At one point during my wanderings around New York City a few weeks ago, I ended up on the east side of Central Park near a plot of grass that’s apparently a dog run. As you can see, lighted collars are now stylish accessories for nighttime canine frolics, and when I saw these dogs the first thing I thought of was Star Wars. I was really hoping they’d both run around and play with each other, allowing me to get some kind of picture that looked like a light saber duel, but no such luck. They just meandered around a bit while their owners argued politics.

A few years ago the city of Seattle passed a law that gradually raised the minimum wage to $15+ per hour starting in 2015. A group of researchers at the University of Washington immediately got to work to track the effect of the law. Their most recent study was released in October and was met with a little too much nyah-nyah-nyahing for my taste. The plan all along was to follow the data over time and write progress reports every few months, so the fact that one report comes to different conclusions from a previous one doesn’t really mean much. It doesn’t mean they “changed their minds” or anything like that. It just means they added another couple of data points and reassessed their results.

So. Anyway. Here are the results for hourly wages through the end of 2016:

The methodology used is to compare wages in Seattle with wages in a “pseudo-Seattle” that’s constructed of other cities that, when mushed together, very much resemble ordinary Seattle. More here if you’re interested. In this case, wages increased far more in Seattle than in the synthetic version of Seattle, which is no surprise. But did this affect the number of hours worked?

There’s a small disemployment effect of about an hour a week for low-wage workers. Now let’s combine the higher wages with the lower hours worked:

Low-income workers in Seattle, on average, earned about $17 more per week than they would have if the minimum wage law hadn’t changed. The researchers also come to a few other conclusions:

• All the gains accrued to experienced workers. Inexperienced workers earned the same weekly wages as before.
• Some of the gains came from working extra hours outside Seattle.
• Job turnover decreased.
• New entries into the workforce also decreased.

That’s it so far: mostly good news, with a bit of concern about the effect on new workers entering the job market. Personally, though, I’d caution not to take any of this very seriously until we see a change in the economy. If a big change in the minimum wage were going to have an effect, I’d expect to see it first when the economy turns down, as it eventually will. There’s a good chance that we won’t get a final read on the Seattle experiment until 2020 or later.

This isn’t really more predictions, just the same one as yesterday but presented with a little more context. I finally got tired of reading endless headlines about how the Democratic lead in the generic ballot is “shrinking” or “widening” or “eroding” or whatever, so I went off to 538.com to get the answer. Here it is:

There you have it. Ignore the individual polls, which can easily change by several points for no particular reason. You may truthfully say that “the Democratic lead has grown substantially since summer” or “the Democratic lead has shrunk slightly since September.”  Whichever one supports your narrative is OK. But those two choices are about all you have.

I have no intention of making any predictions about Election Day.  The New York Times, for example, suggests that Democrat Katie Porter will edge out incumbent Republican Mimi Walters here in my home district, the California 45th deep in Orange County. That would be so great! But having lived for 60 years—represented by Republicans in every single one of those years—it’s just hard to believe it might finally happen. We’ll see.

Anyway, with no polling days left before Election Day, here are two projections for the House. The first is from Sam Wang:

He puts the generic congressional ballot at +8 percent for the Democrats, which would be enough to barely win control of the House with a pickup of around 30 seats. And here’s Nate Silver:

This appears to be a bit more optimistic, with a projection of a 38-seat Democratic gain and an 85 percent chance of winning control of the House. Keep your fingers crossed and keep ringing those doorbells!

POSTSCRIPT: As for the Senate, don’t ask. No one has much confidence the Democrats can run the table and take back control. Unless something big happens, it looks like Mitch McConnell has two more years left as majority leader.

Following up on the whole NAFTA thing, I thought I’d collect a few summaries and charts that would help explain the consensus on NAFTA a little better. First off, here’s a survey of economists:

This is the mainstream view of NAFTA. There are some lefty economists who would disagree with this, for the same reason they dislike most trade agreements: the benefits tend to skew to high-income workers while displacing some number of low-income workers. But economic elites and national leaders tend to like trade agreements a lot. Here’s the Economist on the 20th anniversary of NAFTA:

For its first six years the pact’s success was remarkable. Cross-border investment ballooned. Rapid growth led to the three countries’ share of global production hitting a peak of 36% in 2001. But the progress stalled. Fears of terrorism in the United States after September 11th 2001 led to a security clampdown on its borders that still hampers trade today (the drug war in Mexico has not helped). Even more destabilising, after China joined the World Trade Organisation in 2001, American firms in Mexico upped sticks and headed across the Pacific. Trade within North America eventually started to grow again; but global trade—principally with China—mushroomed.

This is an important point: after 2000, when China joined the WTO, it gets very hard to analyze the impact of NAFTA. The trade data simply becomes overwhelmed by the Chinese tidal wave in both manufacturing and agriculture. Nonetheless, by 2017, when NAFTA was two decades old and the Chinese tsunami had taken its full effect, mainstream American economic opinion continued to be moderately positive toward NAFTA.

But what about Mexico? Has NAFTA been good for them? There’s no simple survey of Mexican economists that I’m aware of, but here’s a 2006 survey taken in Mexico that asked generally about international trade:

So a decade after NAFTA, Mexican public opinion toward trade was still generally favorable and elite opinion was extremely favorable. And what about economic researchers? Do they think NAFTA was good for the Mexican economy? Here’s the generally reliable Congressional Research Service:

A number of studies have found that NAFTA has brought economic and social benefits to the Mexican economy as a whole, but that the benefits have not been evenly distributed throughout the country….positive impact on Mexican productivity….positive effect on stimulating the productivity of Mexican plants….economic effects have found that the net overall effects on the Mexican economy tended to be positive but modest….NAFTA helped Mexico get closer to the levels of development in the United States and Canada.

Other studies suggest that NAFTA has been disappointing in that it failed to significantly improve the Mexican economy or lower income disparities between Mexico and its northern neighbors.

….Many critics of NAFTA say that the agreement led to a large number of job losses in Mexican agriculture, especially in the corn sector. One study estimates these losses to have been over 1 million lost jobs in corn production between 1991 and 2000. However, while some of the changes in the agricultural sector are a direct result of NAFTA…many of the changes can be attributed to Mexico’s unilateral agricultural reform measures in the 1980s and early 1990s….These reforms coincided with NAFTA negotiations and continued beyond the implementation of NAFTA in 1994. The unilateral reforms in the agricultural sector make it difficult to separate those effects from the effects of NAFTA.

Ah, the “corn sector,” source of a million arguments. A lot of people continue to believe that “NAFTA killed Mexico’s rural corn farmers,” but there’s a big problem with this: Mexico initiated sweeping economic reforms that took effect from the late 80s to the late 90s, and NAFTA came right in the middle. Here’s the basic data, courtesy of the USDA:

This is raw data, and you can read it a lot of different ways. But a few things are clear:

• Mexican corn production began to increase in the late 80s and has continued to increase ever since.
• After NAFTA, US corn exports to Mexico increased.
• This produced a brief pause in Mexican corn production, which was flat between 1995 and 2000.
• It also reduced the price of corn by about a quarter.
• However, between 1994 and 2017, Mexican corn production has increased from 18 to 26 million tonnes. That’s a rise of 42 percent.

All of these things taken together unquestionably had a big effect on Mexican corn production, especially on rural farmers with low efficiencies. But the biggest point of pain for rural farming communities was not NAFTA. More likely, it was the Mexican government’s decision to reduce price supports for agricultural commodities like corn starting in the late 80s, well before NAFTA was signed:

The bottom line is that Mexico’s rural corn farmers, especially in the poorer south, did have a hard time. But this was due to a whole host of things, and NAFTA was only part of it:

• Like most developing countries, agricultural production in Mexico has for many decades been shifting to large-scale farms and capital-intensive food processing, which puts pressure on small-scale farms and household farmers.
• Economic reforms reduced the price of corn, and NAFTA reduced the price further.
• Access to credit for small farmers became harder after government subsidies and guarantees were removed.

And now for the key question: was NAFTA responsible for the huge immigration flows from Mexico into the United States during the 90s and aughts? Probably not. Here’s the conclusion of a Carnegie report from 2004:

NAFTA does not appear to be the culprit in this acceleration of rural out-migration, however. Taylor and Dyer found no indication that NAFTA created any sort of “break point” in the growth of migration from rural areas. Rather, migration from rural Mexico to the United States had been accelerating well before the onset of NAFTA, and the trend continued afterward.

….Even as agricultural prices dropped and Mexico’s trade deficit in agricultural goods with the United States widened after NAFTA, Mexico’s agricultural GDP increased. The gains in the value of Mexican agriculture were accompanied, however, by a seemingly paradoxical decline in employment in the Mexican agricultural sector, from 8.1 million jobs in 1993 to 6.8 million in 2001. A number of factors besides NAFTA were at work, however.

The first, and one that is often overlooked, is the natural, perhaps inevitable, process of rural-to-urban movement that all countries experience as their economies advance. The share of agricultural workers as a proportion of Mexico’s workforce has declined steadily, from over 50 percent in 1960, to 36 percent in 1980, to less than 25 percent and falling since 1995. The second factor is the continued reform of Mexican agricultural policy. Reforms of the ejido system of landownership that began in 1992 have allowed land sales and rentals and have been accompanied by cuts and changes in the structure of agricultural subsidies. Both changes have encouraged increased productivity and production, but not always in ways that have resulted in greater agricultural employment or have encouraged rural people to stay put.

Thus, NAFTA has played only a minor role in the continuing acceleration of rural outmigration during the decade since its enactment. The choice of whether to migrate within Mexico or to the United States, however, has been shaped by the larger and more structural general migration forces outlined in this essay, and by the unavailability and low quality of jobs in Mexico’s cities.

This got a little long. Sorry. But too many people see that (a) US corn exports to Mexico spiked after NAFTA and (b) lots of rural farmers pulled up stakes at around the same time and emigrated to El Norte, and conclude that (c) Bill Clinton and big, bad American corn wrecked rural Mexico.

It’s a good story, and it even has a kernel of truth to it. But that’s all. The real story is more boring, a combination of globalization, economic reforms, and a booming American economy that started long before NAFTA and continued long after it.

Oregon gets a lot of attention every election year because it’s gotten rid of polling places and gone to 100 percent mail-in ballots. That’s great, but Oregon’s real contribution was its leadership in something much more important: automatic voter registration. If you have a driver’s license, you’re registered to vote. End of story. It’s a trend that’s now spreading across the country:

Now, you will notice that Georgia and West Virginia are the only Republican states to have adopted automatic voter registration so far. So this is not quite the bipartisan initiative we might have hoped for. Still, it’s progress.

Of course, what would really be progress would be national ID cards. Everyone with an ID that marks them eligible to vote gets a ballot by mail, full stop. If you don’t have an address, you drop into a polling place, show your ID, and vote. If you try to vote twice, a computer catches you and a warrant is issued for your arrest. Done.

Now all I have to is convince all you laggards that a national ID card would be a good—no, a great—thing. So far, I’ve had no luck on that score.

There are times when I think I’m repeat-posting a picture, but I checked and this one has never appeared before. The problem, I guess, is that I don’t live in a mansion, and there are only so many places the cats can be. In this case, Hilbert is up on the fence peeking out from our neighbor’s bougainvillea plant. You can tell it was taken a while back since this plant has since been hacked away almost to bare stems. Neither of our cats is nearly as interested in the neighbor’s yard as they used to be back when it was a jungle.

Here’s a piece from Kevin Williamson today. This is the entire post:

That Thing That Never Happens but Keeps Happening

While I’m on the subject of Texas politics: You know that voting fraud that our Democratic friends are always indignantly insisting does not exist? A former Democratic-party official in Texas has been named as the financier behind a voting-fraud ring that has resulted in four arrests and dozens of felony indictments. The fax machine that prosecutors say was used to transmit fraudulent voting applications belonged to a former Fort Worth elected official, who, as you will guess, is not a Republican.

Voter fraud! I figured I had to check it out, which I did via the clever investigatory mechanism of clicking the link. Here’s what I dug up:

• The fraud that Democrats say “never happens” is a very specific type of voter fraud: in-person voting under a false name. This is the fraud that motivates photo ID laws.
• The fraud that Democrats say we should take seriously is mail-in ballot fraud. However, Republicans don’t like to worry about this because mail-in voters tend to be middle-class Republicans.
• The fraud that allegedly happened in Texas was mail-in ballot fraud.

You can stop now if you want since it’s already obvious that nothing much happened here. But there’s more:

• The main allegations aren’t related to voting at all. They’re aimed at faxed applications for mail-in ballots. In other words, it’s related to the standard-issue voter canvassing that’s a big part of GOTV operations for both parties in nearly all elections.
• Vote canvassers are usually paid by the local party, so there’s nothing odd about the fact that they received money “from funds.” In fact, the local party probably paid dozens or hundreds of canvassers.
• A specific group of four canvassers was indicted. Neither of the Democratic officials Williamson mentions was part of the indictment.
• It’s not uncommon for aggressive canvassers to push the boundaries of the law, which includes altering ballot applications if the voter filled out part of it wrong. If this actually happened, it’s a bad thing but hardly the end of the world.
• From the Star-Telegram story: “The state’s notice also accuses Sanchez of illegally voting by marking ballots or encouraging others to mark ballots in more than a dozen instances without the voter’s consent or knowledge. She is only charged, however, with one count of illegal voting and 16 counts of providing false information on an application. She also is accused in the notice of committing, or aiding others in committing, more than 100 counts each of forgery, tampering with a governmental record and providing false information on an application.”
• Note also that these are only charges. No one has actually been convicted of anything yet.

According to Texas Gov. Greg Abbot, “Largest Voter fraud Investigation in Texas History Underway in Tarrant County. We will crush illegal voting.” Golly. If this is the biggest voter fraud investigation ever in a state with a population of 30 million, I think we can all breathe a sigh of relief at just how honest our elections are.

After that we can go back to asking Republicans for examples of in-person vote fraud. Then, after we’ve finished watching the paint dry, the grass grow, and downloaded Facebook over a 300 baud modem, we can go about our business. Because they don’t have any examples. As in this case, it always turns out to be something else.

The LA Times has an interesting op-ed today about housing prices. It’s written by a professor and student pair at UCLA who say that estimates of housing prices are generally much higher than reality:

The inflated numbers come from private firms such as Zillow, Reis, Apartment List and other real estate brokerage firms, which generate data that are often cited by reporters and political candidates. But the data from these firms, while accurate for their individual listings, can be wildly inaccurate measures of overall price levels because they are based upon only part of the market.

Sensibly enough, they recommend using figures from the Census Bureau’s American Community Survey instead. I didn’t know the ACS tracked this, but indeed they do. For example, here is average rent since 2005 in the Big Three expensive cities:

That…doesn’t look so bad. I used gross rent just because it was easy, but the ACS also allows you to search, say, for 2-bedroom apartments or by the year the structure was built. Generally speaking, though, the UCLA folks are right: most pieces I read about housing in big cities suggest that a typical apartment in San Francisco will set you back $3,000 or more. But apparently not.

On the other hand, if you look at rent as a percentage of income, things are a lot worse:

The ACS provides this directly, but instead I did this chart by hand using national median income, not median income in each city. After all, median income in big cities is artificially high because only people with high incomes can afford to live there, which skews the picture in a way that’s not easy to fix. Instead, this chart shows, on average, what percentage of your income it’s likely to cost you if you decide to move to one of these cities. It also shows that rent as a percentage of income in big cities has increased by about 10 percentage points. That’s because median income hasn’t grown much over the past couple of decades.

These are very basic charts that I threw together just out of curiosity. If you want to be more careful, you can eliminate things like extremely low rents charged to family members or you can look only at rents for newly occupied apartments. If you do both of these things, median rents increase by about $300.

I don’t have any big point to make here. I just thought the data was interesting. Here’s how the op-ed concludes:

We agree that rent affordability in California is a real problem. After adjusting for inflation, median rents in California are about 50% higher than they were in 1980. This is partly because the average quality of housing is much higher today than it was 40 years ago….Housing prices have also risen because of the severe restrictions in many parts of the state on new construction — in other words, there is a housing shortage. High rental prices mean that even among households with incomes above $30,000, nearly a third devote more than 30% of their income to rent.

The best response to the affordability problem is a matter of legitimate debate. But thoughtful debate must start with accurate data.

Fair enough.

POSTSCRIPT: There’s no point in yelling at me in comments about how you live in one of these cities and these figures are obviously wrong. Go yell at the Census Bureau. But before you do, keep in mind that my readers and their circle of friends tend to have higher than average incomes and live in more expensive than average neighborhoods. So that probably biases your view of the rental market.

Do you remember that plan from the Trump administration about making pharmaceutical companies cut the price of some of their most expensive drugs to match prices in Europe? It was, oh, let’s just check here—well, it was last week. An eternity in Trump time, I know.

Anyway, the fatal flaw in the plan is that it’s all about how much Medicare pays for drugs, but Medicare has no leverage. It can negotiate all it wants, but in the end it’s required to cover virtually all drugs and it’s required to pay whatever the manufacturer says. It’s not allowed to drop coverage of a drug because it’s too expensive.

Still, you’d expect drug companies to be publicly polite about the whole thing, if only for the sake of PR. But David Lazarus of the LA Times reports that they aren’t even bothering to pretend. Pfizer’s CEO, Ian Read, announced a 45 percent earnings increase this week and was then asked if it would reconsider price hikes in January due to Trump’s proposal:

“I expect our approach by the end of year will be, what I would characterize as business as normal,” Read answered during a conference call with analysts. “We price to the marketplace,” he said. “We price competitively, and we will make those decisions towards the end of the year and early in January.”

In other words, no more Mr. Nice Guy. Pfizer will once again reach as deeply as possible into people’s pockets, regardless of what President Trump might want. Because let’s face it: For all his talk of drug companies “getting away with murder,” Trump has been all bark and no bite when it comes to sky-high drug prices. And the industry knows it.

In fairness to Trump,¹ there’s little he can do on his own. Only Congress can change the law that requires Medicare to cover everything, and Trump has never demonstrated either the desire or the ability to sway Congress. Not that it would probably matter: congressional Republicans just won’t do it, no matter how much their constituents might like the idea. So in only a few days, Trump’s plan is already effectively dead, and he’s unlikely to care much about it after the election is over.

¹See? I can be fair to Trump once in a while.