Confession: I’m an occasional stress-eater. Not so much the archetypal unhealthy stress-makes-me-binge-eat-every-time stress-eater, but when under pressure I occasionally cave and have fried chicken or a pair of donuts. And right now, I’m in the middle of a move. The stress is killing me and I was hungry to begin with and my apartment is a pleasant walk from an around the clock donut place. So I went for donuts.
I was enjoying the walk when crossing a nearly-deserted intersection someone yelled, “Faggot!”
I wheeled around to see my every stereotype confirmed. A 20-something I can only describe as a “douche-bro*” was leaning out of red, moving pickup truck. My mouth dropped open. I was genuinely surprised.
First off, I have passing privilege! Folks on the street don’t peg me as gay so long as I don’t, you know, talk. I have only twice had this kind of drive-by harassment in Bloomington. Once it was truly a surprise, the other time I was holding my then-boyfriend’s hand. I know LGBT folks in Bloomington get street-harassed, but I have the privilege of “passing” as straight and usually don’t. It’s always jarring to lose it.
But also—this is Bloomington!
That’s why I’m putting this down. This happens to less “straight-seeming” folks than me on a regular basis. (Those who feel comfortable, please, please consider testifying to that fact in the comments either here or on Facebook.) It’s easy to sit back and think, “Ah, Bloomington! A place where gays can be gay!” But it’s not so simple.
Homophobia, both casual and aggressive, still permeate many spaces. This is just one facet. It spans from micro-aggressions and tokenization to outright violence. One must still be careful in some bars, mind what they say to some professors, and in general worry about things straight folks still don’t have to.
Don’t mistake me. Bloomington has been very good to me, as a gay man and beyond. I feel a bit guilty that these will be among my parting words. But if I learned anything here, it is the importance of Justice and calling out these sorts of things.
So to everyone, Bloomington’s great. Gay friendly. I’ve told younger LGBT folks that, if the University is a good fit academically, they needn’t worry too much about being out and proud in this town. My boss here endorsed me giving my number to a flirty costumer. If I have a complaint, it’s that the gay men here are a little too comfortable with the status quo.
But to those who are still going to be here, and those of you who’ve left, we have work to do. Bloomington isn’t perfect. Whether we’re talking about what to do moving forward or thinking back to what Bloomington did right we want to carry elsewhere, we should remember that.
Because, dammit! I know most of you want to make it all the better.
P.S. For those who might be worried, I’ll be fine. I’m well past the point that a little street-harassment will do more than unsettle me for a few minutes. I post because not everyone is as confident as I feel.
*Yes, “douche-bro” is hardly a term that shows grace on my part, but I’m not going to apologize for this either.
Kansas cut taxes hoping it would get enough growth. It hasn’t happened, and we shouldn’t expect it to have.
We Bore the Freshmen With This
The concept Kansas is working with is called The Laffer Curve. In theory it’s a sound idea, but it’s a good example of wishful thinking all the same.
The intuition is easy enough. The higher the tax, the lower consumption. When the tax is zero, the state makes no money. At some point, the tax will be so high that the thing in question won’t be sold, and the revenue will also be zero. (Well, legal sales, anyway, but modeling the black market is unnecessary. If you have properly estimated the legal market, the black market is implicitly accounted for in your curves.) That means that state revenue as a function of the tax rate will have a characteristic “hump” shape. This curve is called the Laffer Curve.
This is one of those things that is true for most markets. It’s possible to mathematically contrive situations where the Laffer Curve won’t have that hump, but those goods would be extraordinary. They’d violate the laws of supply or demand! So, Laffer isn’t wrong and those who follow him make a fair point.
But that dries up fast.
An Elementary Example
Most intro classes put supply and demand down as lines a plane. This stylization is just that—so we must keep in mind that any results we derive from this are limited by our model’s assumptions. Still, this graph will look familiar:
What you find is that state can increase its revenue only to a point, as suggested by the Laffer intuition I detailed above. That point happens to be a 300% tax rate! The state can tax up to 300% before increasing it further costs it money!
Granted, the state may have other goals besides revenue collection. After all, by sheer numeric coincidence, the deadweight of this policy happens to be half of the taxes being collected. Whether or not its worth it to collect those taxes is open to debate, but it will never be worth it to collect more than 300%.
Pretty much whatever models you use, you get a large turn-around point. Without getting to much into the technical details, the simple models give positive results more often when you are measuring over smaller intervals; it may be a bit hasty to suggest that we have data to estimate the turn-around point. Still, Kansas should hardly have assumed itself near it!
Now Comes the Ass-Covering
The economist Stephen Moore was a proponent of this model. Cut taxes, increase revenue. As a testable proposition, even he concedes he failed. He’s written a piece for The Kansas City Star. I’m not here to take the piece to task, but it’s full of factual inaccuracies. (Krugman and the President are pretty far apart on economic strategies, for example.) What I am here to point out is that he’s engaging in some pretty simple misdirection:
As for Kansas, the tax cut has been in effect a mere 18 months—not a lot of time to measure the impact.
You’re kidding, right?. The Laffer argument is based on conditions on the ground. If you have have to appeal to long-term changes, you’re not doing a Laffer analysis! People should be aware that if they cut taxes, their state will likely move towards the red. In the long-run, those states might recover—though if you’re running a deficit, it’s hard to argue your state is attractive.
Taking the piece at face value, there may be pseudo-Laffer effect whereby cutting taxes leads to an increase in population. Two things must be noted. The piece takes data from sustainable cases; the states in question aren’t in the red. Second, this may still imply a per-capita falling of welfare—welfare here in the economic sense of GDP per person. If you double GDP but triple population, the average person is worse off. By the same token, if you double population but keep government spending the same, it has effectively fallen.
There’s a lot more to parse in that piece—and one might reasonably support sustainable cuts to government—but don’t fall for the misdirection: Kansas’s government was not able to make up the difference from growth.
Well, Did It Create Growth?
Tax cuts are expected to create more consumption on net, and so more jobs. A good deal of the difference is lost to rearrangement. Cutting a 5% tax rate in half in the above example increases the amount of money in the economy by a mere quarter of a percent.
But this assumes the markets being intervened in are unaffected by the tax cut—which is likely wrong. If government intervention is correcting an imbalance there, then it will eat that small margin. People like Laffer and Moore are optimistic that those interventions are inefficient as well, and so the margin might be bigger than a quarter of a percent.
This is notoriously hard to measure. It’s why a good policy mix isn’t just a matter of empirical fact. (Though, yes, that matters.) A good rule of thumb, then, is to consider what’s being lost when we make the tax cut.
Laffer was wrong, again. Cutting taxes, when taxes are not exorbitant, will not lead to an increase in government revenue.
Freshmen in college are shown these models. There is simply no reason for Governors and economists to misunderstand them. Extraordinary claims require extraordinary evidence. That lowering the rate you collect revenue should raise the amount of revenue is extraordinary.
Don’t fall for it.
I have to confess, I never understood the hatred of mathematics.
Actually, that’s a lie. I too had to survive math class. It’s an easy mistake to conflate the torture of 45 minutes of high school math every day for years with the untethered power and unspeakable beauty of mathematical exploration. You have to realize that the loathing of math is a relatively recent phenomenon*. The ancients were absolutely entranced by the intrinsic elegance of mathematics.
One of the most beautiful objects of study is the so called “Golden Ratio”. I’m not going to belabor its derivation because I’ve gone through it before, but it is the division of a length such that the whole divided by the larger part is equal to the larger part divided by the whole. That ratio is:
Pages about phi—and the internet is simply filled with them—usually present a pair of approximations that are simply mystifying. The first is an iterated square root:
The other is a continued fraction:
Neither of these are very intuitive. For a long time, I was convinced these were found through good luck; an infinite number of mathematicians playing with strange structures for an infinite amount of time will eventually find all interesting numbers. But mathematicians aren’t monkeys. They are methodical. And while structures like this are reached by good luck sometime, something this strange was probably found analytically.
And then it hit me.
Back to Basics
If you go through the work of deriving phi from its definition above, you eventually find out phi is the positive solution to a quadratic. (If you haven’t read my derivation of it, consider following the link.) This quadratic in fact:
Using the quadratic equation will yield phi.
The quadric equation is solving for x, but what if we solved for x badly? In the spirit of bad math students everywhere, we might end up with:
Taking the square root of both sides and switching the order of x and 1 for convenience later,
This is called a recursive definition. It is a definition where the term defines itself. (Mathematicians aren’t terribly hot on these; computer scientists love them. There’s a deep meaning to be derived from that, I know it!) If we replace the x on the right with the entire right:
By repeating this ad infinitum we will arrive at the first approximation!
Trying another “bad” solution for x, we can put all the terms with x on the left side and all others on the right:
Dividing both sides by x:
and then “solving” for x on the left side:
Again, this gives us a simple recursion. Substituting the right into itself:
And once again we could do this again to get the second approximation!
It is commonplace to present these mathematical discoveries as facts from the Gods. And maybe they are. But we discovered them anyway. The internet doesn’t need more mystifying approximations of phi, but rather, clear explanations of the power the algebra.
Because that’s what makes math awesome!
*Of course, so is how many people we educate, so maybe Greek peasants would have hated geometry as well.
Can I get a cheer? Marriage equality in Indiana! I don’t care that it’s going to be staid, that we’ll have to wait for the Supreme Court, or really anything else. Tonight, there are couples in Indiana who are married who weren’t before. We can deal with the AG in the morning.
What I do want to draw your attention to while it’s still an active story is the most immediate dimension of this fight: County Clerks have a fair amount of leeway in implementing this ruling*. Some Clerks are issuing liscences, some are waiting to weigh the options, and some have already denied them. That is, right now, marriage equality is a county-by-county patchwork.
This map, provided it’s still up, tracks which counties are issuing and which aren’t in real time**. What’s striking is how much it’s a map of Indiana’s stance on same-sex marriage. Most of the counties denying at the time of posting are in the extraordinarily conservative East-Central region. While there are some surprising counties, by and large those issues are the more reliably liberal.
There’s a case to be made that County Clerks shouldn’t be elected, but to some degree that’s moot: even if they were appointed, it is still local elections that determine this. And it’s not just the clerk. Got a strong feeling about sexual assault? Sheriff and coroner matter. Even positions like Appraiser can become politicized when relevant Court rulings come down.
So let this be a lesson: how we vote in local elections matters.
*This may not be legally the case, and I want to specifically welcome comments that can shed light on that question. That said, as a de facto consideration, it is clear there is some leeway.
**So, yeah, this hyperlink may not age gracefully.
Paul Krugman, Nobel Prize Laureate and New York Times columnist is one of my least favorite columnists. Among other things, he routinely claims consensus on issues where none exists, and then excoriates his opponents for not having heard that there was agreements among columnists. It gets rather tiring rather fast.
Take Krugman on the minimum wage (emphasis mine):
Now, you might argue that even if the current minimum wage seems low, raising it would cost jobs. But there’s evidence on that question — lots and lots of evidence, because the minimum wage is one of the most studied issues in all of economics. U.S. experience, it turns out, offers many “natural experiments” here, in which one state raises its minimum wage while others do not. And while there are dissenters, as there always are, the great preponderance of the evidence from these natural experiments points to little if any negative effect of minimum wage increases on employment.
Reading this, you’d get the impression most studies find that the minimum wage doesn’t impact employment and the few that do are from the kinds of professors we don’t take seriously anyway. If Krugman were glossing over a few dissidents as he implies, I wouldn’t blink an eye; he is writing a column, not a literature review.
The problem is that the actual literature reviews do not show any kind of consensus. University of California, Irvine Professor David Neumark at the end of an extensive review:
Although the wide range of estimates is striking, the oft-stated assertion that the new minimum wage research fails to support the traditional view that the minimum wage reduces the employment of low-wage workers is clearly incorrect.
You’ll be unsurprised to learn that’s because the majority of the studies found disemployment effects:
Indeed, in our view, the preponderance of the evidence points to disemployment effects. For example, the studies surveyed in this monograph correspond to 102 entries in our summary tables. Of these, by our reckoning nearly two-thirds give a relatively consistent (although by no means always statistically significant) indication of negative employment effects of minimum wages, while only eight give a relatively consistent indication of positive employment effects. In addition, we have highlighted in the tables 33 studies (or entries) that we view as providing the most credible evidence; 28 (85 percent) of these point to negative employment effects. Moreover, when researchers focus on the least-skilled groups most likely to be adversely affected by minimum wages, the evidence for disemployment effects seems especially strong. In contrast, we see very few—if any—cases where a study provides convincing evidence of positive employment effects of minimum wages, especially among the studies that focus on broader groups for which the competitive model predicts disemployment effects.
You read that right, the studies they found most credible had disemployment effects. If you’re interested in the gory details, the study goes on to highlight concerns with the new minimum wage literature and is very comprehensive.
If Krugman wanted to say that he gave more weight to the studies that found the steady employment rate, he need only say that. He’s a Nobel Laureate and respected economist; his professional judgement is worth something. But by taking the demonstrably incorrect stance that the preponderance of evidence is on his side, he calls into question that very professionalism.
To whit: Krugman is one of the dissenters, and he undermines his cause when he says otherwise.
It’s not secret that I’m partisan against a minimum wage increase, mostly because the balance of the evidence is against. It brings me some joy to call Krugman out on lying about that.
Unfortunately, he does that on things I happen to agree with him on. Krugman on stimulus:
What about positive evidence for the benefits of stimulus? That’s trickier, because it’s hard to disentangle the effects of the Recovery Act from all the other things that were going on at the time. Nonetheless, most careful studies have found evidence of strong positive effects on employment and output.
A generous reading of this is that he’s making that professional judgment I alluded to above. But why not be more explicit? Why not just say that, as Paul Krugman, he believes the end of the estimation range has the highest quality studies?
Krugman’s closer to the mark here, to be fair. Valerie Ramey gives a range of .8-1.5 for the multiplier—though note she’s not a fan of Zero Lower Bound assumptions. Still, Krugman calls 1.3 “conservative“.
And that studies don’t find the zero lower bound assumption to hold is important as well! Krugman is want to act like that is a foregone conclusion, something everyone who has read the literature finds. But studies frequently find the opposite, as Ramey points out*.
Krugman gives an impression that most of the literature supports the kind of high multipliers he’s—and, to be fair, I’m—biased towards. In reality, those are outside estimates. It’s one thing to say that we think the way those studies are conducted are superior. It’s also fine to say that we are inclined to take them more seriously because of our a priori beliefs from models. It’s quite another to claim that everyone reasonable should plainly see it our way!
*Krugman, I think, may be aware of the problems with Ramey and other’s approach to finding that—though I’m just reading between the lines. Critics have pointed out that some of those studies use too high a threshold for near-zero, and that if you’re more stringent, you find the ZLB assumptions hold. Regardless, Krugman still takes the tone that this is a settled conclusion, as is his habit.
The Krugman Problem
Truth, as is popularly said, is not a democracy. Even within the scientific fraternity, it is acceptable to give greater weight to greater voices. And it is normal for academics—including economists—to bifurcate on what the best way of figuring out truth is. Those who have contributed disproportionally to a field, those who have demonstrated a greater grasp, are often given a greater voice.
Which is what makes Paul Krugman’s habit of claiming consensus on matters where no such thing exists so vexing! Krugman is a Nobel Prize Laureate and a brilliant mind, especially in the sub-field of international trade. If he was taking the position that, in his estimation, Card and Krueger type studies are the preferred methodology, it would carry some weight. But in scientific disciplines, the consensus (or lack thereof) is a knowable thing. His opinion on what the general opinion is carries no weight next to the evidence of such things.
It’s hard to know if Krugman is being duplicitous or simply lazy. It doesn’t really matter, if you consider it. Either way, it’s bad.
First of all, dear reader, I promise to soon get back to my regularly scheduled posts about more traditional econ and political science fair. But, there’s been no shortage of feminist-oriented material as of late, and some of it intersects handily with the kind of data and experiment design my econ/political science training is good for. I plan to have my next post be about Paul Krugman’s terrible habit of claiming false consensus and not mention anything about feminism!
But, today, I wanted to briefly address Lara Stemple’s The Sexual Victimization of Men in America: New DataChallenge Old Assumptions. It’s behind a paywall, sadly, so if you want to read it you’re going to have to find it for yourself. It’s in this month’s (June 2014) edition of the American Journal of Public Health.
The Slate also has a write-up by Hanna Rosin. It does a good job highlighting the strong points of the study, but if you read it, know Rosin has an ideological axe to grind. She has put forward a hypothesis that the post-industrial world benefits women more than men. I’m not here to critique Rosin, but Rosin has no incentive to be critical: the idea that women get away with rape backs her thesis.
Stemple’s study has ample merits for Rosin to highlight. Stemple argues, I would say correctly, that the changes to the definition of rape and sexual assault only begin to combat underreporting by men. The fact that “carnal knowledge of a female” was used by the Feds into the second decade of this century is embarrassing, not to mention sexist. These amendments will give us a better picture of how close Stemple’s estimates are to the mark.
But that’s the weakness of this study. Stemple’s research is based on survey data. It’s an important piece of the puzzle—we need to ask people if they’ve been sexually assaulted, and we need to ask those questions well.
One result I was alarmed to see taken at face value was the finding that teenaged boys, when asked if they’d had sexual contact with the staff in detention said yes—and said it was overwhelmingly with the women. Trepidation is warranted in pointing this out, but it might say more about teenagers in juvenile detention that than the staff that they’re claiming this about. The Federal Government should take these figures seriously and try to mitigate these sexual assaults, as many are almost certainly happening. Responsible scholars and commentators should remember survey data from teenagers are notoriously unreliable and note what’s been done to try to head off false reporting.
But, deeper, why try and conflate teenage boys with adult men? For that matter, why try conflate the non-institutionalized population with the institutionalized one? Rapes in jails and prisons—regardless who they’re by or against—deserve equal consideration. But aggregate statistics reporting them together tell us far less about how to help either population than the disaggregation, doubly so if you start including teenagers.
Stemple’s argument, to be fair, seems more aligned towards tearing down the notion that women automatically the victims and men are the perpetrators. Insofar as we’re reading her that way, I have no beef. But let’s head off the interpretation that because things are happening in prisons, we can draw broad-based conclusions about gender, sex, and power outside of prison. Presumably, prison is different sort of environment and requires a different sort of policy.
Another point of some concern for me that Stemple downplays that the only study to collect lifetime data found a disparity between the last 12 months and lifetime sexual assault. In the 12 months leading up to the study, apparently, the same number of men and women were sexually assaulted. In the lifetime for those same people, women experienced comparable acts of assault between 2 and 3 times as often. There could be a lot of reasons for this ranging from a shift in perpetrators, underreporting of old assaults by men, to flaws in the survey causing men to overreport recent assaults. This merits attention, as a discord like this indicates something is up!
And, really, that’s how I feel about Stemple’s work here in general. This merits attention. If these results are ultimately critiqued as spurious and unrepeatable, we’ll have thoroughly explored it and be able to do better policy. I suspect we’ll find that the truth is somewhere between the low stats for male victimization we see now and Stemple’s findings, but we should find that out.
Regardless of the numbers, however, it is unlikely that rapes of men and rapes of women are essentially the same phenomenon. In a society that aggressively bifurcates between men and women, it is possible to get similar gross figures and have the dynamics of power be different. For an example I found more data on, women are consistently found to be victimized in violent crimes by intimate partners more often than by strangers despite men seeing more total victimization. Gender interacts with these sorts of things in funny ways. It is important we ask how we can help men who have been raped. But, derailing conversations about helping women to contentiously insert men helps neither male nor female rape victims.
Ultimately, I think Stemple’s work is an excellent jumping off point and a call for more research. It is just too soon to call it definitive or game-changing.