July 2015 at 06:35
@Brian Donohue
It’s not about what I am trying to accomplish. I’m merely pointing out that if the Fed is effectively targeting unemployment at some particular rate, then there is a monetary offset to the minimum wage debate that needs to be considered (just like a monetary offset needs to be considered in debates about fiscal policy). So if you want to increase the wages of minimum wage employees, and believe that labor demand slopes downward, and believe that the Fed targets unemployment, then increasing the minimum wage will do nothing for real wages, and therefore that person should not support an increase in the minimum wage, *even if that person was willing to accept the compromise of higher unemployment for higher wages*
Note that this also applies to academic studies. Just as studies of the “fiscal multiplier” are all over the place because they don’t understand that the fiscal multiplier is zero if the Fed wants it to be, studies of the effect of minimum wage on unemployment are all over the place because they don’t account for the fact that the Fed will actively work to reduce whatever impact there actually is on unemployment. At the very least it will make intertemporal comparisons very difficult.
27. July 2015 at 07:33
Great stuff by Lorenzo! “The Rotten Heart of Europe”
http://lorenzo-thinkingoutaloud.blogspot.com/2015/07/the-rotten-heart-of-europe.html
27. July 2015 at 07:35
In other French news: “French farmers angry over low prices turned back hundreds of trucks at the German border on Monday, looking for cargos of foreign meat and milk products.
An Associated Press photographer at the German frontier saw farmers stopping refrigerated trucks to verify their contents on Monday, and one of the protest’s organizers said 300 trucks had been turned back since the morning. Other vehicles were allowed to cross freely.
Police in France tend to avoid intervening in peaceful protests, and French President Francois Hollande on Monday said he backed the farmers and called for a high-level meeting of European agricultural officials.
“Between now and then, we will continue to pressure, so that the farmers are certain, protests or not, that we are at their side,” he said.”
http://wapo.st/1MvOupS
27. July 2015 at 10:03
@MF – Thanks for the link. The Mises Daily site says: “Further, an injection of new money necessarily causes a change in relative prices, and this, in turn, affects the production of consumption and investment goods. Any change in the money supply “” be it free-market money or government-controlled money “” will have the implication outlined above.” – if this is true, then the Austrians have conceded Monetarism works. But no econometrics study has shown it works.
@Sumner – show me the econometric study that shows monetarism works. BTW forget cows and quantum mechanics, a recent study found plants ‘understand’ (and use) quantum gravity during photosynthesis. They’re not as dumb as you and your research assistant Sadowski are, who thinks 10M people crossing a border looking for work is the same as ten.
27. July 2015 at 15:26
Ray,
I knew it.
The reason you repeatedly say you believe in neutral money is to convince yourself of a reason to reject Monetarism.
Like I said, money neutrality is NOT a prerequisite to Monetarism being a destructive ideology. In fact, one of the reasons Monetarism is destructive is precisely because money is not neutral. The absence of money neutrality is related to the destruction it unleashes.
Or, to put it a different way, the reason Monetarism fails, doesn’t work, destroys rather than constructs, is on account of non-market driven money non-neutrality, as opposed to market driven money non-neutrality.
Money being neutral is in part what enables it to serve as a means for poeple to coordinate their activities.
It is precisely money non-neutrality that “drives” resources and labor going to their most urgent uses in a free market. Non-neutrality of money is what allows money to serve as a meana of exchange that minimizes waste.
Or, to put it in a still different way, if money were neutral, then it could not serve as a means of exchange.
Ray, for Pete’s sake, you’re wrong on this. The fact that I actually agree with Sumner on this should be a clue that it is so trivial that even ideological opponents can agree on it.
Agreeing that money is not neutral does NOT require you to “admit” or “concede” that “Monetarism works”. It is precisely the opposite! Arguing that inflation destroys economic growth, PRESUPPOSES that money is non-neutrality.
Argh, the whole concept of non-neutrality is an Orwellian word game. By attempting to convince people that money is “non-neutral”, this gives tacit credence that the frame of reference, the standard, is money neutrality and that the only alternatives can be what is not that. Sort of like going to a westernized Indian restaurant and seeing “vegetarian” and “non-vegetarian” categories of food on the menu.
Perhaps a more honest phrase would be “Money is affective”
Money is not the sole driver of production, but it is a crucial driver. The reason why inflation destroys capitalist expansion, is because of money effectiveness turning destructive rather than constructive.
If history is a guide, then the dollar collapses.
27. July 2015 at 15:27
Typo:
Meant to write:
“Money being NON neutral is in part what enables it…”
27. July 2015 at 15:44
Monetarism is at root a frightened child.
“It is better to be certain now without planning, regardless of future destruction, then to be uncertain now and plan for an uncertain future.”
It would be quite “pragmatic” for a frightened child to pretend that an adult, any adult really, is to be trusted, obeyed and respected not by merit, but by status quo and tradition.
Monetarists just haven’t learned that the parent-child structure ought not be imposed on society as a whole. But they just don’t know any better. It is what they have come to accept on faith.
At root, most of them feel antagonistic towards libertarianism because they consider it an attack on the parent-child structure, and, since Monetarists are mostly childlike thinkers, they identify themselves as the children, the state as their parents (Mommy is Mrs. Democrat and Daddy is Mr. Republican), and thus Libertarianism as an attack on their own parents, at some subconscious level.
This is why I have hatred and vortiol spewed at me from the child like thinkers on this blog. They view me as a destroyer of families, of their parents, of wanting a sort of Lord of the Flies children only anarchy.
How irresponsible isn’t it? Grown men ruling themselves and no other. Where are the parents?
27. July 2015 at 16:40
Ray Lopez,
Marcus informs me your comment wasn’t deleted. Evidently you just forgot where you commented:
https://thefaintofheart.wordpress.com/2015/07/06/the-monetary-base-and-the-bond-yield-channel-of-monetary-transmission-in-the-age-of-zirp/
Sounds like a slight touch of senile dementia to me.
27. July 2015 at 17:02
@MF – you say: “Arguing that inflation destroys economic growth, PRESUPPOSES that money is non-neutrality.” – I would take issue with “destroy”. I agree there are search costs and menu costs associated with high inflation–let’s define that as in Brazil, around 15% a year for several generations–but that these search costs are not that great. Brazil actually grew (in real terms) with such high inflation for nearly 40 years after WWII. Argentina also had inflation but did not grow as much. So inflation is a bit overrated. Hyperinflation however is not, that much we can agree on. Put another way: is the battle over Austrians and Monetarists over menu costs? Over shoe leather costs? If so, it’s much ado about nothing.
@Sadowski: thanks. A slight touch of dementia I doubt; I have an IQ says one test of 120. One-twenty! That’s in the upper echelon. It used to be 140 so one or both tests are wrong, or I’m getting dumber as I age; further research is needed.
27. July 2015 at 17:16
@Mark A. Sadowski – could not log onto Nunes’ site to make a reply to your comment. Graphics heavy site that uses Javascript, and slow internet in the Third World where I live. If you read this see the last paragraph. Bye.
@Mark A. Sadowski [re Nunes’ site blog post]
– “data mining” in my view is pejorative.
“Thirty observations is the usual minimum to be considered a large sample, although more is usually preferred” – correct, I think 5% error rate comes from this size sample for normal distributions. I use this sample when calculating my chess Elo vs calibrated PCs.
Re 4) – data mining: price of butter in Bangladesh correlates with world GDP. Also see this site, on publication bias but the same principle applies: http://www.jerrydallal.com/LHSP/multtest.htm
re 5)- I think you misunderstood, I am simply saying that you are measuring a routine business expansion, and not that the Fed creates such an expansion short term, which your paper implies. “Granger causes” is the key term. Playing with time periods. Perhaps, by using “Granger causes”, you are willing to concede that effect is not cause.
BTW, I’m looking for an econometrics study that shows monetarism works. The only one I found is only to 60% confidence, not the usual 95%: Lawrence Christiano, Martin Eichenbaum, Charles Evan, “The Effects of Monetary Policy Shocks: Evidence from the Flow of Funds”, Review of Economics and Statistics, February 1996, 78-1, cited in Olivier Blanchard “Macroeconomics”, 2nd edition., p. 96, 5-6, ‘Does the IS-LM Model Actually Capture What Happens in the Economy?’ Surely there must be another more robust study? My email is: raylopez88 at gmail dot com. Thanks and goodbye.
27. July 2015 at 17:25
Ray,
Destroys what could have been, I.e. a higher standard of living, more economic liberty (since the destruction from inflation has historically served to increase general government intervention).
Socialism in all of its forms destroys. There only seems to be possibility of non-destruction because of anchor bias, frame of reference bias, etc. The notion that we ought to settle with 2% real growth as “normal” and “healthy”. As if blood letting the victim slow enough to still allow a temporal growth in the victim’s health is evidence that the blood letting caused a growth against what would otherwise be less growth without the blood letting.
The fact that a victim grows temporally, that is, its growth over time is increasing, does not suggest inflation isn’t all that bad. That is a slave mentality and morality. Make do with far less than otherwise because it is difficult to imagine just how much better things could be.
Summer’s cheerleading for certain fleeting instances of growing temporal AGGREGATE real wealth around the world, is only to distract himself from the actual effects of what he advocates, which is billions of people worldwide having a far lower standard of living than they otherwise would have and could have had, had money production and distribution be subjected to the very same market forces he, at least in written form, says should apply to capital goods.
And since reduced material wealth in general always always always hits the very poorest and destitute the hardest, who are juuuust priced out of the market, not because of nominal prices only, but because of lower real wealth being produced in general. People have been killed because of Summer’s ideology in action.
He has literally zero comprehension or appreciation of just how pronounced the negative effects have been. He sees city lights and highways, and like a child visiting an amusement park for the first time, gives way to euphoria founded on ignorance.
No, the battle is not over “menu costs”. It is over economic calculation, temporal coordination, and property.
27. July 2015 at 18:21
Ray Lopez,
“The only one I found is only to 60% confidence, not the usual 95%: Lawrence Christiano, Martin Eichenbaum, Charles Evan, “The Effects of Monetary Policy Shocks: Evidence from the Flow of Funds”, Review of Economics and Statistics, February 1996, 78-1, cited in Olivier Blanchard “Macroeconomics”, 2nd edition., p. 96, 5-6, ‘Does the IS-LM Model Actually Capture What Happens in the Economy?'”
Actually that’s one of my favorite papers. Here’s the link:
http://faculty.wcas.northwestern.edu/~lchrist/research/fofa/flowoffunds.pdf
The dashed lines are one standard error bars. To get 95% confidence intervals (two standard errors) you simply have to double the distances (the authors are assuming their readers are capable of making the mental leap).
Look at page 22 and you’ll see that the effect of a shock to the federal funds rate on output, employment, unemployment, commodity prices, retail sales, retail trade profits, non-financial corporate profits and manufacturing inventories are all easily statistically significant at the 5% level at some point.
There are literally hundreds of similar Vector Auto-Regressive (VAR) studies showing similar results.
27. July 2015 at 18:52
Travis, Thanks for the links.
27. July 2015 at 20:40
Scott,
“…trying to explain economic concepts to Ray is like trying to teach quantum mechanics to a cow.”
I always thought it was more like trying to explain gravity to a chicken.
27. July 2015 at 20:42
@MF: “This is why I have hatred and vortiol spewed at me from the child like thinkers on this blog.”
Nope, that’s not why. You’re wrong again.
27. July 2015 at 21:15
Geddis also beleives I want to take away his mommy and daddy.
28. July 2015 at 07:08
@Mark Sadowski – did you type your post with a straight face? You think nobody reads these comments so you can just say anything?
First, the Christiano paper in the URL you cite is taking forever to download (I’m in the Philippines where the internet is spotty). If you care to email it to me I would appreciate it…wait it just came though.
But it seems to be data mining (in the bad sense) when you say:
“The dashed lines are one standard error bars. To get 95% confidence intervals (two standard errors) you simply have to double the distances (the authors are assuming their readers are capable of making the mental leap).”
What? So to use a chess analogy, if I, rated class A, which is in the top 10% of all chess players, beat by chance a master (2200 Elo), and you call that one unit distance, then you can assume I’ll beat a grandmaster someday, who is further out on the curve? Reasonable maybe for me, but try that now with a “Class E” (beginner who just learned the moves) and you’ll quickly see this sort of “interpolation” is bogus. Instead of ‘assuming’ show me the actual data where a class E player beats a grandmaster. I’ve not seen a single such game in all my years of studying chess, though in theory it’s possible.
“Look at page 22 and you’ll see that the effect of a shock to the federal funds rate on output, employment, unemployment, commodity prices, retail sales, retail trade profits, non-financial corporate profits and manufacturing inventories are all easily statistically significant at the 5% level at some point.” – but those bands are at 60% confidence! Not 95%?! Is that not so? “At some point” says Mark… love that fib. He’s essentially saying, ‘this VAR study is bogus data mining, and if you run the simulations long enough by chance you’ll get whatever you want ‘at some point someday'” – lol.
“There are literally hundreds of similar Vector Auto-Regressive (VAR) studies showing similar result” – hundreds? name one please. Even Christiano et al upon a careful review fails to show what it purports to prove–and that’s not me talking but the Dallas Fed. The problem I just found from the few pages of Christiano et al that did download is that the authors assume (p. 18) that they have an answer to the well known “prize puzzle” that blows a huge hole in monetarism. But this prize puzzle cannot so glibly be explained away. In fact, using this very same Christiano et al paper (in a pre-published form in 1994, but essentially the same paper), the Dallas Fed (Balke et al) found, using VAR and your favorite “Granger causation”, that Christiano et al’s explanation could NOT explain the prize puzzle of the 1960-1979 period (see: dallasfed.org/assets/documents/research/er/1994/er9404b.pdf see page 18 of this paper)
C’mon Mark, you are just b.s-ing dude. Let’s take this offline if you want to debate further. You got nothing and you know it. You’re just trying to bamboozle the hapless readers who can’t follow this debate properly.
28. July 2015 at 08:09
Was reading some interesting papers in the VAR sphere for some related work I was doing in completely different area that may be of interest to Mark (the evidence IMO is much more convincing than the traditional VAR literature). Further, it’s fun to nail Ray because he is so adamant he knows things which he clearly does not.
Basically, I find the factor augmented VAR evidence much much more convincing because it directly addresses unaccounted for information and data.
Off the top of my head, I recall this paper (more respected for those that care about names… I am looking at you Ray):
http://www.federalreserve.gov/pubs/feds/2004/200403/200403pap.pdf
And a much more interesting yet relatively unknown work on Bayesian FAVAR with sign restrictions:
http://www.diw.de/documents/dokumentenarchiv/17/diw_01.c.423811.de/amirahmadi_uhlig_b-favar.pdf
28. July 2015 at 08:24
@Ray Lopez: “the well known “prize puzzle” … this prize puzzle … the prize puzzle”
Fascinating, Ray! I’ve never heard of this “well known” prize puzzle. Couldn’t even find it in the paper you referenced. Do tell more. Please teach me! What’s a “prize puzzle”?
28. July 2015 at 08:27
Price puzzle
28. July 2015 at 08:56
https://s-media-cache-ak0.pinimg.com/736x/05/89/84/0589841b13c9245d9553901b1c58ee90.jpg
28. July 2015 at 19:29
@Student- thanks a lot. I find both papers interesting, both papers either supporting my position or flawed, and I invite you to discuss this further with me off-line, as this forum is not conducive to debate.
First paper [A] by Bernanke et al.- background most interesting on why VAR on occult variables is a data-mining bust, but even assuming it works, look at their conclusion: monetary policy contributes between 3.2% to 13.2% of the total in output such as cap utilization, new orders, unemployment, industrial production. That is a trivial. Monetarism then is about 3% to 13% of the total? I’ll take that as a concession of defeat.
Second paper [B] by Ahmadi et al.- sign restrictions in their model are a huge red flag. They are essentially taking the absolute value of variables, to augment the variables and give them “bite” (by their own admission they use this term). Analogy: if I’m a professional gambler at a poker table and only count the times I’ve won, not the times I’ve lost, then by the end of the year I owe the IRS a huge tax bill. That doesn’t sound logical and fair does it?
Needless to say, Sumner would never discuss these papers or points, as he doesn’t understand them. And Sadowski replies with the ‘face palm’ ad hominem reply, showing he surrenders as well.
RL
[A] Measuring the Effects of Monetary Po licy: A Factor-Augmented Vector Autoregressive (FAVAR) Approach * Ben S. Bernanke et al(2003) Apart from the interest rates and the exch ange rate, the contribution of the policy shock is between 3.2% and 13.2%. This sugge sts a relatively small but still non-trivial effect of the monetary policy shock. In particular, the policy shock explains 13.2%, 12.9% and 12.6% of capacity utilization, ne w orders and unemployment respectively, and 7.6% of industrial production
[B] MEASURING THE DYNAMIC EFFECTS OF MONETARY POLICY SHOCKS: A BAYESIAN FAVAR APPROACH WITH SIGN RESTRICTION – Amir Ahmadi et al. (2010)
Since the first draft of this paper, Mumtaz and Surico (2009) too have introduced sign restrictions to FAVARs. There are key differences between their approach and ours, however. These authors use factors as summaries of blocks of data, constructing, for example, a real activity factor. They then apply standard (contemporaneous) sign restriction and conventional identification techniques to the small-scale VAR in these factors. The difference of their paper to the existing literature is thus the application of sign restrictions to these constructed variables, but is conventional otherwise. By contrast, we do not impose a structural interpretation to our factors, and instead impose sign restrictions directly on the responses of the various variables, exploiting the variations in factor loadings across large sets of, say, price responses. It is the latter feature which gives our procedure “substantial bite” and constitutes the major innovation compared to the existing literature
Identifying assumptions are key. We propose to identify monetary policy shocks with the help of sign restrictions, introduced by Dywer [1997], Faust [1998], Canova and Pina [2005], Canova and De Nicolo [2002] and Uhlig [2005]. A survey and various extensions are in Rubio-Ramirez-Waggoner-Zha(2010). In this approach, the response of key macroeconomic variables such as prices and inter- est rates to a monetary policy shock is restricted to accord with a priori theory, in order to achieve identification.
28. July 2015 at 19:55
@myself – there is of course another more general criticism of these papers: they are Bayesian but they don’t up front define what their priors are (maybe it’s buried in the back).
Simple example: I run Bayesian VAR tests to show that two data sets are correlated as follows. The first data is [sic] roosters crowing in the morning. The second data is the sun rising. Treating the rooster as the independent variable, and the sun as the dependent variable, I conclude from VAR and impulse functions that you get the sun to rise when the rooster crows. Not realistic. Now I reverse the variables and I run another VAR test to show that when the sun rises the rooster crows. Realistic. You can see that the first Bayesian prior is not realistic. It all depends on what you are considering as your independent and dependent variables.
Further, and more fundamentally, I suspect the ‘granularity’ of the data sets used in all the papers (said data going back to the 1960s) is not fine enough to distinguish what the authors are trying to prove. The authors are saying when the Fed acts, something happens. But since data lacks fine granularity, it does not show that the Fed is responsible (except for 3.2% cause and effect, lol) for that something: it could be due to another, hidden, unmeasured variable. For example, the Fed could raise rates because already they see the futures market is anticipating greater rates (the hidden variable not in the data), not because of what the Fed does, but because business conditions are improving. The VAR tests don’t have sufficient granularity to account for this (correct me if I’m wrong, I’m not in the field, but that’s my impression).
28. July 2015 at 20:04
@myself – just to be clear on this point in my last post: “hidden, unmeasured variable” is not just a variable not in your data set. It is data that is not sufficiently granular to adequately capture what you are measuring, e.g., if Fed responses and market future rate data are quarterly, but the Fed responds a week after the futures market shows an expected rate increase, you cannot tell from the data if the Fed acted due to the market or acted independently (of course the two markets are intertwined, with one watching the other, anticipating the others’ moves like a game of chess, making the entanglement more difficult still).
In short, show me a study that blow-by-blow shows an unanticipated Fed response–like say 1994’s surprise interest rate rise, which we all would agree was a ‘bolt from the blue’–had a real effect on the economy that was persistent more than a very short time. Won’t happen since it doesn’t exist.
28. July 2015 at 21:15
@Ray Lopez: “I’m not in the field”
What?! Shocking. I never would have guessed!
Then how did you know so much about the “well known “prize puzzle”“?
29. July 2015 at 03:22
You are wrong about so many things in what you said here I don’t have the time to address them.
The priors are specified, that tells me you didn’t read them… try again before I waste any time. This time, however, don’t skip the math and read only the intros and conclusions.
29. July 2015 at 06:06
@Student – you have a lot to learn, not just about the T-distribution either, but about life. Please then tell me what the priors are, if you’re so well read. You didn’t read the paper either. But I read the conclusion, and I understood it, outlined above. See below and replace ‘communist’ with ‘money communist’ = monetarist as MF would say, and replace “Marx and Lenin” with “Friedman and Sumner”.
@Don Geddis – you seem to have been overawed by Student and Sadowski, and are reduced to making schoolyard taunts from the sidelines like a boy. Don’t be that way, I liked the old Don Geddis better. They got nothing, just a few concepts borrowed from electrical engineering and statistics that they are attempting to build a career over. Pretty narrow career if you ask me, but they’d probably be thrilled to make about $75k a year (my starting salary, well over 20 years ago, and I quickly was making six figures a few years after that. Semi-retired now, as I’ve made all the money I care to make.)
RL
How do you tell a communist? Well, it’s someone who reads Marx and Lenin. And how do you tell an anti-Communist? It’s someone who understands Marx and Lenin. – Ronald Reagan
29. July 2015 at 06:44
The priors are fully specified in both papers. It is not my role to read the papers for you and then spoon feed to you what those specifications imply. You should do that yourself prior to commenting on things you have no idea about.
Further this is a rather straight forward case because they are using the same prior specifications, which are actually very conventional (see page 32-34 in the fed paper and page 20 in the other).
Your simply example (provided below) clearly demonstrates you have absolutely no idea what you are talking about. Its not even in the ball park. I dont think I could construct something more wrong than that on purpose.
It demonstrates a complete lack of understanding of what a prior is, what role it plays, how its specified, what it implies, etc. The ordering of the variables has nothing to do with it. You are absolutely utterly clueless and at the same time you are delusional and think you know what you are talking about.
You began by stating that there were no econometric studies on the issue. Then you declared there were no VAR studies on this issue without even doing a simple google search on your own. You expected Mark to educate you on the topic. Once that was pointed out, you scanned the introductions and conclusions of three of several hundred studies to construct a nonsensical response, completely misinterpreting nearly everything you were reading.
I dont believe anything you say and will go back to completely ignoring you for being the troll you are. You spout off complete nonsense and somehow think that by trolling a useful blog you are somehow making contributions to discussions on things you dont have a clue about. You are clearly delusional and are not worth spending but a few minutes every couple of months making fun of.
“Simple example: I run Bayesian VAR tests to show that two data sets are correlated as follows. The first data is [sic] roosters crowing in the morning. The second data is the sun rising. Treating the rooster as the independent variable, and the sun as the dependent variable, I conclude from VAR and impulse functions that you get the sun to rise when the rooster crows. Not realistic. Now I reverse the variables and I run another VAR test to show that when the sun rises the rooster crows. Realistic. You can see that the first Bayesian prior is not realistic. It all depends on what you are considering as your independent and dependent variables.”
29. July 2015 at 10:12
@Student – you lose again. Story of your life. A priori assumptions are in fact fundamental to how data is ordered in Bayesian analysis. It’s actually the very definition of Bayesian, “The Bayesian design of experiments includes a concept called ‘influence of prior beliefs’. ” – with your a priori beliefs being akin to postulates in mathematics.
Google is your friend here, perpetual Student. Use it.
Now go back to school, and stay there. Your appeal to unspecified ‘hundreds’ of papers is rebutted, but head, by the conclusion of Bernanke, yes, that Bernanke, who stated (and I repeat): “Ben S. Bernanke et al (2003) Apart from the interest rates and the exchange rate, the contribution of the policy shock is between 3.2% and 13.2%. This suggests a relatively small but still non-trivial effect of the monetary policy shock. In particular, the policy shock explains 13.2%, 12.9% and 12.6% of capacity utilization, new orders and unemployment respectively, and 7.6% of industrial production”
Got that? 3.2%. That’s what you are studying, something that has 3.2% effect. You are ignoring the 96.8% for the 3.2%. And constantly shifting the goal posts by citing yet more of your ‘hundreds’ of bogus papers won’t change that fact. You’re dismissed. And jobless. And not in the 1%. And … disses x infinity. Like in Epic Rap Battles of History, you’re in the dustbin.
29. July 2015 at 13:29
Ben Bernanke says: “the contribution of the policy shock is between 3.2% and 13.2%”
Ray Lopez says: “That’s what you are studying, something that has 3.2% effect.”
Hmm … who to believe? Ray or Ben? They’re saying different things. I wonder which one is more credible.
29. July 2015 at 16:15
@Don Geddis – hey Don, I think this thread is dead but thanks for the reply. You don’t have to believe me, believe Ben. He says monetarism’s contribution to various outputs is between 3.2% or 13.2%, said endpoints being equally likely says frequency probability.
BTW a good slide I found online to explain Bayesian priors is here: http://faculty.washington.edu/kenrice/BayesIntroClassEpi515.pdf go to slide #27. You can think this through and see I am correct. The Bayesian monetarist would say: “I know variable X is the result of the Fed raising rates” and then assign a probability to it. But suppose that money is neutral and the Fed has no power. Then this ‘prior’ is completely wrong. Now suppose by chance that X occurs when the Fed raises rates. To the Bayesian this is ‘proof’ but it’s a false dawn. Simple analogy: a monetarist thinks the sun rises due to a rooster crowing first. Assume roosters usually crow after the sun rises. But by chance, on occasion, for reasons known only to roosters, they crow before the sun rises. Every time, by random chance, that the rooster crows before the sun rises, say this happens 5% of the time, it’s Bayesian ‘proof’ of the monetarist’s ‘priors’ are confirmed, and it goes into their fancy VAR. But the results if given a large data set will reflect poorly on the conclusions (though 5% of the time they will be ‘right’). Yes, on occasion, about 5% of the time, the Bayesian monetarist is proved correct. But it’s a false conclusion, a false dawn. If the Bayesian updated their prior to drop the ‘rooster crows then sun rises’ prior, the real truth would come out: that the sun rising causes the rooster to crow, that monetarism is a chimera, money is neutral.
Catch you in another thread…I doubt I reply anymore here.
29. July 2015 at 16:24
Geddis:
“Hmm … who to believe? Ray or Ben? They’re saying different things. I wonder which one is more credible.”
Lol, are we still stuck in the immature appeal to authority fallacy?
“You better have faith in THIS priest, rather than THAT priest.”
How about looking actual arguments? Google “Ben Bernanke was wrong”.
29. July 2015 at 16:57
@myself, Don:
MF is right Don, you cannot appeal to authority, you must look at the facts.
Another good Bayesian tutorial I found is here: http://www.austintek.com/bayes/
Note the ‘prior’ here. Also note the priors in slide 30 of the PDF link. Thought experiment: if the rooster crows before the sun rises 0.000001% of the time, it’s unlikely anybody would say the rooster causes the sun to rise, whether they are Bayesian or ‘frequency based’ statisticians. But 5% is not that ‘rare’ so you will get people like the VAR econometric statisticians who will build a model that data mines data to show this 5% phenomena (5% also being the cutoff for most frequency based normal distribution ‘true or false’ calculations too). But these VAR monetarists are, sadly, being ‘fooled by randomness’ as Taleb would say.
29. July 2015 at 18:18
Major –
I’m surprised at you. Ray quotes Ben Bernanke as saying the “the contribution of the policy shock is between 3.2% and 13.2%” Then Ray appeals to Ben Bernanke’s authority in claiming the number is 3.2%, leaving out the range.
“Got that? 3.2%. That’s what you are studying, something that has 3.2% effect.” – Ray
Then you criticized Don for pointing out that Ray incorrectly appealed to Bernanke’s authority. Ray was the one appealing to authority.
Also, I suspect Don was making a reference to Ray’s first comment on this thread “Who to believe? ..”, which was an explicit appeal to authority.
29. July 2015 at 18:25
@MF, Ray: You’re telling me not to appeal to authority?! To look at “actual arguments” and “the facts”? LOL. Talk about the pot calling the kettle black…
MF: Ray’s only “facts” were that very Bernanke quote, which he helpfully included in his own comment. And then proceeded to mis-summarize. Of course appeal to authority is stupid (not that it has stopped either of you). Hence the little joke I was making. Since the real humor is that Ray isn’t even careful enough to avoid contradicting quotes he himself pastes into his own comments.
Ray, hint for you: I know all about Bayesian probability already. I don’t need to look it up on Google. BTW, you seem quite obsessed with crowing roosters. Perhaps you’ll find this of interest.
P.S., MF: “How about looking actual arguments?” Having a productive intellectual discussion requires the participants to be honestly seeking truth. You and Ray don’t qualify, you’re just trolling. Long experience has shown that there’s no point in attempting to engage you with any actual content.
29. July 2015 at 18:29
@Negation: Ah, at least someone got the joke! I was beginning to think that my efforts at humor had been wasted.
29. July 2015 at 18:56
Don he cant comprehend. His is talking nonsense. No use even wasting your time. He has no clue and there is nothing u can do about it.
29. July 2015 at 19:21
@Don, @Negation – I think the difficulty you are having is that you don’t realize that in statistics involving a confidence interval, when you say: “between 3.2% and 13.2%”, it is EQUALLY TRUE that it could be 3.2% or 13.2% or anything in-between. Google it, it’s Stats 102.
@Student – your true face comes out. Rather than an intellectual, you are a schoolboy who is surfing the net on their smart phone and making solecisms like ‘cant’ to save time on your typing (or you don’t know correct punctuation, either is equally likely). Vamos! You’re dismissed little boy.
29. July 2015 at 19:33
Student, You said;
“You are absolutely utterly clueless and at the same time you are delusional and think you know what you are talking about.”
As good a description of Ray as we are ever likely to get.
29. July 2015 at 19:55
Blah blah blah a rooster crows… blah blah blah… Therefore Bayesian priors are garbage… Blah blah blah… I am a 1 percenter…blah blah blah…. I have an IQ of 120…. Blah blah blah… You have no job… Blah blah blah…. I am a chess player…. Blah blah blah… I win.
Hahahah
29. July 2015 at 20:12
Geddis,
I suppose for someone in your position the easiest route to take is to call arguments “trolling” when they are too difficult to refute when they make you feel bad.
Not once have you actually demonstrated that anything I have said to you or anyone here constitutes “trolling”.
Trolling is defined as posting comments for the sole purpose to flame and get an emotional rise out of people. It is impossible for that to apply to me because everything I have ever said here was for the sole purpose of having arguments and debating arguments and comparing arguments. I don’t care whatsoever about your feelings or emotions, nor what upsets you, nor what makes you feel like trolling yourself. Yes, you are in fact a troll because you post comments for the purpose of getting a rise out of people. You spew ad hominem, make crude jokes, and you seek to play tribal warfare.
The real reason you no lo longer engage the substance of my arguments has long been known: you can’t actually refute the arguments.
So you try to convince yourself that I am trolling so that you can rationalize why you have chosen to be an intellectual coward.
You are not seeking truth, you are seeking acceptance and approval. Between you and I, I am the person who is more concerned with truth and intellectual honesty. You are lying to yourself. You are not interested in truth because you resort to advocating for violence. Seeking truth requires absence of violence. You do not care about the truth of a free market in money. You ignore it, support the violence backing it, out of some dishonest appeal to pragmatism.
Those who are actually seeking truth and wanting to be honest, don’t behave the way you do.
29. July 2015 at 20:23
Geddis:
“You’re telling me not to appeal to authority?! To look at “actual arguments” and “the facts”?”
Yes. You have shown appeals to authority on multiple occasions.
“Talk about the pot calling the kettle black…”
You don’t see me making the kinds of appeals to authority statements you’re making. You just said “Between you and Bernanke, Bernanke has more credibility, so…”
You don’t care about arguments and facts being the basis. You care about names, agreements, associations, and approvals. It is why you post here all the time. To get approvals in your echo chamber.
I post here to actually engage and test my arguments. Yes, you’re being used, but that is what free association in a context of peace is for us humans. Each human being an end in themselves in a context of peace makes all individuals both ends for themselves, and means for others who are ends in themselves.
You just haven’t grown up mentally enough to not get upset.
29. July 2015 at 20:33
Negation:
Ray wasn’t appealing to the authority of Bernanke, he was telling Geddis “If you don’t believe me, then believe Bernanke” precisely because Geddis responds to authority more than arguments.
Ray’s arguments Geddis takes as wrong because they come from Ray, and Ray is not credible. Ray knows Geddis is not engaging the arguments, so Ray refers to Bernanke as someone who agrees so that Geddis will engage the arguments.
This is similar to what Sumner does. Sumner also makes arguments and then tries to juice them up by saying “Friedman agrees with me” or “Bernanke agrees with me” (even though he disagrees with Friedman’s call to end the Fed in his later life).
This blog is run by a kind of economist but more a political strategist bobblehead on network news shows, who encourages appeals to authority.
I am not saying Ray never does it. Of course he does. But in this case, Geddis did.
30. July 2015 at 06:09
@MF – correct about Geddis and the issue at present. And notice the reply by “Student”, a clear troll post at 29. July 2015 at 19:55 ‘Blah blah blah a rooster crows…’ and the base and debasing encouragement by Sumner to Student at ‘As good a description of Ray as we are ever likely to get.’ at 29. July 2015 at 19:33. Wow, I’ve never seen a host of a supposedly reputable site encourage such trollish behavior.
I think this site has sunk to new lows. Then again both Murray Rothbard and Milt Friedman apparently made ad hominem a key part of their argument, so Sumner may be taking a page out of their playbook. As they say in law: when the law is on your side, cite the law; when the facts are on your side (but not the law), cite the facts; when neither the facts nor the law are on your side, ‘pound the table’ (appeal to emotion, like Sumner is doing now). Sumner and his minions are ‘pounding the table’ when I challenge them to find me an econometric study that shows the Fed moves the market in a substantial way (I’m willing to concede it’s 3.2%, as in Bernanke’s paper).
30. July 2015 at 07:10
@Ray Lopez: “when you say: “between 3.2% and 13.2%”, it is EQUALLY TRUE that it could be 3.2% or 13.2% or anything in-between”
Ah. Perhaps. But the conclusion you actually made from that quote was: “Got that? 3.2%. That’s what you are studying, something that has 3.2% effect. You are ignoring the 96.8% for the 3.2%.” Which is not at all the same thing. Your statements do not follow, from what Bernanke wrote.
Are you really so incapable, of admitting even the tiniest error, no matter how obvious or trivial?
(Of course you’re incapable of admitting error. That’s one of the hallmarks of being a troll.)
30. July 2015 at 18:22
@Don Geddis:
I’m sorry if, in the haste of giving a reply, I gave the impression you could ignore a range in a probability analysis. In fact, the range is useful because it gives an indication of how ‘sure’ the estimate is: a range of 3.2% to 3.299% to 95% confidence is a lot more ‘sure’ than a range of 3.2% to 99.2%, even though in both instances the 95% confidence includes either endpoint and everything in-between equally.
I apologize if I mislead you.
“(Of course you’re incapable of admitting error. That’s one of the hallmarks of being a troll.)”
Wrong again Don. Story of your life.
30. July 2015 at 18:59
The fact you are talking about “confidence” in a Bayesian setting reveals you don’t really know what you are talking about. Haha. Everytime you try it just makes you look more ignorant.
I suggest you do a little reading about the difference between confidence and credible intervals with respect to the Frequentist and Bayesian viewpoints. You are confused yet again. This is introductory level stuff here…
30. July 2015 at 19:36
As well, we are talking about a Bayesian framework here, so unless the posterior is completely uniform within the interval (which I guess is theoretically possible but not in this case), then the values have different likelihoods. Again, this is basic stuff you are butchering here. Come on man….
31. July 2015 at 00:12
Lol, the comments here still manage to be somewhat amusing. How about a quiz to show off your math prowess? If a robot leaps three times in a row, but in random directions, what’s the probability it ends up within one leap of where it started? Show your work. BQ: what’s the final density function?
31. July 2015 at 01:07
Trolls: don’t let this happen to you.