From the Wall Street Journal:

Not Everyone Wants to Shop on Amazon

[Chris] Outwater is among the 17% of U.S. primary household shoppers who say they never shop on Amazon, according to data from Kantar Retail ShopperScape. While the percentage has steadily declined over the past five years, roughly 22 million American households didn’t use the retailer this year.

Those Amazon holdouts tend to be older than U.S. shoppers overall, with an average age of 57 versus 49, respectively, according to Kantar, and they tend to earn less—$45,700 in annual income, compared with $62,800 among all shoppers. They are less likely to have or live with children.

Talk about burying the lede. Maybe I just haven't been paying attention, but the news to me is that 83 percent of "primary household shoppers" shopped at Amazon at least once last year. Compare that to this from a June story in the Journal:

While total online spending comprised 7.8% of all retail purchases in the first quarter, according to the Commerce Department, more than half the population, or about 190 million U.S. consumers, will shop online this year, according to Forrester.

This is roughly 80 percent of all US adults, which jibes pretty well with the Kantar number. Basically, about 80 percent of all US adults shop online at least a little, and every single one of them has shopped at Amazon once or more. In other words, among online shoppers, close to 100 percent have shopped at Amazon. Is there any other retailer on the planet who can claim anything close to that number?

A few days ago I wrote about an odd Fox News report which said that SNAP (food stamp) fraud was at an "all-time high" of $70 million. The odd part of this was not that Fox was lying about SNAP fraud being at an all-time high. What do you expect from Fox? The odd thing was that actual SNAP fraud clocks in at nearly a billion dollars. Why did Fox lowball it? I couldn't figure out where Fox got its number, and as near as I could tell the Department of Agriculture hadn't published anything about it more recently than 2011 anyway.

Today, the Washington Post's Erik Wemple does some shoe-leather reporting and makes a few phone calls:

The Agriculture Department is asking Fox News to correct a report from Tuesday morning’s edition of “Fox & Friends” alleging new heights for food-stamp fraud in the United States....“We are not quite sure where this came from,” a USDA spokesperson tells the Erik Wemple Blog. “We saw that there was a story on Breitbart. We have not issued a report on this recently. There is no new rate that we’ve published. So we’re not quite sure why they’re so interested in stirring this up.”

....A Fox News spokesperson indicates that this matter will be addressed on tomorrow’s program.

OK then. It sounds like someone at Fox just made this up, though I'm sure they'll put a more positive spin on it than that. I'm also sure they'll be eager to correct the fact that SNAP fraud is actually far higher than they suggested. But given all this, why did the Post's Philip Bump include the following graphic in his story about this on Wednesday?

It's fine to write about Fox's misinformation, but why was it credited to the Department of Agriculture? That's how Fox credited it, to be sure, but that credit shouldn't have been passed along without verifying it independently.

One final note: this story demonstrates the value of being numerate. The reason I first noticed it was because the $70 million figure was so obviously absurd. SNAP is a $70 billion program, and it's nuts to think that it could have a fraud rate that low. Mother Teresa's mission in Calcutta probably had a higher fraud rate than that. Anybody with even the smallest working knowledge of SNAP, normal fraud rates, and basic arithmetic should have heard alarm bells going off immediately.

But no one did. Nearly all of the response to the Fox report was either (a) mockery of their suggestion that SNAP should be ended, or (b) mockery of their belief that 0.1 percent was a high rate of fraud. Nobody really seemed to notice that it couldn't possibly be correct. More numeracy, please.

Today the FBI and the Department of Homeland Security released a joint report directly accusing the Russians of hacking into the servers of "a U.S. political party," including a spearphishing campaign "launched as recently as November 2016, just days after the U.S. election." I'm not quite sure why they're being so precious about naming the DNC, which has gotten several miles of press coverage, but the ways of intelligence organizations are mysterious:

This document provides technical details regarding the tools and infrastructure used by the Russian civilian and military intelligence Services (RIS) to compromise and exploit networks and endpoints associated with the U.S. election....This activity by RIS is part of an ongoing campaign of cyber-enabled operations directed at the U.S. government and its citizens.

The report doesn't actually say an awful lot about how they know this hacking comes from Russia, but it does include one bit of signature code to watch out for. We also learn about lots of funky code names for Russian hacking operations:

APT28, APT29, Agent.btz, BlackEnergy V3, BlackEnergy2 APT, CakeDuke, Carberp, CHOPSTICK, CloudDuke, CORESHELL, CosmicDuke, COZYBEAR, COZYCAR, COZYDUKE, CrouchingYeti, DIONIS, Dragonfly, Energetic Bear, EVILTOSS, Fancy Bear, GeminiDuke, GREY CLOUD, HammerDuke, HAMMERTOSS, Havex, MiniDionis, MiniDuke, OLDBAIT, OnionDuke, Operation Pawn Storm, PinchDuke, Powershell backdoor, Quedagh, Sandworm, SEADADDY, Seaduke, SEDKIT, SEDNIT, Skipper, Sofacy, SOURFACE, SYNful Knock, Tiny Baron, Tsar Team, twain_64.dll, VmUpgradeHelper.exe, Waterbug, X-Agent

So what are we going to do about this? Here's the New York Times:

The Obama administration struck back at Russia on Thursday for its efforts to influence the 2016 election, ejecting 35 Russian intelligence operatives from the United States and imposing sanctions on Russia’s two leading intelligence services.

The administration also sanctioned four top officers of one of those services, the military intelligence unit known as the G.R.U., which the White House believes ordered the attacks on the Democratic National Committee and other political organizations....In addition, the State Department announced the closing of two “recreational facilities” — one in New York, another in Maryland — that it said were used for Russian intelligence activities, although officials would not say whether they were specifically used in the election-related hacks.

This brings back memories, doesn't it? It's just like the Cold War, and Russia will no doubt expel 35 Americans in a few days. However, this also puts Donald Trump on the spot. Will he reverse the sanctions on GRU when he takes office? It's one thing to do nothing and hope everyone forgets about it, but it's quite another to actively reverse sanctions that are based on the conclusions of our own intelligence agencies. What are you gonna do, Donald?

Last year, several insurance companies abandoned Obamacare because they were losing money. This year, premiums have spiked 25 percent on the exchanges. As a result, Paul Ryan says insurance markets are "collapsing," and Republicans are promising to repeal Obamacare practically on Day 1 after Donald Trump takes office.

But a funny thing has happened on the way to the collapse: Obamacare is more popular than ever. Charles Gaba is the go-to guy for Obamacare enrollment data, and the simplified chart on the right is based on his more detailed versions here and here. Last year at this time, a little over 11 million people had signed up on the exchanges. This year, a little over 12 million have signed up. Here's what this means:

With premiums up 25 percent, a little arithmetic suggests that insurance company revenue from Obamacare will be more than a third higher this year than last year.

Time will tell what this means, but it doesn't look like a death spiral. The premium hikes are painful, but clearly people find health insurance worth it even at the higher price. What's more, about 85 percent of Obamacare customers qualify for federal subsidies, and for them premiums are up only slightly—and not even that much if they also shop around.

So: insurance companies will be in pretty good shape this year. A few people have seen sharp premium increases, but the vast majority haven't. Uncle Sam is picking up most of the tab. In other words, 2017 is shaping up to be the healthiest year on record for Obamacare.

Was there a huge crime wave in 2015? There are two main sources for crime rates in the United States. The FBI produces the Uniform Crime Report (UCR), which is based on reporting from police agencies. The Bureau of Justice Statistics produces the National Crime Victimization Survey (NCVS), which conducts surveys of ordinary Americans and asks if they've been a crime victim in the past year. Rick Nevin breaks down the numbers:

The 2015 NCVS property crime rate (household burglary, motor vehicle theft, and other theft) was down 6.3% from 2014...2015 UCR burglary rate...down 8.5%...UCR larceny-theft rate...down 2.5%...UCR property crime rate...down 3.4% from 2014....roughly consistent with the NCVS data showing the property crime rate falling 6.3% in 2015 to a record low.

The UCR violent crime rate (murder, rape, robbery, and aggravated assault) should be roughly consistent with the NCVS serious violent crime rate (sexual assault, robbery, and aggravated assault),1 but the UCR violent crime rate increased 3.0% in 2015 as the NCVS serious violent crime rate fell 11.7%....

OK, hold on. Everyone agrees that property crime is down, but the FBI says the reported violent crime rate increased 3 percent while the NCVS survey data says it decreased 11.7 percent? What's going on? The biggest components of the violent crime index are robbery and aggravated assault. Both the UCR and the NCVS agree closely about the robbery rate, so that means there must be some kind of discrepancy in the aggravated assault rate:

The 2015 UCR aggravated assault rate was up 3.8% from 2014....NCVS total aggravated assaults were down 25.2% in 2015, and NCVS aggravated assaults reported to police were down 20.7%.

Yikes! Long story short, Nevin shows that this divergence between UCR and NCVS has been increasing for the past decade. The culprit, apparently, is exactly the opposite of the frequent allegation that police departments understate serious crime in order to make themselves look better. "The fact that NCVS victims are reporting aggravated assaults far below UCR recorded aggravated assaults suggests that police have become far more expansive than crime victims are when it comes to defining aggravated assault, perhaps to protect against allegations that the police undercount serious violent crime."

Most likely, then, there's a longstanding issue of how aggravated assault is reported and categorized. Basically, police departments underreported it in the past and are now overreporting it. Aggravated assault probably decreased or held steady in 2015, which means the overall rate of violent crime was also either down or steady.

There was an increase in the murder rate last year, from 4.44 in 2014 to 4.88 in 2015 (per 100,000). This is a significant jump, and it was apparently fueled by an especially large jump in about a dozen big cities. This is cause for concern, especially since the murder rate usually correlates roughly with the overall violent crime rate. The divergence last year is unusual, and we don't yet know what explains it. It might just be a random spike, or it could suggest something worse.

But while murder gets the headlines, it's only one small component of the overall crime rate. Overall property crime was down last year and overall violent crime was probably down too. These are, by far, the crimes that actually affect most people. With the exception of a few pockets of increased homicide, America continues to get safer and safer.

1The NCVS numbers don't include homicide because you obviously can't survey murder victims. However, homicide is a tiny part of the overall violent crime rate, so that doesn't account for the difference between UCR and NCVS figures.

Politico reports on Donald Trump's search for a Secretary of Agriculture:

Trump met Wednesday with two Hispanic politicians at his Mar-a-Lago resort in Palm Beach to discuss the possibility of taking on the agriculture post: Dr. Elsa Murano, a former U.S. agriculture undersecretary for food safety, who is Cuban-American, and Abel Maldonado, a Mexican-American who is a former California lieutenant governor and co-owner of Runway Vineyards.

I imagine Trump's interior monologue for his cabinet choices has gone something like this:

Lessee. Solid, silver-haired white guy for State. Check. Retired general for Defense. Check. Personal financial crony for Treasury. Check. What else? Teachers are all women, so Betsy is good for Education. Urban is code for black, so Ben will fit in at HUD. Lotta oil wells in Texas, so maybe a Texan for Energy. Perry can do it. Somebody exotic-looking for UN ambassador. Nikki really looks the part. Asians are bad drivers, maybe Elaine can get through to them at Transportation. Fill out the rest with a bunch of dull white guys. I'll let Pence take care of it. And Agriculture. Hmmm. Gotta be Hispanic, right? They're the ones who pick all the crops. But who?

If only I were just joking with this.

The nanny-staters at the EU are at it again:

E.U. regulations adopted in 2012 require that all newly manufactured trucks heavier than 3,500 kilograms (3.9 tons) be fitted with an advanced emergency braking system. The systems use radar and cameras to detect obstacles and warn the driver. In the event of a crash, the brakes may stop the vehicle entirely. Although the driver can override the brakes, an inexperienced user may not know how.

They just can't get enough, can they? This raises the cost of trucks, raises the cost of goods carried in trucks, and probably has only a minuscule—

Wait. What?

The truck attack on a Christmas market in Berlin last week may have been cut short when the truck automatically deployed its brakes, the result of European Union regulations that require automatic braking systems on large trucks.

The new detail, revealed jointly by Germany’s Süddeutsche Zeitung newspaper and broadcasters NDR and WDR on Tuesday, may explain why the truck came to a stop after a few hundred feet. In the end, twelve people were killed in the attack, but there are indications that the E.U.-mandated advanced emergency braking system may have prevented more deaths.

How about that? The regulatory state FTW.

So what has President Obama done over the past month to get a few last-minute liberal priorities in place before Donald Trump takes over? Obama has moved forward on eight substantial executive actions so far:

  • Enacted a permanent ban on offshore oil and gas drilling in areas of the Arctic and the Atlantic Seaboard.
  • Refused to veto a UN resolution condemning Israel's settlements in the West Bank.
  • Designated two new national monuments totalling more than 1.6 million acres: Bears Ears Buttes in southeastern Utah and Gold Butte in Nevada.
  • Instructed the Department of Homeland Security to formally end the long-disused NSEERs database, which Trump could have revived as the backbone of a new Muslim registry.
  • Instructed the Army Corps of Engineers to deny final permits for the Dakota Access Pipeline where it crosses the Missouri River near the Standing Rock Sioux reservation.
  • Issued a final rule that bans the practice among some red states of withholding federal family-planning funds from Planned Parenthood and other health clinics that provide abortions.
  • Finalized rules to determine whether schools were succeeding or failing under the Every Student Succeeds Act.
  • Began an investigation into charges of Russian hacking during the presidential campaign.

This last-minute flurry of activity is actually fairly normal, but Trump is annoyed anyway, saying he's doing his best to "disregard the many inflammatory President O statements and roadblocks." Too bad, Donald: there's more to come. According to Politico, "As many as 98 final regulations under review at the White House as of Nov. 15 could be implemented before Trump takes office. Seventeen regulations awaiting final approval are considered “economically significant,” with an estimated economic impact of at least $100 million a year." Here are fifteen of the most important ones:

  • A new policy making it easier to hire and retain highly skilled immigrants.
  • A new rule forcing state regulators to tighten oversight of for-profit colleges that operate online courses in their state.
  • New energy efficiency standards.
  • Regulations designed to discourage speculation on commodities trading.
  • A new rule that would regulate air pollution from the oil industry.
  • A change in the way Medicare drug payments are administered.
  • Reform of Medicare payments to doctors, moving toward a system that better evaluates the quality of care they provide.
  • Finishing up an investment treaty with China (though it would require Senate approval in 2017).
  • Speeding through a backlog of debt relief claims from students at ITT Tech and Corinthian Colleges, two for-profit colleges that went out of business under pressure from the Obama administration.
  • A ban on cellphone calls on commercial flights.
  • A rule requiring that most freight trains have at least two crew members on duty.
  • Rules for the 2018 version of the Obamacare state insurance marketplaces.
  • Regulation of methane releases from oil and natural gas wells.
  • A major rule on leases for wind and solar projects on federal land.
  • A rule that aims to ensure poor and minority students get their fair share of state and local education funding.

Some of these actions could be overturned either by Trump or by Congress, but not all of them. Congress is restrained by the fact that it has limited floor time to review new rules. Trump is restrained because agency rules go through a lengthy rulemaking process before they're finalized, and he would have to start up this entire process all over again to repeal them.

Of course, all of these actions are also susceptible to court fights, just as they always are. There's no telling how that might turn out.

We've had a busy day of Trump news. I know you all want to be on top of things, so here's the latest. First, Trump was asked what he thought about Sen. Lindsey Graham's statement that sanctions were due against Russia and Vladimir Putin for their hacking during the election. Check out his reply:

I think that computers have complicated lives very greatly. The whole age of computer has made it where nobody knows exactly what is going on. We have speed, we have a lot of other things, but I’m not sure we have the kind the security we need. But I have not spoken with the senators and I will certainly will be over a period of time.

Later, asked about Israeli settlements on the West Bank, Trump produced another bit of word salad that made it clear he had no idea what a settlement even was. This is probably why Trump hasn't spoken to the press in such a long time. This kind of callow blather might have been entertaining when it was coming from a buffoon candidate who had no chance of winning,1 but not when it's coming from the president-elect.

In other news, Politico reports that Trump was irritated by President Obama’s comments at Pearl Harbor yesterday. Obama said, "even when hatred burns hottest, even when the tug of tribalism is at its most primal, we must resist the urge to turn inward. We must resist the urge to demonize those who are different." Those are fairly boilerplate remarks, but "these felt to Trump like direct criticism of the president-elect, according to two people close to Trump." Gee, I wonder why?

Finally, Trump announced that Sprint was bringing 5,000 jobs back to America. "I just spoke with the head person," Trump told Bloomberg. "He said because of me they're doing 5,000 jobs in this country." Here's how it played in the nation's press:

The skepticism in these headlines turns out to be warranted. Trump did indeed desperately try to take credit for this, and you will be unsurprised to learn that he was lying. First of all, Sprint announced these jobs back in April. Here's the Kansas City Star: "Sprint Corp. is launching a nationwide service to hand-deliver new phones to customers in their homes. The Direct 2 You service, which first rolled out in a Kansas City pilot, will lead to the hiring of about 5,000 mostly full-time employees as it spreads nationwide."

Second, the Japanese owner of Sprint, Softbank, announced in October that it was creating a huge tech investment fund.

Third, in December, Softbank's CEO announced the fund again after a meeting with Trump, and said that one part of the whole package was the creation of 50,000 new jobs. Today, Sprint reluctantly conceded that its 5,000 jobs were part of the previously announced 50,000 jobs.

And finally, these jobs were announced yet again today.

That makes four times these jobs have been announced. Donald Trump was responsible for none of them.

1Actually, it wasn't entertaining even back then.

UPDATE: It turns out the April jobs are different from the October/December jobs. So they've only been announced three times.

A few days ago I posted an update on how the Republican economic revolution was going in Kansas under Gov. Sam Brownback. Answer: not so great. But that got me curious: what about other states that turned red in the great tea-party wave of 2010? Not just a little red, but all red. I'm talking about states that switched from blue or mixed to fully Republican—governor and legislature—and stayed red for the next six years. This would give tax-cut fever plenty of time to do its thing.

There were nine states that filled the bill, and then I added Louisiana even though it only stayed fully red for four years. I did this because Bobby Jindal was such an annoying loudmouth about his conservative governance, but I was fair: I counted just the four years of full redness to his record. He didn't get any of the blame for the next two years.

Anyway, the results are below. Only three of the ten states managed to beat the national average. That whole supply-side thing might need a little work.

UPDATE: I forgot Michigan! Fixed now.