Feed aggregator

Has the United States reached peak incoherence in foreign policy?

President Obama spent seven years expanding the war on terror and intervening inconsistently and incoherently in the Middle East, only to acknowledge in recent interviews that Libya was his greatest mistake. The frontrunners in the presidential campaign are no better. Hillary Clinton supported the Iraq War before she later opposed it, and promoted the TTP and TTIP while Secretary of State but now says they’re a bad idea. Donald Trump’s foreign policy views have doubled back on themselves so often it’s hard to tell where he stands.

Observers on both the right and the left agree that the United States has lost its vision of how to succeed in the international arena. This makes it impossible to craft sound strategy or to build any sort of public consensus around it. What the United States needs is a new paradigm to help cut through the many conundrums that currently have the United States flummoxed. This new vision must clarify the primary goals of American foreign policy and identify the appropriate means for achieving those goals.

Crafting consensus around such a vision will be challenging, however, and in the meantime we need strategies to save us from our worst tendencies towards threat inflation, overspending, and excessive military intervention around the world.

Today I would like to propose a mechanism for doing just that. I call it the Switzerland Test. Applying the test is simple. When assessing threats, making decisions about defense budgets, or thinking about whether to intervene in another civil war, our political leaders should just ask: What would Switzerland do?

Switzerland provides a compelling vision for American foreign policy for several reasons. First, the Swiss assess security threats rationally. The Swiss are lucky, surrounded by mountains as they are, which has allowed them to fend off would-be invaders and occupiers for most of their history. Even Hitler didn’t bother. And today, the Swiss share borders with friendly neighbors. As a result, the Swiss waste little time or money on unnecessary national security initiatives. The most heated security debate in Switzerland recently has been whether or not even to have an army.

The United States will surely keep its army, but notice the parallels here. The United States, like Switzerland, is surrounded by imposing natural obstacles and friendly, militarily weak allies. Beyond this the United States also enjoys the safety of a secure nuclear deterrent. Unlike Switzerland, unfortunately, the United States sees threats everywhere and thus the legacy military-industrial complex from the Cold War continues to rumble along, sucking up trillions of dollars that the private sector economy could put to far better use.

Second, Swiss foreign policy is mostly geared toward improving its economic well being by expanding international trade through the development of cooperative agreements. And because it is so tiny, Switzerland has had to become competitive economically through investments in its people, its educational infrastructure, and its physical infrastructure.

The United States could learn a lot from this effort as well. Rather than focus on military power, which costs a lot but does little to advance the nation’s economic interests, the United States should focus on improving its economic power. Rather than flirt with protectionism like Clinton and Trump have done throughout the campaign, the United States should embrace the cause of free trade much more fully.

Third, Switzerland accepts the world as it is and does not seek to change it or to control it, only to make money by working with it. Unlike the United States, Switzerland does not seek to control the behavior of other nations, to resolve other nations’ internal conflicts through force, or to reshape the world to make it safe for democracy.

To be sure, as a tiny country with zero ability to project military power, Switzerland has little choice in this regard. But the benefits to Switzerland are many. In the United States pundits decry the turmoil in the Middle East and worry about which group of moderate rebels to support in Syria, Yemen, Libya, or Iraq. Does anyone think Switzerland won’t end up making money by trading with whichever government winds up emerging? There is simply no reason for the United States to struggle so hard against what is. Instead, like Switzerland the United States should accept the world as it is and work to get rich and happy by working with the world.

Finally, Switzerland ensures that it will not wind up doing stupid things by staying neutral and maintaining its strategic independence. Rather than entangling themselves in the EU or in NATO, the Swiss stay neutral in most things but remain willing to cooperate to solve problems on a situational basis.

The United States would benefit from the Swiss example here as well. Though the threat of Soviet attack justified American participation in NATO during the Cold War, today there is no reason for the United States to hitch its wagon to Europe’s security. Today, neither Russia nor the Islamic State poses the sort of military threat that requires American assistance. Likewise, though China certainly scares its neighbors in the Pacific, but many of them can take care of themselves and none of them are worth the United States risking major military conflict with China. The United States has more to gain by trading with China than confronting China.

The Switzerland Test is simple, clear, and easy to apply. The primary goals of U.S. foreign policy are national security and prosperity. Since the security angle is almost entirely taken care of by geography and nuclear weapons, most U.S. foreign policy should be focused on expanding America’s economic connections with the rest of the world and enabling those efforts with the appropriate investments in education and infrastructure.

When confronting a foreign policy question, the United States just needs to ask: What would Switzerland do?

Two experiments in a new working paper find some evidence that some drivers using popular ride-hailing platforms discriminate against riders. In some cases African-American riders faced longer wait times or higher probabilities of having a driver cancel on them. To some extent, these findings should temper hopes that these new technologies and platforms would succeed in quickly rooting out these forms of discrimination. It is important to note, however, that this study examined whether there was discrimination within these platforms, it did not compare ride-hailing platforms and traditional taxi companies. These new platforms do offer some advantages over the status quo when it comes to identifying the channels of discrimination through richer data, while the rating system gives riders some recourse to penalize discriminating drivers and an incentive for drivers to maintain a high rating.

In the study the researchers undertook two large-scale experiments to try to determine whether there was a pattern of discrimination for riders using these ride-hailing platforms.

In the Seattle experiment research assistants of different racial backgrounds requested rides along pre-determined, randomized routes. They logged speed of service, directness of the route, and the passenger ratings received. For each of the services analyzed (UberX, Lyft, and the traditional taxi-hailing app Flywheel), African-American riders had longer acceptance times than white riders. In UberX, they also waited about 30 percent longer to be picked up, while there were no significant differences in the other two.

In Boston, research assistants varied the names they used to request rides, with one more “white sounding” and the other a “distinctively black name.” Here wait times and acceptance times did not differed significantly, in one platform the rate of trip cancellations did: in UberX the rate of cancellation for hailers using African American-sounding names was 10.1 percent, compared to 4.9 percent for the white-sounding names, while there were not signs of racial discrimination in Lyft (and women using the African-American names actually had a lower cancellation rate).

Some of this variation in discrimination might be due to differences in the way platforms provide rider information to the drivers. Lyft drivers see the passenger name and photo (if the passenger has chosen to upload one into their profile) before accepting a ride request, in UberX the drivers only get a name after accepting, and in Flywheel there are no passenger photos in traveler profiles at all. The authors suggest that platforms could alter the timing and content of the rider information provided to reduce the possible channels for discrimination.

Again,  this study examines whether there are signs of discrimination within these ride-hailing platforms, that is, differences in wait times, routes, and acceptance rates by gender or race. As the authors are careful to point out, they “do not claim that [transportation network companies] are “worse” than the status quo.”

The authors do provide one brief illustration of the discrimination present in traditional taxis, and find that there is also a substantial different in acceptance rates for traditional taxis when research assistants of different racial background tried to hail a cab.  The first taxi stopped roughly 60 percent of the time for a white hailer, compared to less than 20 percent of the time for an African-American rider. White riders never had more than four taxis pass before one stopped, while African-Americans saw six or seven taxis pass by in 20 percent of the time. So there is unfortunately a substantial amount of discrimination in the traditional taxi industry, but it is has been opaque and harder to measure or understand.  Also, riders who are discriminated against in those instances have fewer options to choose a different platform, give the offending drivers negative feedback, or provide information to the taxi company to make the extent of the discrimination clear.  

Number of Traditional Taxis Passing Waiting Travelers, by Race

Source: Ge et al. (2016).

While these ride-hailing platforms may offer many improvements over traditional taxis when it comes to convenience and cost, this new paper shows that some individual drivers discriminate against riders on the basis of their race or gender. It’s not clear how this level of discrimination compares to traditional taxis, and the study shows that there is a significant amount in the old framework as well. These new platforms make it easier for riders and companies to identify which drivers are doing this, and they give riders more recourse to penalize offending drivers and more options to choose from that might be better on these measures.

Armed with this new information and a fuller understanding of how this driver discrimination is taking place, platforms can experiment with changes to the timing and content of the rider information they provide to drivers to try and address these issues. Ride-hailing platforms have not yet solved the difficult problem of discrimination in that sphere, but they might still make significant progress in reducing it, and they can do it without the need for heavy-handed new government intervention.

The Guardian reports on calls by German chancellor Angela Merkel for internet platforms to “divulge the secrets of their algorithms”:

Angela Merkel has called on major internet platforms to divulge the secrets of their algorithms, arguing that their lack of transparency endangers debating culture.

The German chancellor said internet users had a right to know how and on what basis the information they received via search engines was channelled to them.

Speaking to a media conference in Munich, Merkel said: “I’m of the opinion that algorithms must be made more transparent, so that one can inform oneself as an interested citizen about questions like ‘what influences my behaviour on the internet and that of others?’.

“Algorithms, when they are not transparent, can lead to a distortion of our perception, they can shrink our expanse of information.”

An algorithm is the formula used by a search engine to steer a request for information. They are different for every search engine, highly secret and determine the significance or ranking of a web page.

Merkel has joined a growing number of critics who have highlighted the dangers of receiving information that confirms an existing opinion or is recommended by people with similar ideas.

“This is a development that we need to pay careful attention to,” she told the conference, adding that a healthy democracy was dependent on people being confronted by opposing ideas.

“The big internet platforms, through their algorithms, have become an eye of a needle which diverse media must pass through [to access their users],” she said.

My sense is that some Europeans are frustrated at how American companies dominate many aspects of the Internet. However, instead of trying to compete with the American companies in the marketplace (which would be a welcome development, as more competition is good), they have decided that regulating these companies (e.g., through antitrust scrutiny) is their best strategy for reducing American dominance.

Here, the demand to divulge algorithms strikes me as very odd. Would Chancellor Merkel ask newspapers such as Der Spiegel to divulge the decision-making process for how they decide to present information to their readers? Of course not. That would be absurd. Similarly absurd is telling Google and Facebook to divulge their algorithms for presenting information.

Now, these companies might have their own incentives to be transparent about their decision-making, in response to market forces. But there’s no need for governments to get involved with any of this, and Chancellor Merkel’s efforts may be mostly based on anti-Americanism and frustration with losing in the competition for Internet dominance.

They’ll be watching you: King County (Seattle) uses grocery loyalty card data to figure out who owns pets, according to a new report from local station KOMO. It then sends them letters warning of a $250 fine if they do not license the animals. The “county said they pay the company who pays stores such as Safeway …for access to customer data contained in every one of those reward card swipes.” And “the mailers work. Just last year they brought in more than $100,000 in new pet licenses.”

But remember, government needs access to Big Data to fight terrorism.

In 2008, Georgia’s General Assembly enacted the Qualified Educational Tax Credit Program in an effort to expand educational opportunities for schoolchildren and provide alternatives for parents concerned about underperforming public schools. Under the program, individual and corporate donors can receive a credit against their state income tax liability in exchange for contributions to qualified, nonprofit Student Scholarship Organizations that aid Georgia families in paying tuition at qualified private schools of their choice.

Unfortunately, opponents of school choice are once again trying to restrict parents’ ability to select the best education for their children. Because many of the scholarship students use them to attend religiously affiliated schools, the plaintiffs in this case argue that the tax-credit program entangles government in religion. Specifically, they claim that the program violates the Georgia constitution’s No-Aid Clause—one of the historically anti-Catholic Blaine Amendments—which forbids the taking of money “from the public treasury, directly or indirectly, in aid of any church, sect, cult, or religious denomination or of any sectarian institution.” They also allege a violation of the Gratuities Clause, which says that “the General Assembly shall not have the power to grant any donation or gratuity or to forgive any debt or obligation owing to the public.” Several families who have benefitted from the program, represented by the Institute for Justice, have intervened to defend the law.

The trial court held that plaintiffs lacked standing to challenge the tax-credit program. It further ruled that, even if they had standing, plaintiffs’ constitutional arguments failed because tax credits are not government funds. Violations of the No-Aid Clause require that public funds be spent in aid of a sectarian institution, and the Gratuities Clause could not have been violated because “the General Assembly cannot donate or give what it does not own.” Plaintiffs appealed and Cato has now filed an amicus brief, in collaboration with Neal McCluskey and Jason Bedrick of our Center for Educational Freedom, before the Georgia Supreme Court.

We urge the court to affirm the determination that the tax-credit program does not violate the state constitution, focusing on the fact that it does not involve spending public funds for any sectarian purpose. Because the program makes no expenditures from the public fisc, it cannot violate the No-Aid Clause. Taxpayers choose to donate voluntarily using their own private funds and receive a tax credit for the amount of the donation; no money ever enters or leaves the treasury.

The challengers attempt to get around this fact by claiming that the credits constitute an indirect public expenditure, but this argument relies on a budgetary theory known as “tax expenditure analysis” that finds no support as a legitimate means of constitutional interpretation under Georgia (or federal, or any other state) law. Indeed, the U.S. Supreme Court rejected this type of reasoning in Arizona Christian School Tuition Organization v. Winn (2011).

The argument that the program constitutes an unconstitutional gratuity is likewise incorrect because the tax credits are not public funds, and the government cannot give away that which it does not own. Even if Georgia were giving up something of value, it would not be a “gratuity” because the state receives a substantial benefit in return: increased educational attainment, plus the secondary effects that increased competition and a more educated citizenry create.

The Georgia Supreme Court should affirm the lower court’s decision and uphold the state’s Qualified Educational Tax Credit Program—ensuring educational choice for Georgia families, regardless of how much money they make.

This week and last, the Cato Institute filed amicus briefs urging the Supreme Court to take up two cases dealing with the constitutional status of “cell site location information,” or “CSLI.” This data, collected of necessity by cellular communications providers, creates detailed records of their customers’ movements. The briefs invite the Court to accept these cases so it can revise Fourth Amendment practice to eschew doctrine and more closely adhere to the language of the Fourth Amendment.

The Fourth Amendment states that “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated.” Presumably, when called upon to determine whether a Fourth Amendment violation has occurred, courts would analyze the elements of this language as follows: Was there a search? Was there a seizure? Was any such search or seizure of “their persons, houses, papers, [or] effects”? Was any such search or seizure reasonable?

And in cases involving familiar physical objects, courts usually do a sound textual analysis, at least implicitly. But in harder cases dealing with unfamiliar items such as communications and data, courts retreat to “reasonable expectation of privacy” doctrine that emerged from Katz v. United States in 1967, and offshoots of it like the “third-party doctrine.” The “reasonable expectation of privacy” test asks whether defendants’ feelings about things government agents accessed were reasonable. The corollary “third-party doctrine” cancels Fourth Amendment interests in information and things that are shared on the theory that expectations of privacy evaporate in that context.

The “reasonable expectation of privacy” test is the product of one non-essential concurrence in Katz, and the third-party doctrine was wrong when the Supreme Court created it in 1976 to ratify a law that deputized banks into financial surveillance. That doctrine grows further out of synch with each step forward our society takes in modern, connected living. Today, third-party service providers collect incredibly deep reservoirs of information about us: Cellular telephone networks, Internet service providers, search engines, and payment systems have data that can throw open windows onto our relationships, feelings, health conditions, business dealings, sexuality, emotions, and more.

In United States v. Carpenter, the U.S. Court of Appeals for the Sixth Circuit turned aside the appeal of two men convicted of armed robbery, whose CSLI had been used as evidence against them. Relying most heavily on an application of the “reasonable expectation of privacy” test in Smith v. Maryland, the court said that, while the contents of communications are protected by the Fourth Amendment, routing information is not. The court backhandedly dismissed the crucial question of whether the defendants had a property right in the data the government had seized, even though contracts and regulation both allocate property rights in data about communications use to consumers.

And in a similar case, Graham v. United States, the Fourth Circuit held that CSLI isn’t constitutionally protected because “an individual enjoys no Fourth Amendment protection ‘in information he voluntarily turns over to [a] third part[y]’” (quoting Smith v. Maryland).

Rather than relying on misshapen precedent as the appeals courts did, the Supreme Court should find that communications and data are items that can be seized and searched. Consistent with precedents both longstanding and recent, the Court should recognize that telecommunications customers can have property rights in such data, and that when the government seeks to seize and search such data, it generally requires a warrant. This will permit courts below to address seizures and searches of communications and data forthrightly, confidently assessing the reasonableness of government searches and seizures even when communications and data are involved.

The briefs join a merits brief filed this summer in United States v. Benbow, a CSLI case pending in the D.C. Circuit.

A large number of Republican candidates are openly opposing the Donald Trump’s immigration position. As I’ve noted before, 10 of the 11 GOP Senate candidates have campaigned on pro-immigration platforms. In the closest 40 House races according to Cook Political Report’s ratings, a majority of Republicans representing have already expressed openness to a pathway to legalization for unauthorized immigrants.

The large number of supporters of a compassionate immigration policy could be seen as surprising not only because the Republican presidential nominee has strongly opposed this approach, but also because only 6 of the districts are currently held by Democrats who back a legalization measure. However, it is important to note that these members are not out of step with their Republican constituents who, polls show nationally and on a state by state basis, support allowing unauthorized immigrants to stay.

Because some members were not directly asked about their views on legalization, support could be even greater, but here are the 21 members in tight races who have taken humane, pro-immigration positions:

1. Don Young (AK): “We want our country’s 11 million undocumented individuals to be welcomed and to have a place to belong – free of fear…. Immigrants have always recharged our country have been reliable sources of economic growth, cultural diversity and innovation…. like our colleagues drafting the bill, we believe this legislation should… provide a clear and responsible path to citizenship for those already here.”

2. Scott Jones (CA-7): “For those folks who are here illegally, I would advocate, as I have always advocated, a pathway to legal status for each any every one of them if they can pass a background check.”

3. Denham (CA-10): “I believe our immigration system is broken and in need of real and effective reform. I support providing an earned path to citizenship for those who want it…. Reform will ensure that all undocumented immigrants are added to the tax rolls, ensuring that everyone pays their fair share.”

4. David Valadao (CA-21): “Immigration reform is something I’m still very supportive of and continue to work on behind the scenes…. Every bill that I’ve been a part of is about allowing immigrants who are here to go through a process to become legal… For the people who are working hard in the fields, in the restaurant industry, in the service industry, I mean we’ve got to come up with a system that addresses the 11 million who are here. You have to a process that makes sure that you have guest worker programs that work, visa programs that work.”

5. Darrel Issa (CA-49): “Beyond border security, any reform package must make an immediate determination of who stays and who goes, based on our national interests. Those who demonstrate the ability to contribute to our society in a meaningful way should have a path forward to guide them, be placed at the end of the legal-immigration line, meet the strict standards established and face a rigorous but fair application process. Those who are migrant workers should be put into a temporary guest-worker program.”

6. Coffman (CO-6): “Immigration reform… has to be compassionate about keeping families together…  I cosponsored legislation this summer that would give [immigrant children of unauthorized immigrants] a legal status and then a path to citizenship… For the adults who knowingly broke the law who are here today, I think they ought to have the opportunity to come out of the shadows and have a legal status.”

7. Jolly (FL-13): “I support… comprehensive immigration reform that has remediation and penalties and so forth…. I don’t support a pathway to citizenship for people who came here illegally. But I do support a pathway to legal status and residency.”

8. Brian Mast (FL-18): “I do not support a pathway to citizenship… But I am open to a conversation of a pathway to legal work status for people who are here in the United States.”

9. Curbelo (FL-26): “I have been for comprehensive immigration reform… We need to have a guest worker program… We need to create a path to citizenship for those families who are undocumented but are contributing to our economy. A lot of people talk about the undocumented but they don’t realize that these are some of the hardest workers.”

10. David Young (IA-3): “Why are they coming in illegally? It’s probably because our legal system is so messed up… We need to put a human face on it, have some compassion… allow them to come out of the shadows, and have that guest worker program where they are allowed to work, receive a paycheck, not worried about going to jail, so they can live in peace with their families… I am not one who wants to round up and deport people.”

11. Dold (IL-10): “Last year, I had the honor of bringing a Round Lake DREAMer named Erika Martinez to the State of the Union Address. Every day that Congress fails to act on immigration reform, families like Erika’s are forced to continue living in fear of being torn apart. Although divisive, partisan rhetoric may make for better TV, today’s forum is proof that there is bipartisan support to move immigration reform forward.”

12. Tim Walberg (MI-7): “There’s a comprehensive plan that’s needed… we need to understand that without immigrants we don’t have the resources for the jobs that are here.” Walberg said there needs to be a “cooling off” period to stop “some of this anti-immigrant, anti-refugee rhetoric that’s out there.” “They’re in the agricultural industry and construction industry and high tech industry. There are people that could add to this melting pot in very positive ways.” WKAR: “Many of those people we’re talking about giving a pathway to citizenship aren’t necessarily asking for citizenship. Many are simply asking for a temporary work opportunity.”

13. Erik Paulsen (MN-3): Via MPRNews: He is open to giving illegal immigrants citizenship or permanent residence. “It’s not a realistic proposition to round up 11 million people and send them out of the country,” Paulsen said.

14. Don Bacon (NE-2): “Once we do that and employer enforcement and some other measures to secure the border, I would take a more moderate approach to handle the 12 million folks who are here… Some people are good neighbors are doing good in their community, we ought to provide a pathway to legalization.”

15. Hardy (NV-2): Hardy said he believes most of them don’t want to become citizens but want to become legal so they can work and live here without fear of deportation. “I think you give them a pathway to legal status, not necessarily to citizenship.”

16. John Faso (NY-19): “Farm employers must also have better access to seasonal labor. I don’t support citizenship for those illegally in the country; I would support legal status however — but not citizenship — for those who have not committed crimes and are willing to pay a fine. We should emphasize immigration for those who bring specific skills beneficial to the economy.”

17. John Katko (NY-24): “You get back to what immigration is all about. That’s that big group of people who are leading law-abiding lives here, maybe even starting families here and are employed. Those individuals we have to have a discussion about. And the discussion is really going to boil down to do you create a path to citizenship for millions of people who are here illegally or do you do some sort of resident alien status for them? I’m leaning towards the latter, but I still want to discuss it more with individuals.”

18. Mike Fitzpatrick (PA-8): “Once you secure the border, the next question is how to deal with the undocumented immigrants that are in this country, and I believe that they need to be dealt with humanely that are becoming of American values. Immigration is not a bad thing. It’s a good thing. We are a nation of immigrants. I am the grandson of Irish immigrants… Mass deportation is a silly idea. It shouldn’t even be discussed. It’s not workable. It’s not humane.”

19. Lloyd Smucker (PA-16): “If you’re here, you want to work hard, provide for your family or future family and want to contribute, we should encourage that… We are a nation of immigrants. We should encourage people who want to come here, who want to work hard, who want to give back, who want to provide for their families. That’s what has made America great in the first place. We need a comprehensive immigration reform that starts with securing our borders and that provides opportunities for those who want to live the American Dream… I don’t want you taking from the system. I want you to be able to get a good job. I want you to be able to pay taxes. I want you to be able to be part of society here.”

20. Will Hurd (TX-23): “We need to increase access to legal immigration. We are a nation of immigrants. We have benefited from the brain drain of every other country for the last couple of decades. I want to see us continue that. But I also want to see us benefit from the hard workers drain too. If you’re going to be productive member of our society, I want to get you here as quickly as possible. I’m against giving people who broke the law a special pathway to citizenship, but I think there is a way to get them into the system… whether paying fines and back taxes, making sure that they have a job and are going to be educated.

21. Mia Love (UT-4): Question: Should there be a path to citizenship for undocumented residents? “I am committed to immigration and legal immigration… We have to secure the border… We also have to track people… It is our job to make sure that we have a front door in, and we close the backdoor. In order for us to fix it and give people opportunities whether they want to become citizens or just have legal status, we have to fix those three first…. This is an area in which we can show our compassion as Americans and fix the problem permanently…. If we fix it permanently, we would make sure that families don’t have to go through the difficult situations that they are going through, that they don’t know where to go right now.”

I’m in Sweden today, where I just spoke before Timbro (a prominent classical liberal think tank) about the US elections and the implications for public policy.

My main message was pessimism since neither Donald Trump nor Hillary Clinton support genuine entitlement reform.

But I’ve addressed that topic many times before. Today, motivated by my trip, I want to augment my analysis about Sweden from 10 days ago.

In that column, I highlighted some research from Professor Olle Kranz showing that Sweden became a rich nation during a free-market era when government was relatively small. And as you can see from his chart (I added the parts in red), this is also when per-capita economic output in Sweden caught up with - and eventually surpassed - per-capita GDP in other advanced countries.

Then Sweden began to lose ground. Some of this was understandable and inevitable. Sweden didn’t participate in World War II, so its comparative prosperity during the war and immediately afterwards was a one-time blip.

But the main focus of my column from last week was to show that Swedish prosperity began a sustained drop during the 1960s, and I argued that the nation lost ground precisely because statist policies were adopted.

In other words, Sweden enjoyed above-average growth when it relied on policies I like and then suffered below-average growth when it imposed the policies (high tax rates, massive redistribution, etc) that get Bernie Sanders excited.

Today, let’s build upon Professor Kranz’s analysis by extending his calculations. He did his research in the early part of last decade, and we now have many years of additional data that can be added to the chart.

But before doing that, it’s worth noting that the years of additional data basically coincide with a period of market-oriented reforms in Sweden. A study from the Reform Institute in Stockholm explains some of what happened, starting with the stagnation caused by the era of big government.

The seventies and eighties saw Sweden’s tax burden rise from an average European level to the world’s highest. The public sector expanded vastly. All facets of the welfare system were made more generous in international comparison. Meanwhile, labour market regulation increased… Throughout these years, Swedes’ individual after-tax real income stagnated, private sector job creation ceased, and public debt spiralled higher. This culminated in a severe economic crisis in the early 1990s. By then, Sweden had fallen to 14th place in the GDP per capita rankings of OECD countries.

That’s the bad news.

The good news is that this economic misery led to market-oriented reforms.

When the onset of the financial crisis coincided with election of a market-oriented centre-right government in 1991, the reform process began in earnest. Most emphasis at the time was placed on reforms that opened significant sectors in the economy to greater competition. Moreover, an important feature of these regulatory reforms was that the crisis spurred local authorities to implement less burdensome regulation. …significant changes were introduced to the tax system, macroeconomic policy framework, and social insurance system. …every aspect of the Swedish economy has changed due to implementation of reforms. …public sector employment has declined.

To be sure, none of the means Sweden became Hong Kong. It is currently ranked only #38 by Economic Freedom of the World, and its score only improved from 6.92 in 1990 to 7.46 today, hardly a huge jump.

But we nonetheless can now check whether this period of modest reform yielded any dividends. And, looking at an updated and extended version of Professor Kranz’s chart, there certainly seems to be a clear relationship between pro-market policy and Swedish prosperity.

Call me crazy, but it seems like there’s a lesson here about the right recipe for growth.

P.S. The 16 countries in the comparison are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Norway, Japan, Sweden, Switzerland, the United Kingdom, and the United States.

P.P.S. If you’re interested in more information about market-oriented reforms in Sweden, check out Lotta Moberg’s video and Johan Norberg’s video.

A recent piece by Mark Krikorian, executive director of the Center for Immigration Studies (CIS), argued for mandatory E-Verify as “an important enforcement tool” and metaphorical “wall” that would prevent the hiring of illegal immigrants.  Krikorian did not mention many of the problems with E-Verify so I will do that here after a brief description of the system.

The 1986 Immigration Reform and Control Act (IRCA) created the rudimentary employment verification known as the I-9 form that every new employee must fill out.  An E-Verify mandate would add another lay on top of the I-9 whereby employers, after collecting I-9 those forms, would enter the information on the form into a government website.  The system compares these data with information held in the Social Security Administration (SSA) and Department of Homeland Security (DHS) databases.  The employee is work authorized if the databases decide that the data are valid.  A flag raised by either database returns a “tentative non-confirmation,” requiring the employee and employer to sort out whatever error has been flagged.  If the employee and employer cannot sort out the errors then the employer must terminate the new employee through a “final non-confirmation.” 

First, Krikorian erroneously labels E-Verify as a “free online system.”  E-Verify is not a gift from heaven, it was created by the federal government and funded by taxpayers.  E-Verify is also not free because of the opportunity cost of the using the system.  The current I-9 form costs employers an estimated 13.48 million man-hours each year.  A national E-Verify mandate would add to that – perhaps substantially.  Those are a lot of hours that employers could otherwise spend on growing their businesses but instead must waste complying with government rules. 

Most E-Verify checks do not take much time but 46.5 percent of contested cases in 2012 took DHS eight work days or more to resolve.  During that time, employers are justifiably reluctant to train new employees who might not be work authorized.  Employers will likely avoid that cost by pre-screening job applicants and rejecting those who come back as tentative non-confirmations.  Workers could thus get rejected from every job they apply for but not know a simple and correctable error in the E-Verify database is the reason.  Although pre-screening employees would be illegal under a national E-Verify mandate, we shouldn’t expect it to work because the entire point of the system is to stop illegal behavior by employers in the first place.

Second, E-Verify is ineffective at detecting illegal immigrant workers. On top of that, E-Verify’s accuracy rates are notoriously difficult to judge.  An audit of the system by the firm Westat found that an estimated 54 percent of unauthorized workers were incorrectly found to be work authorized by E-Verify because of rampant document fraud.  E-Verify relies upon the documents presented by the workers themselves to their employer.  Frequently, identity information comes from deceased Americans – a loophole the government seems incapable of closing.  For instance, SSNs for roughly 6.5 million Americans who are 112 years old or older do not have a death date attached which means they can easily be used by illegal workers and nobody would complain.  An illegal worker using the SSN of a deceased American would likely end up work authorized.

Another gap in E-Verify’s wall is far more difficult to fill.  Many employers ignore E-Verify even when it’s mandated, just like they ignore other government immigration enforcement rules. Alabama, Arizona, Mississippi, and South Carolina all mandate usage at the state level, yet usage and enforcement have been lax.  In 2014, only 56 percent of employers in Alabama, 57 percent in Arizona, 43 percent in Mississippi, and 54 percent in South Carolina used E-Verify for new hires despite their state laws mandating that 100 percent of employers must use the system.  Since E-Verify isn’t enforced in states like Arizona currently, we can reasonably assume that there is no hope that it will be well-enforced nationwide.  E-Verify, like every enforcement system, must actually be enforced by people and these states show about as little interest in doing so as the federal government does in enforcing I-9 requirements.  Any of the proposed national E-Verify mandates will not change that dynamic.    

Third, some Americans would be kicked out of the labor market due to E-Verify.  E-Verify’s accuracy problems mean that Americans are and will continue to be barred from work due to false positives.  Roughly 0.15 percent of all E-Verify queries result in a false “final non-confirmation.”  While that is an admittedly small percentage, if applied nationwide to an American labor pool of roughly 125 million workers, it would result in 187,500 wrongly issued FNCs to American workers each year. 

Fourth, E-Verify is supposed to help curb illegal immigration by turning off the jobs magnet.  In the real world, E-Verify barely affects the wages of suspected illegal immigrants.  In Arizona, the E-Verify mandate lowered the expected wage gain of immigration from Mexico from 253 percent to 241 percent – hardly diminishing the strength of the wage magnet.  That small effect could even overstate E-Verify’s effectiveness because it includes a period of time when employers and employees were learning how to get around the system. A national mandate in the near future would confront many millions of employers and illegal immigrants who now know how to get around the system thanks to Arizona and other states.

Fifth, CIS’ chosen legislative vehicle for mandatory E-Verify is the so-called Legal Workforce Act.  Versions of this bill have been proposed in every new Congress for years and each version has failed to make it to the floor for a vote.  A new version of the bill is likely to be proposed next Congress and it is likely to be as expensive and problematic as previous versions. At the federal level, expanding E-Verify through the Legal Workforce Act would, per the CBO, cost the federal government $635 million from 2018 to 2023 while imposing $10 million in annual compliance costs on states and localities. The CBO also estimates a minimum $200 million cost to the private sector following passage, as employers are forced to verify an additional 50 million employees.  That’s a high price tag for a supposedly “free” program.

Sixth, government identity requirements for employment incentivize identity theft.  A huge cottage industry of forged identity documents sprung up after the government first mandated that employers check the identification of new hires in 1986.  Just as IRCA gave a big boost to the black market 30 years ago, nationally mandated E-Verify would subsidize it even further.          

Seventh, E-Verify’s errors and loopholes mean that the system will be quickly rendered useless if it is ever mandated universally – which is the most positive thing I’ll say about E-Verify until it is significantly reformed.  The government will not react to E-Verify’s failure by throwing up their hands and calling it a day.  The government would then integrate other biometric information like fingerprints or perhaps even DNA into a national identity system to close the E-Verify loopholes in an attempt to make the system more effective.  Such a beefed up E-Verify system could easily be used for other purposes that its proponents would be horrified by – like creating a national gun registry.  It is unwise to mandate participation in a new government identity tool that will expand in the future, especially in an era of serious privacy scandals.       

Interestingly, the government is already starting to link E-Verify to DMV photos in some states to reduce fraud.  Mississippi is one of those states but, as I noted above, it has the worst rate of employers actually checking new employees through E-Verify.  A fancy new biometric identity system combined with E-Verify does not exclude illegal immigrants from the workforce if nobody uses it.

This post will end with a few value judgments.  The government should enforce its own laws rather than conscripting employers in its attempt to regulate the labor market.  If the government cannot enforce its own laws then that is a signal that its laws should change.  Also, Americans should not have to ask government permission to work.  Krikorian will respond that Americans already have to ask through the I-9 form.  In reality, current workers fill out a government form that nobody ever sees again before getting a job.  That is very different from having a government computer reject or accept a worker for employment.

There are many proposed fixes to the problems above but they will likely be just as disappointing as the lockdowns on SSNs.  Many Americans are opposed to mandatory E-Verify not out of some Democratic Party-inspired desire for open borders but rather because they are justifiably skeptical of an error prone, easily fooled, and deceptively expensive federal mandate that has great potential to diminish the privacy of Americans and lead to a national biometric identity program.  Krikorian and other E-Verify supporters should address the real arguments against E-Verify rather clumsy strawmen.      

The German Marshall Fund has just published an essay that I wrote in conjunction with a conference convened in Berlin six weeks ago. For a few hours in September, as Donald J. Trump’s poll numbers hovered around parity with Hillary Clinton’s (that is, before they fell and rose again), an array of American and European scholars contemplated Trump’s emergence as the de facto leader of one of the United States’ two major political parties.

What, if anything, did Trump’s rise portend for the future of transatlantic relations?

“Not much,” would be the safe, short and simple answer. After all, quite a number of GOP national security officials have raised the #NeverTrump banner. One group openly doubts that Trump has the “qualifications, the judgment or the temperament” to be president. Others bristle at his claim that U.S. allies are feckless free-riders, and reject his call for making them pay for the protection that they receive under the U.S. security umbrella. Hillary Clinton is even more adamant that NATO and other Cold War-era alliances should never be subject to scrutiny. As she said several years ago, American global leadership, which manifests in such alliances, “is in our DNA.

In short, if Trump loses next week, it isn’t obvious that his ideas will outlast his unconventional campaign. The reigning transatlantic consensus, in which U.S. taxpayers subsidize European countries that might otherwise be inclined to spend slightly more on defense, will survive, and foreign policy elites in both the Democratic and Republican parties will work hard to ensure that no one like Trump ever emerges to challenge them.

However, as I note in the piece:

The mere possibility that America’s obligation within the NATO alliance might be open to interpretation should…serve as a powerful incentive for European countries to hedge their bets and get serious about developing a credible defense capability, one that is capable of acting without the United States in the lead.

Regardless of whether that occurs, it was probably unwise for Europeans to have relied so much on a U.S. political system over which they had no control. The fact that a bipartisan consensus among U.S. foreign policy elites sustained the transatlantic bargain for decades didn’t mean that that consensus was permanent.

Defenders of the status quo, whether they call it primacy, deep engagement, or simply American leadership, often claim that our dominant position in the world order will never be challenged because the United States is uniquely poised to reassure others. Core classical liberal principles, such as the importance of the rule of law, and governmental transparency and accountability, make Uncle Sam a reliable partner. Checks and balances between competing branches supposedly ensure that no one individual is able to effect dramatic changes in the way that the United States conducts its affairs.

Over the course of his campaign, Trump has torched these ideas. But he didn’t start the fire. Trump and Clinton both score poorly on transparency. Meanwhile, the Clintons’ critics often complain that they operate under a different set of rules than everyone else. Others worry that the president’s powers are not constrained by Congress or the courts. 

America’s so-called liberal hegemony lacks another key feature of liberalism: self-determination.

individuals must have a say in who governs them and how they are governed. Yet, the vast majority of people that fall under the dominion of the United States will have no role in selecting its next leader. They can’t vote in U.S. elections. And the awkward realization that they have entrusted their security to another government, liberal though it may be, over which they have no control, is crashing down hard on them.

I conclude:

Trump’s emergence provides an opportunity for leaders of all countries to reconsider how much trust they wish to place in the United States as the guarantor of global security. Some, perhaps many, will…revisit their decision to sub-contract governance to an unpredictable partner…

Without intending to do so, or even realizing that he’d done so, the mercurial Mr. Trump may be responsible for a welcome change in global security policy. If the reaction against him creates a more resilient international order, one that is less dependent on the military power of a single country, that would be a silver lining to Trump’s otherwise dark cloud.

You can read the whole essay here.

Last night Jimmy Wales, the founder of Wikipedia, delivered the first Joseph K. McLaughlin Lecture at the Cato Institute. He talked about the vision, history, organization, and impact of Wikipedia, and the influence of F. A. Hayek and his essay “The Use of Knowledge in Society” on his own initial conception of a crowdsourced encyclopedia. He also discussed Wikipedia’s occasional influence on public policy decisions, such as the defeat of the Stop Online Piracy Act (SOPA) in 2012. But I was particularly struck by this line (about 43:00 in the video):

Far too often lawmakers propose laws, and it’s fairly clear that they do not even have the most rudimentary understanding of how the internet works.

It reminded me of something Bill Clinton said at the Clinton Global Initiative in 2010:

Do you know how many political and economic decisions are made in this world by people who don’t know what in the living daylights they are talking about?

That’s a lesson policymakers ought to keep in mind whenever they contemplate legislating about health care, marriage, minimum wage laws, net neutrality, banking regulations, overtime pay, or anything else. Do they really understand how the particular market or industry works? Do they really understand how the impact of a new law or regulation will ripple through affected industries? In most cases they don’t, as Aaron Powell wrote about the lessons of SOPA:

SOPA was not the exception to the rule. Instead, it was just how things are done in Washington.

An Oct. 29, 2016, article by Danny Hakim in The New York Times gives a decidedly skewed view of the benefits of agricultural biotechnology. It is based on the author’s presumption that genetically modified organisms (GMOs) were supposed to accomplish two things: (1) increase crop yields; and (2) reduce the use of chemical pesticides.  In essence, Hakim sets up two straw men and proceeds to knock them down using questionable analysis.

Hakim compares crop yields in the United States and Canada, where GMO use is widespread, with yields in Western Europe, where GMOs generally are not allowed.  He finds that North America had “gained no discernible advantage in yields” relative to France, Germany, and other European countries.

As an example, Hakim includes a chart showing that the trend lines for yield increases of rapeseed in Europe and canola (a variety of rapeseed) in Canada are parallel, so that both added a similar amount of output per hectare. What he neglects to explain (or perhaps doesn’t appreciate) is that the percentage increase in European yield from 1995 to 2014 was modest. It rose from about 3.1 metric tons (MT) to 3.7 MT per hectare, an increase of approximately 19 percent.[i] Canada’s yield grew about the same quantity per hectare – from 1.4 MT to 1.9 MT – but the percentage increase was much greater at roughly 36 percent.  This is because Europe grows winter rapeseed under conditions that allow for high yields, while Canada grows spring canola under conditions in which lower yields are the norm.

So which farmers experienced a greater increase in profitability? Compared to their 1995 earnings (assuming constant prices), European famers managed to increase their per-hectare revenues by 19 percent. Canadian famers, on the other hand, achieved an increase that was almost twice as high – 36 percent. Casual observation would suggest that Canadian canola growers have become relatively more profitable over time than their European colleagues. One factor that appears to have increased Canadian profitability is the planting of GMO seeds.

As for Hakim’s argument that herbicide use has increased in the United States, especially on soybeans, it’s not clear that this is a bad thing. Most of the soybeans grown in North America have been genetically modified so that they are not harmed by Roundup (glyphosate), an herbicide that kills many weeds. Roundup has the advantage of being less toxic than some other herbicides, and it breaks down quickly in the soil. Anyone who has spent hours in a soybean field on a hot, humid day pulling weeds by hand is not likely to object to the idea that the same result could be achieved by modern biotechnology.

The bottom line is that Hakim largely ignores the reality that many thousands of individual farmers make decisions each year regarding which seeds to plant and which agronomic practices to use. The fact that so many farmers choose to plant costly GMO seeds (in countries where they are allowed) indicates that the added expense produces a real benefit. It seems improbable that those agricultural entrepreneurs all are making poor decisions about what is best for their businesses. The safe assumption is that users of biotechnology expect it will lead to a marginal increase in revenue that is greater than the marginal increase in cost. In the real world, it looks like the use of GMOs is being driven by favorable economics.

Mr. Hakim may think that GMO “technology has fallen short of the promise.”  The marketplace, however, understands things very differently.

[i]The data included in the article are in graphic form, so it’s not possible to determine precise numbers.

Scope of practice (SOP) restrictions in health care professions are often portrayed as a necessary intervention to protect consumer health and safety. Given how common this argument is, there have been surprisingly few studies trying to determine whether SOP restrictions actually have any impact on such outcomes. A new working paper seeks to fill this gap in the literature by determining whether SOP laws for certified nurse midwives (CNMs) affect health outcomes. On average, it turns out that the restrictions do not have a significant impact on maternal behaviors or infant health outcomes. Instead, they “primarily serve as barriers to practice and removing these restrictions has the potential to improve the efficiency of the health care system for delivery and infant care.”

SOP laws are determined at the state level, and regulate which activities and tasks certain professions can perform within the state. Physicians are generally unaffected, but other health practitioners are—in this case, CNMs specifically. Their level of restriction ranges from states with “no barriers,” where CNMs do not have oversight requirements, to states with “high barriers,” where they have to be under the direct supervision of a physician and may not write prescriptions. In heartening news, more states seem to be recognizing the wasteful nature of these laws. The recent trend for this specific case has been a move towards a more relaxed scope of practice environment.  

Scope of Practice for Certified Nurse Midwives by State, 1994 vs. 2013

Source: Markowitz et al.

There are also differences between the states as to when such SOP laws were implemented. It is these sources of variation that allow the authors to analyze the impact of CNM SOP regulations on health outcomes. They find that states that allow for full practice by CNMs have “on average, little or no differences in maternal health behaviors or infant health outcomes as compared to states with more restrictive SOP.”

While the authors fail to find significant effects there, they do find differences when it comes to rates of labor inductions, elective labor inductions, C-Sections, and elective C-Sections. This has implications for overall costs in the health care system: births at delivery cost almost $40 billion in 2013, with almost half of that amount funded through the Medicaid program. As the authors note, one of the biggest drivers of the costs associated with a delivery is whether it is via C-section. Women giving birth in “high barrier” states that are more restrictive of CNMs are 3 percentage points more likely to be induced and 1.6 percentage points more likely to have a C-section compared to those in “no barrier” states. There are small but statistically insignificant effects when comparing states that aren’t on the two extremes of SOP restrictions. The implied savings of moving states from “high barrier” to “no barrier” exceed $85 million a year due to the resulting reduction in C-section births.

These new findings are particularly illuminating when considered with another recent study that tried to determine whether occupational licensing in health care professions improved patient safety. Those authors looked at the licensing of midwives in the beginning of the 20th century and presented some evidence that the introduction of those laws reduced maternal mortality. It’s worth noting that the surrounding context a century ago was markedly different when it came to asymmetric information and a host of other factors. The new working paper highlighted here examines a recent situation that is more directly relevant to the discussion today, and finds that these SOP laws do not now have a significant impact on the related health or behavioral outcomes, on average.

Both studies seek to address a gap in our understanding of health care regulation, by finding out whether restrictions on practice actually improve health outcomes or consumer safety. While licensure for midwives may have had a positive effect a century ago, many things have changed in the intervening decades. Today, SOP restrictions act mostly as barriers to care. Where that’s the case, removing them could improve efficiency without harming patient safety.

I have chosen a provocative title, but it is fully justified. Fed officials are flying on autopilot, but the controls don’t work anymore, or at least not reliably. Fed watchers are largely clueless. The investment community and the economy may be collateral damage.

Let me begin by briefly reviewing the recent past. All through last year, Fed officials were signaling they would begin a program of rate increases. At first, there were going to be 8 increases of one quarter point. As the year progressed, the first increase faded into the future. Finally, in December 2015, the Fed finally hiked its new interest-rate targets by 25 basis points. In my opinion, the FOMC did so largely to keep its credibility.

At the time, I wrote that “the chief effect of Wednesday’s action and accompanying statement is to once again increase uncertainty in financial markets” (O’Driscoll 2015). I became convinced that, promises to the contrary notwithstanding, the Fed would not raise interest rates again before December 2016. Instead, policymakers would dither all year. I forecast the earliest rate hike would be in December 2016. Note, I did not predict the Fed would actually raise rates this December, just that they would not do so before. I think I have been vindicated.

The reasons for Fed inaction were both political and economic, and they reinforced each other. On the political front, most Fed watchers naively ignored that 2016 was a presidential election year. Yes, the FOMC has raised interest rates in election years. But I felt there would be significant pressure on the Fed no to do so this year, and, frankly, I felt the Fed has become more politicized over the years. So, political considerations argued against a rate hike.

I turn next to economic considerations. The conventional economic case for increasing short-term interest rates was really quite weak. Neither of the two Fed mandates provided much justification. Early on, some Fed officials forecast that inflation would rise as the economy neared full employment. Yet the inflation rate remained stubbornly below the two-percent target. Next, Fed officials suggested the economy was approaching full employment. With or without inflation, they argued, the FOMC needed to start raising interest rates.

In fact, this recovery has been weak by any standard measure. It is a long recovery, but a weak one. That is true whether measured by output growth or employment growth. The contention that we are at or near full employment ignored labor-market realities. The Civilian Labor Force Participation Rate has been in free fall since the Great Recession. It has only recently shown an uptick. For men, the picture is even more dismal. The participation rate is near an all-time low. It is a largely unnoticed crisis, which Nicholas Eberstadt (2016) has chronicled. He estimated that there is a “male jobs deficit” of 10 million. The economy is not close to any reasonable concept of full employment. There was no employment justification for raising short-term interest rates.

I have come to the conclusion that the stated reasons for raising rates are not the real reasons, at least not for some Fed officials. They are constrained to Humphrey-Hawkins criteria (inflation and unemployment). But there is concern over the compressed margins for banks and money-market mutual funds. For banks, spread between borrowing rates and lending rates are compressed. For money funds, yields on eligible assets are very low, and it is difficult to pay a positive return to shareholders after costs. Can you imagine Fed Chair Yellen or New York Fed President Dudley saying we need to raise interest rates to make banking more profitable?

There are also international considerations. By now, central bankers are beginning to see the folly of negative interest rates. For the ECB to get rates just back to zero, however, U.S. rates must move higher into positive territory. Then, too, there are the EU’s banking problems. The cost of bailing out European banks could be monumental. A little extra economic growth from a weaker Euro would help finance that cost. Can you imagine if Fed officials said we need to raise U.S. interest rates to bail out the Germans?

There are good reasons for raising rates. They have been too low for too long. Capital is being misallocated all over the place as a result. Instead of stimulating investment in the real economy, low interest rates have fueled stock buybacks, M&A, and financial engineering more generally. The financial sector has grown at the expense of real economic activity. Asset bubbles have been created. The Fed’s policy of buying MBSs (Mortgage-Backed Securities) has channeled credit into housing. If these considerations were paramount, however, interest rates should have been higher long ago.

Some might say that, if I am correct, the Fed is in an impossible position. The economy is too weak to raise interest rates, but asset bubbles necessitate higher rates. That would be true if current monetary policy were stimulating the economy. In fact, unconventional monetary policy has been contractionary. To repeat, the policy of very low interest rates has been contractionary, not expansionary. The Fed has been dampening economic growth.

As already noted, very low interest rates have compressed bank’s margins. One can see that in the relatively flat yield curve. The point is even more compelling if one factors in risk. Lending is a risky activity. A bank must expect to earn a positive risk-adjusted rate of return. Right now it can earn 50bp risk free on reserves on deposit at the Fed. Moreover, those reserves satisfy the mandated liquidity balances of Dodd-Frank. Speaking of Dodd-Frank, its restrictions on risk-taking greatly reinforce the effects of the Fed’s low interest rate policies. Bank lending is discouraged. That has had an especially large impact on small business lending. Small businesses are the job creators for the economy. That is another reason why employment growth has been weak.

In a forthcoming article in the Cato Journal, Professor Charles Calomiris of Columbia University has estimated that the value to a bank of acquiring a deposit is approaching zero. Banks are actually turning deposits away. That implies, of course, that the value of all those branch networks is approaching zero since they are basically deposit-gathering facilities. The traditional business of banking may not be a viable model under current monetary and regulatory policies (Calomiris 2017).

To recapitulate, higher interest rates might actually spur economic growth. And they would be good antidotes to asset bubbles. What, then, are the prospects for higher interest rates?

The Federal Reserve’s unconventional monetary policy has robbed it of the ability to influence interest rates as it has in the past. That is particularly true for raising interest rates. The necessary linkages in monetary policy have been broken, or at least damaged.

In the past (pre-2008), how did the Fed raise interest rates? It sold short-term government bonds, Treasury bills. That pushed up the interest rates on bills and put upward pressure on other short-term interest rates. Banks, attracted to higher returns, competed for federal funds to lend. That drove up the fed funds rate, which restrained lending. The fed funds market was a choke point.

All these linkages are gone.

The Fed has zero Treasury bills on its balance sheet, so it has no short-term assets to sell. The policy of large-scale asset purchases — QE — flooded banks with reserves. Banks do not need to borrow reserves; reserves are abundant. The fed funds market, greatly shrunk in size, now mainly consists of transactions between GSEs — chiefly Federal Home Loan Banks — and a few banks, mainly foreign.

So what did the Fed do in December 2015 to increase short-term interest rates? It increased the interest rate paid on reserves from 25bp to 50bp. And it posted a rate of 25bp for reverse repos at the New York Fed. It then set a target range of 25-50bp for the fed funds rate. Did it hit that fed funds target? Yes, about in the middle of the range. Did it matter? Not really, for the reasons just mentioned. The availability and pricing of fed funds no longer constrains most banks. Moreover, the goal was surely to raise market interest rates. The opposite occurred. I quote former Cleveland Fed president, Jerry Jordan, on point: “Yields of market-determined interest rates subsequently fell and remain below the levels that prevailed before the increase in administered rates” (Jordan 2016: 26).

The Fed is now largely engaged in a futile exercise to control the economy. The massive expansion of its balance sheet flooded banks with excess reserves. “Operation Twist” robbed it of short-term assets. Its influence over short-term, market interest rates is attenuated at best. Nor does it seem able to control the money supply, though that is a more complicated issue (Jordan 2016).

What financial markets have feared the most, they in reality dare not hope for. The Federal Reserve is not positioned to raise short-term, market interest rates. If the feared spike in inflation materializes, the Fed is not positioned to contain it. Markets are expecting or fearing the Fed will do things it can no longer do. The misalignment between market expectations and Fed capabilities is very dangerous, and I fear it will not end well. Not an out-of-control central bank, but a not-in-control central bank is the problem markets must confront. The Fed is not positioned to control inflation when and if that becomes a problem.

___________

This post is a revised version of a speech given to the CFA Society of Nevada, October 27, 2016.

References

Calomiris, Charles W. (2017) “The Microeconomic Perils of Monetary Policy Experiments.” Forthcoming in Cato Journal, Winter 2017, based on remarks delivered at the Shadow Open Market Committee’s Fall 2016 Meeting.

Eberstadt, N. (2016) “The Idle Army: America’s Unworking Men.” Wall Street Journal (October 2).

Jordan, J. (2016)  “Rethinking the Monetary Transmission Mechanism.”  Paper prepared for the Cato Monetary Conference, November 2016.

O’Driscoll, Jr., G. P. (2015) “The Fed’s Uncertain Leap Forward.” Wall Street Journal (December 17): A19.

[Cross-posted from Alt-M.org]

In a recent editorial, the Dallas Morning News inveighed against expanding educational choice in Texas, arguing that legislative leaders should “focus on improving public schools” instead. What the DMN editorial board means, of course, is “spend more money,” as they make clear in the penultimate paragraph. Yet although the national average annual expenditure per pupil for district school students has, after adjusting for inflation, nearly tripled in the last forty years, student performance remains flat. Moreover, there is little evidence that merely increasing spending improves school performance or student outcomes. Nevertheless, the DMN has reservations about the possible effects of expanding educational choice:

One proposal would create education savings accounts. If a parent decides against public schools, the money that would have gone with the student to the local school district would instead go to the account, for parents to use on private school.

That could decimate public schools. What about the quality of education for the students left behind?

Good question. Fortunately, a few dozen high-quality studies provide an answer. Contrary to the DMN’s assertion that educational choice “would divert scarce taxpayer dollars from already struggling public schools and do nothing to help improve them,” the research overwhelmingly finds that expanding choice and competition has a positive effect on the performance of district schools. For example:

  • A 2016 study by Anna Egalite of North Carolina State University looked at the impact of the Louisiana Scholarship Program (LSP) on Louisiana public schools. Egalite found, “The competitive threat of the LSP ranges from negligible to modestly positive in the public schools exposed to the threat of competition, with effect sizes growing in magnitude as the competitive threat looms larger.”
  • A 2014 study by David Figlio and Cassandra Hart of Northwestern University examined the competitive effects of the Florida Tax Credit Scholarship Program on public schools. They learned that more access and variety of private schools increased the competitive pressure on public schools in the wake of the policy announcement. They state in their conclusion, “The fact that we observed generalized improvements in school performance in response to the competitive threats of school vouchers, even in a state with rapid population growth, suggests that voucher competition may have effects elsewhere.”
  • A 2011 peer-reviewed study by Jay Greene of the University of Arkansas and Marcus Winters of the University of Colorado – Colorado Springs looked at the impact of Florida’s McKay special education voucher program on Florida public schools. Greene and Winters found there was approximately a “12 percent reduction in the probability that a fourth- through sixth-grade student” was diagnosed with a learning disability in a public school with average levels of competition. They also found that “being in a public school surrounded by the average number of McKay-accepting private schools was related to an increase in academic proficiency of about 0.01 standard deviations in both math and reading. The positive but very mild competitive effect is consistent with what has been found in previous research evaluating more conventional school choice policies.”
  • A 2009 study by Jay Greene and Ryan Marsh of the University of Arkansas considered the systemic effects of expanding school choice in Milwaukee. Greene and Marsh found that public school students in Milwaukee fare better academically when they have more free private options through the voucher program. They concluded, “It appears that Milwaukee public schools are more attentive to the academic needs of students when those students have more opportunities to leave those schools. This finding is robust across several different specifications of the model.”

In other words, the research finds that when parents have multiple schooling options, schools have to work harder to attract them. Throwing money at district schools hasn’t worked, but expanding educational choice does not mean abandoning the schools where most children are enrolled. Rather, empowering families with expanded educational choice gives those schools a stronger incentive to meet the needs of students and their parents. That means using resources more wisely to provide families with a learning environment that they want to affirmatively choose, rather than one where they are merely assigned based on the home they could afford.

It is irresponsible for media outlets to make grand pronouncements about the supposed effects of a policy without first looking at what the research literature actually says. Far from “decimating” district schools, dozens of studies show that educational choice policies foster improved performance. The Dallas Morning News should correct the record.

Recently, the UK Daily Telegraph ran a remarkable Op-Ed written by William Hague, the just-retired Conservative politician and former UK Foreign Secretary. The title alone was startling: “Central bankers have collectively lost the plot. They must raise interest rates or face their doom.”

Now I confess that I may be a little biased. Lord Hague’s article made many of the same points that Martin Hutchinson and I set out in a paper we have just prepared for Cato’s 34th Annual Monetary Conference next month, “From excess stimulus to monetary mayhem,” and I am pretty confident that he hasn’t seen our draft. What encourages me most about his piece is that it hints that the monetary policy Zeitgeist is changing — the failures of recent central bank policies are becoming increasingly obvious to anyone without a completely closed mind. The Overton Window might at last be moving in our direction.

In this post, I would like to summarize his argument and offer a few observations of my own.

He begins with a nice little nugget about the Tories’ response to Gordon Brown’s pronouncement, on the heels of Labour’s landslide election victory in 1997, that the Bank of England would be operationally independent from governments:

We [the Tory “survivors”] decided not to demur. In private, we had considered doing the same thing ourselves. The idea that central banks should be free of political pressures and the electoral cycle as they set interest rates had become a prevailing one across the world — with good reason after the many wild swings in inflation and interest rates over previous decades.

Many of us had guessed what the Tory leadership were thinking on this issue, but it is good to see it confirmed.

Hague goes on to suggest that central bankers have badly misused the powers that were granted them and are “now in deep trouble,” continuing to pursue “emergency policies” that are “becoming steadily more unpopular and counter-productive.” Unless a course correction comes soon, central bankers “will find their independence increasingly under attack.” He continues:

In 2008 the central banks reacted to a massive crisis they had completely failed to foresee by cutting rates to record lows and embarking on “quantitative easing” … The trouble is that eight years later they are, to varying degrees, still doing it. Like doctors keeping their patients on a drip many years after an operation, they are losing credibility and producing very dangerous side effects.

To use the Marxist phrase — although one that Hague did not use — central banks are facing a legitimation crisis.

He then outlines no less than 10 serious drawbacks from ZIRP and QE that could be “politically explosive or economically unwise if continued indefinitely:”

#1: Savers find they can’t earn a worthwhile return and are driven into riskier assets whose prices rise further.

#2: Asset holders get much richer, but others are left out, seriously exacerbating social and political divides and fueling the anger behind populist campaigns.

#3: Pension funds have poor returns and therefore suffer huge deficits, causing businesses to have to put more money into them rather than finance expansion.

#4: Banks find it harder to run a viable business, a problem very evident now in Germany and Italy.

#5: Those who are able to save more do so, because they need a bigger pot of savings to get an equivalent return, i.e., low interest rates cause those people to spend less, not more.

#6: Companies have an incentive to use borrowed money to buy back shares rather than spend the money on new productive investments.

#7: Central banks are buying up corporate bonds, not just government bonds, so they are acquiring risky assets themselves and giving preference to some companies over others.

#8: Zombie companies are allowed to stay in business only because they can borrow so cheaply, which drags down productivity.

#9: Pumping up the prices of stock markets and houses without an underlying improvement in economic performance becomes ever more difficult to unwind and ultimately threatens an almighty crash when it does come to an end.

#10: When people see emergency measures going on for nearly a decade it undermines their confidence in central bankers, whom they think have lost the plot.

He then concludes:

I am not an economist but I have come to the conclusion that central banks collectively have now indeed lost the plot. The whole point of their independence was that they could be brave enough to make people confront reality. Yet in reality they are blowing up a bubble of make-believe money to avoid immediate pain, except for penalising the poor and the prudent.

Non-economist or not, Hague’s economic analysis is far better than anything I have seen published by any central bank outside of the Bundesbank.

He goes on to explain that he had put this view to senior staff at a major Far Eastern central bank — my guess the Bank of Japan. He was then astonished to discover that they agreed with him, but explained that no single central bank could reverse these policies without causing a recession for their own country, unless there was a coordinated move by all central bankers to gradually raise interest rates. He wryly notes:

The policies of any one central bank may well be perfectly rational …. But so is a decision by any one sheep to run with the flock when in danger. The trouble is that the whole flock might be heading for a cliff.

He then sticks the boot into the argument that interest rates are low merely because of a global savings glut: even if interest rates have to be low because of such a glut, that does not mean those rates have to be driven to zero or subzero levels as they are now. Then he turns on those central bankers who argue that their mandate is only to keep inflation low and that they are merely doing a technical job:

I have bad news for them. The accumulating effects of loose monetary policy globally are intensely political. When pension funds renege on promises, or inequality widens further, or savers become desperate, huge public and political anger is gong to burst over the heads of the world’s central banks.

Finally, he calls on the biggest sheep of them all to wake up to the problem and lead the others away from the cliff before they all go over:

The only way out is for the US Fed to summon the courage to lead the way to higher interest rates, and others to follow slowly but surely. If they fail to do so, the era of their much vaunted independence will come, possibly quite dramatically, to its end.

Lord Hague’s elegant analysis provides a much needed wake-up call. Unfortunately, if there is any indication that the Fed is listening, I must have missed it.

Where I would slightly beg to differ from Hague is where he claims that no central bank other than the Fed can do anything useful because the collective herd mentality is too strong to resist.

Not so. Such pressure is difficult to resist, but it is not impossible and I wouldn’t listen too much to Japanese central bankers whose funny money fantasies are increasingly delusional. Fortunately, the Bank of Canada has shown that less extreme policies can be implemented even in this international environment, and the Bundesbank has been valiantly resisting the pressures for ever looser monetary policies in Europe.

Going back to the UK, there is no sign that the Bank of England is open to his message either: on the contrary the Bank’s Monetary Policy Committee recently voted to reduce Base Rate to its lowest level ever and expand QE, and the senior management of the Bank are too invested in the inherently self-contradictory “we have fixed the banking system but still need more QE” narrative to ever acknowledge that they have had it all wrong all along.

Were I advising the Chancellor, I would be advising him to install a new senior management team in the Bank, reassert control over the Bank with a view to raising interest rates, and prepare contingency plans against the renewed financial crisis that some of us have been warning about for years. But I have zero confidence in the capabilities of either the Bank of England or the UK Treasury to do the analysis properly. That the Bank of England is discredited is amply proven by their demonstrably flawed — because empirically falsified — narrative about UK banks’ capital positions and the Bank’s laughable stress tests. That the Treasury lost its credibility is clear from its alarmist Brexit reports, which were as credible as Tony Blair’s infamous dodgy dossiers.

But just because the authorities are not listening to Lord Hague’s warnings does not invalidate them.[1] That is the problem in a nutshell.

___________________

[1] Hague does not, however, address some of the very serious issues with how central banks should change policy, but I hope to address these in one or more future posts.

[Cross-posted from Alt-M.org]

In 2014, the Cato Institute published a policy analysis that argued that Congress should create a work visa program that would allow states to select some of the foreign workers that the federal government admits—a model that Canada has used for decades. All states would benefit from this approach, since they all suffer from low federal caps on visas and since they all have the best knowledge of their local labor markets. But the states that will likely benefit the most are states with lower populations.

While there is no breakdown of the total number of visas issued to workers by state, we know how many workers enter each state annually. This number is imprecise as a measure of the total number of workers in each state because it double counts people who leave the state and reenter during the year, but it is still useful for looking at the distribution of workers among the states.

Figure 1 presents the distribution of all worker entries and entries of workers under the most important category—the H temporary work visa—by grouping the states into three roughly equal baskets: 1) the very large states with more than five percent of the entire U.S. population, 2) those in the middle with 2 to 5 percent, and 3) those with 2 percent or less. As is readily apparent, the big four states—California, Texas, New York, and Florida—receive a much higher percentage of workers than their share of the population. The middle group of states receives only slightly less than their population share for all worker entries.

On the other hand, the smallest states receive not only fewer workers in absolute terms, but disproportionately less than other states compared to their population. These 34 states are home to 30 percent of all Americans, but receive just 18.6 percent of guest workers overall—an 11.4 percent difference between population share and worker share.  A similarly large gap can be seen for worker entries under the important H category.

Figure 1: Difference Between Share of All Worker Entries and Share of U.S. Population; Share of H Worker Entries and Share of the Population in 2014

 

 

Sources: DHS (Entries); Census (Population)

Figure 2 breaks the third group in Figure 1—the smallest states—into two groups: 1) the 14 states with a population between 1 and 2 percent and 2) the 20 states with a population less than 1 percent of the overall U.S. population. As can be seen, it’s the higher-end of this less populous group that has the greatest disparity between population share and worker share. These are states like Iowa, Minnesota, Colorado, and Wisconsin. The 20 smallest states with populations between 0 percent and 1 of the population also had a difference, but it was a much smaller disparity—2.4 percent compared to 9 percent for states in the 1 to 2 percent range.

Figure 2: Difference Between State Share of U.S. Population and State Share of All Worker Entries and State Share of H Worker Entries in 2014

Source: See Figure 1

We know that this unequal distribution is not explained by differences in state labor market demand since domestic workers are actually moving from larger states to the smaller states. The real explanation is that larger and smaller states have different types of labor demand.

Figure 3 groups the states in the same manner as Figure 1, but breaks the H category into its two skill classes: H-1B high skilled workers and H-2s for lower skilled agricultural and non-agricultural workers (H-2As and H-2Bs). As you can see, the less populated states received a share of lesser skilled H-2 workers in proportion to their populations, but received half as many H-1Bs as their share of the population. For the more populous states, the reverse was true—they received much more high than low skilled workers.

Figure 3: Difference Between Share of H-1B Higher-Skilled Worker Positions and State Share of U.S. Population; Share of H-2 (H-2A and H-2B) Lesser-Skilled Worker Positions and State Share of U.S. Population in 2014

 

Sources: See Figure 1

We can take this conclusion further. Because H-1B and H-2B visas—the nonagricultural part of the H-2 class—are capped, employers actually plan to receive far more workers than they actually get under the federal system. The first step toward applying for an H visa is submitting a certification of need for a certain position to be filled by a foreign worker. The Department of Labor either approves or denies this certification. If it’s approved, then the employer submits a request to the Department of Homeland Security for visas for the workers that it wants, if there are any visas remaining.

 

This certification process allows a very rough comparison between the demand for workers and the number of workers that each state receives. It is very imprecise because other factors limit the number of certifications that can be filed. All H-1B certifications, for example, have to be filed on the same week six months before the workers arrive in the United States, so it underestimates total market demand. Yet as Figure 1 shows, smaller states received an appreciably lower share of worker entries than their shares of certifications of need. In other words, smaller states need a lot more lesser-skilled workers that they are unable to get under the current system.

Figure 4: Difference Between the Share of Certifications for H Worker Positions and Share of H Worker Entries in 2014 By State Share of U.S. Population

Source: DOL (Certifications), DHS

Overall the statistics on certifications for H workers reveal that every state is losing under the current system. On average, the Department of Labor approved an average of 22,438 certifications of need for H worker employment in 2014, but each state received only about 6,000 H workers in 2014. It is important to remember that this discrepancy is with all the current myriad federal rules and with no year-round visa available for lesser-skilled workers at all. The demand for workers surely far exceeds this amount.

The United States has a demand for foreign workers. While reforms to the federal system would be an improvement, Congress has consistently shown that it is incapable of estimating the needs of the states.

The table below presents all of this information for all 50 states and D.C.

Table: State Share of U.S. Population, Share of All Worker Entries, Share of H Temporary Worker Positions, Share of H-1Bs, Share of H-2s, Number of H-2s, and Number of H-1Bs in 2014

  % of US Pop % of All Entries % of Hs % of H2s % of H1Bs H2 Certs. H1B Certs % H2 Certs % H1B Certs California

12.1%

14.8%

15.8%

4.7%

19.8%

8,752

172,318

4.1%

18.5%

Texas

8.1%

10.1%

8.9%

11.1%

8.9%

17,028

84,612

8.1%

9.1%

New York

6.3%

14.0%

7.3%

2.8%

14.0%

7,898

76,105

3.7%

8.2%

Florida

6.1%

6.7%

4.3%

7.9%

5.2%

20,056

29,447

9.5%

3.2%

Illinois

4.2%

3.7%

4.6%

0.8%

4.9%

2,333

50,551

1.1%

5.4%

Pennsylvania

4.1%

2.4%

4.6%

1.7%

2.6%

4,397

48,550

2.1%

5.2%

Ohio

3.7%

2.0%

2.7%

1.1%

2.0%

2,715

28,518

1.3%

3.1%

Michigan

3.2%

6.2%

2.4%

1.2%

2.8%

3,492

24,469

1.7%

2.6%

Georgia

3.1%

2.4%

4.1%

4.5%

2.5%

12,166

34,841

5.8%

3.7%

North Carolina

3.1%

2.1%

3.8%

6.2%

1.9%

18,380

24,539

8.7%

2.6%

New Jersey

2.8%

3.9%

6.0%

1.1%

6.6%

3,252

65,505

1.5%

7.0%

Virginia

2.6%

1.8%

2.8%

2.6%

2.4%

6,631

25,742

3.1%

2.8%

Washington

2.2%

2.9%

3.7%

3.5%

4.4%

10,026

32,010

4.7%

3.4%

Massachusetts

2.1%

3.7%

3.2%

1.0%

4.2%

3,230

33,473

1.5%

3.6%

Indiana

2.1%

0.9%

0.9%

0.4%

0.8%

1,980

8,028

0.9%

0.9%

Arizona

2.1%

2.8%

1.7%

17.0%

1.2%

5,710

13,382

2.7%

1.4%

Tennessee

2.1%

1.0%

1.4%

1.6%

0.7%

3,426

12,868

1.6%

1.4%

Missouri

1.9%

0.7%

1.2%

1.0%

0.7%

2,992

10,938

1.4%

1.2%

Maryland

1.9%

1.4%

1.7%

1.5%

1.5%

4,436

14,788

2.1%

1.6%

Wisconsin

1.8%

0.8%

1.2%

0.2%

0.8%

842

13,088

0.4%

1.4%

Minnesota

1.7%

1.0%

1.8%

0.6%

1.3%

1,867

18,190

0.9%

1.9%

Colorado

1.6%

1.3%

1.5%

2.2%

1.0%

5,225

12,263

2.5%

1.3%

Alabama

1.5%

0.7%

0.4%

1.1%

0.3%

1,911

2,586

0.9%

0.3%

South Carolina

1.5%

0.8%

0.9%

1.5%

0.4%

4,885

4,977

2.3%

0.5%

Louisiana

1.5%

0.9%

1.3%

5.3%

0.3%

12,106

2,450

5.7%

0.3%

Kentucky

1.4%

0.8%

1.2%

3.2%

0.4%

7,438

6,166

3.5%

0.7%

Oregon

1.2%

0.7%

0.8%

0.6%

0.7%

1,639

7,505

0.8%

0.8%

Oklahoma

1.2%

0.4%

0.4%

0.8%

0.3%

1,890

3,154

0.9%

0.3%

Connecticut

1.2%

1.1%

1.8%

0.2%

1.6%

857

20,183

0.4%

2.2%

Iowa

1.0%

0.4%

0.6%

0.5%

0.4%

1,267

5,885

0.6%

0.6%

Mississippi

1.0%

0.4%

0.5%

2.1%

0.1%

5,131

868

2.4%

0.1%

Arkansas

0.9%

0.5%

0.9%

2.0%

0.4%

5,203

5,355

2.5%

0.6%

Kansas

0.9%

0.4%

0.5%

0.6%

0.4%

1,789

4,122

0.8%

0.4%

Utah

0.9%

0.4%

0.5%

0.7%

0.4%

1,644

3,905

0.8%

0.4%

Nevada

0.9%

0.6%

0.4%

0.9%

0.2%

2,138

2,305

1.0%

0.2%

New Mexico

0.7%

0.2%

0.1%

0.1%

0.1%

241

1,328

0.1%

0.1%

West Virginia

0.6%

0.1%

0.1%

0.0%

0.1%

136

574

0.1%

0.1%

Nebraska

0.6%

0.7%

0.4%

0.3%

0.7%

1,604

3,032

0.8%

0.3%

Idaho

0.5%

0.2%

0.4%

1.0%

0.1%

4,013

1,007

1.9%

0.1%

Hawaii

0.4%

0.4%

0.1%

0.0%

0.2%

39

615

0.0%

0.1%

Maine

0.4%

0.6%

0.3%

1.4%

0.1%

1,767

1,583

0.8%

0.2%

New Hampshire

0.4%

0.3%

0.3%

0.6%

0.3%

497

2,879

0.2%

0.3%

Rhode Island

0.3%

0.2%

0.5%

0.1%

0.3%

293

4,896

0.1%

0.5%

Montana

0.3%

0.2%

0.1%

0.3%

0.0%

1,130

248

0.5%

0.0%

Delaware

0.3%

0.3%

0.5%

0.2%

0.3%

359

5,139

0.2%

0.6%

South Dakota

0.3%

0.1%

0.2%

0.4%

0.1%

1,675

479

0.8%

0.1%

Alaska

0.2%

0.2%

0.1%

0.2%

0.1%

1,194

301

0.6%

0.0%

North Dakota

0.2%

0.3%

0.2%

0.5%

0.1%

1,844

748

0.9%

0.1%

Vermont

0.2%

0.3%

0.1%

0.4%

0.1%

1,074

637

0.5%

0.1%

D.C.

0.2%

1.0%

0.5%

0.0%

0.9%

1

5,848

0.0%

0.6%

Wyoming

0.2%

0.1%

0.1%

0.3%

0.0%

594

135

0.3%

0.0%

Averages          

4,141

18,297

   

Sources: See Figure 1, Figure 4

Last July, Dallas police used a robot to kill the man who fatally shot five Dallas-area police officers. Shortly after the shooting I noted that new technologies, such as robots, should prompt lawmakers to find ways to make the face-to-face interactions citizens have with officers safer and less frequent. A recent Amazon patent reveals how new technologies can play a role in improving traffic stops, one of the most common citizen-police encounters.

Amazon Technologies, Inc. recently secured a patent for small shoulder-mounted police drones. The patent abstract explains that, “The techniques and systems can include routines to provide enhanced support for police during routine traffic stops.”

Drones like the one detailed in the Amazon patent could help improve traffic stops. Drones would allow police to examine a pulled-over vehicle before approaching in person. This increased situational awareness would help police officers, providing them with valuable information about how many people are in the car and whether the driver or any passengers have their hands in sight. As drone technology improves it’s likely that police will be able to use similar drones to issue commands. 

If appropriate accountability policies are enacted, these small drones could serve as useful tools in police misconduct investigations. Drone footage of the Philando Castile and Samuel DuBose shootings, for example, would have been helpful to investigators.

But despite the potential for these small drones being useful in misconduct investigations and helping police during traffic stops, citizens may be concerned about the impact such drones could have on their civil liberties. Having a small drone buzzing around your car during a traffic stop may be unnerving, but unless the drone is outfitted with sophisticated surveillance tools it’s unlikely that it will prompt a robust Constitutional challenge.

If these small Amazon drones are equipped with traditional cameras and don’t enter a car during a traffic stop, then they will only be capturing images of material in “plain view.” Nonetheless, citizens should be wary of small police drones being outfitted with surveillance technology that could raise constitutional issues, such as thermal scanners.

New technologies such as drones and body cameras will undoubtedly play an increasingly prominent role in law enforcement. Small drones like the one described in Amazon’s patent could help make routine traffic stops safer for officers and citizens. However, as the ongoing debates about body cameras have demonstrated, these new technologies can only serve as tools for worthwhile criminal justice reform if they’re governed by good policies. It’s not hard to see how small drones could help police and citizens during traffic stops. But as police drones become more common we shouldn’t forget that they can serve as platforms for a host of technologies that threaten civil liberties.

It’s been a little over a year since Bernie Sanders assured America that the public was “sick of hearing” about Hillary Clinton’s “damn e-mails,” and to put it mildly, the claim has not aged well. Even before Friday’s announcement that the FBI had uncovered an additional cache of e-mails from Clinton’s personal assistant Huma Abedin—and the inevitable media feeding frenzy that followed—Clinton’s use of a private e-mail server during her tenure as Secretary of State had remained a central campaign issue. If anything, the controversy had metastasized: The FBI’s investigation into Clinton’s server, culminating in a recommendation that no criminal charges be brought, was received by many as evidence of a corrupt cover-up even more disturbing the underlying offense, a clear-cut case of a Beltway elite getting a pass for conduct that would have seen a normal schlub clapped in irons. It’s this, probably more than any other alleged misdeeds, that has made “lock her up!” a popular refrain at Donald Trump’s rowdy rallies.

As a frequent critic of the FBI’s routine demands for broadened surveillance powers, it’s heartening to see people recognizing that the Bureau is not somehow immune to improper political influence. Moreover, given the Obama DOJ’s unprecedented use of the Espionage Act to prosecute whistleblowers (rather than spies)—his administration has pursued more cases under that law than all his predecessors’ combined—it’s hard not to feel a twinge of schadenfreude when the public concludes that Clinton’s “extreme carelessness” with classified information (as FBI director James Comey characterized it) must surely be criminal too. But in large part because I’m uneasy about normalizing this aggressive approach to the Espionage Act, I think it’s necessary to explain why this widespread perception is wrong, and Comey’s conclusion that “no reasonable prosecutor” would have pursued charges against Clinton on the available facts was pretty clearly right. While it’s impossible to know what other damaging revelations the newly discovered tranche of e-mails may contain, it seems unlikely they will materially alter that basic legal conclusion.

One thing to get out of the way up front: None of what follows is ultimately a defense of Clinton. Beyond the poor judgement implied by her sloppy approach to classified information, the effect (and probable intent) of Clinton’s use of a private server was to hamper government transparency by giving her improper de facto control over correspondence that should be subject to Freedom of Information Act requests—which is to my mind perhaps the most troubling aspect of her conduct. (Much as this former reporter might wish otherwise, circumventing FOIA is not a criminal offense.) This post is narrowly concerned with whether she ought to have been prosecuted under the Espionage Act, specifically 18 USC §793(f). which reads:

(f) Whoever, being entrusted with or having lawful possession or control of any document, writing, code book, signal book, sketch, photograph, photographic negative, blueprint, plan, map, model, instrument, appliance, note, or information, relating to the national defense, (1) through gross negligence permits the same to be removed from its proper place of custody or delivered to anyone in violation of his trust, or to be lost, stolen, abstracted, or destroyed, or (2) having knowledge that the same has been illegally removed from its proper place of custody or delivered to anyone in violation of its trust, or lost, or stolen, abstracted, or destroyed, and fails to make prompt report of such loss, theft, abstraction, or destruction to his superior officer—Shall be fined under this title or imprisoned not more than ten years, or both.

I’m focusing on this clause because it’s the one the FBI itself seems to have considered the only plausible basis for bringing charges. There are other statutes that apply to mishandling of classified information by military personnel or other Defense Department employees, which obviously wouldn’t apply to Clinton. And there are other clauses of the Espionage Act that criminalize “willfully” providing classified information to uncleared persons, or obtaining the same “with intent or reason to believe that the information is to be used to the injury of the United States.” Nobody has seriously suggested that Clinton was deliberately trying to make classified information available to the adversaries of the United States. But what about 793(f), which requires not “willfulness” but only “gross negligence”?

There are three main obstacles to prosecution under that clause in this case: One factual, one statutory, and one constitutional. I’ll take them in turn.

The factual problem becomes clear when you ask what specific actions a prosecutor would charge—actions that could be laid at Clinton’s feet, as opposed to those of her staff. Setting up a private e-mail server, or using it to carry on professional correspondence? That surely wouldn’t fly: Having a private e-mail account doesn’t in itself constitute “mishandling” of classified information, and plenty of officials with high security clearances have done and continue to do the same. That it was Clinton’s own server, as opposed to a Gmail account, makes no difference. Indeed, while it’s certainly relevant to one’s judgment of whether Clinton was seeking to keep her e-mails beyond the reach of FOIA requests, the private server is largely a red herring when it comes to the charge of mishandling classified information. What’s of potential legal relevance is that the server was not an appropriate “place of custody” for whatever classified material ended up there: A personal server is, in that respect, no different from a commercial e-mail account or even an official but unclassified State.gov account. Moreover, being in receipt of classified information on an insecure channel would not, in itself, satisfy the elements of an offense under the statute: A classified document sent to Clinton’s account might constitute a violation by the sender, but it would by definition have to be “removed” and “delivered” by the sender—not the recipient. So the factual obstacle here is that, once we recognize that “setting up a server” is a red herring, it’s not at all clear whether the FBI found instances of Clinton, as opposed to her staff, removing or “permitting the removal” of documents stored in classified systems. (The second prong might conceivably apply if it was sufficiently clear to Clinton that classified material had been illegally removed, though we’d then be left wondering what “superior officer” the Secretary of State is obligated to notify. The president?)

The reference above to the removal of documents is deliberate, and that brings us to the statutory obstacle to prosecution. While several sections of the Espionage Act that apply a “willfulness” standard cover any inappropriate communication of classified information, §793(f) really appears to be about the exfiltration of files or documents, not the mere discussion of classified matters. As Prof. Steve Vladeck, an expert on the Espionage Act, has observed, discussing classified subjects on an insecure channel might be grounds for a reprimand or revocation of clearance, but it has not traditionally been viewed as a criminal offense—and the language and structure of the statute, as well as the history of prosecution, bear out that interpretation.

Start with the language: Discussing classified facts known to you is not, on any ordinary use of English, an instance of “removing” information from a “place of custody,” except perhaps on the strained reading that makes the cleared individual’s brain such a “place.” For confirmation, we can look to differences in the language found in other parts of the statute, as under traditional rules of legal interpretation, differences in otherwise parallel phrases should be construed as corresponding to a difference in legal effect. Whereas §793(f) applies to a person “being entrusted with or having lawful possession or control of” classified documents, it omits the language “access to” found in §793(e). Similarly, §783(f) covers allowing the materials to be “removed… delivered… lost, stolen, abstracted, or destroyed” but not “communicated” as in (e). The reading that most naturally makes sense of those differences is that (f) is about permitting the wholesale abstraction of files & documents, not merely communicating information someone may have gleaned from having access to them (or from access to classified briefings).

Structural considerations support the same conclusion. Consider: If §793(f) covers the same conduct as the rest of the statute, with the same penalties, with the only difference being the standard of “gross negligence” as opposed to “willfullness,” the structure of the whole becomes rather puzzling. Why have a separate clause to prohibit the same conduct under two standards, rather than having a unified prohibition on doing the same thing either “negligently” or “willfully”? Why, for that matter, would it make sense to penalize communication of classified information to uncleared individuals only with a strong intent requirement if communication of the same information even to cleared persons on an insecure channel would be subject to exactly the same penalties under a lower standard of negligence? Such a reading makes the statute into a patchwork quilt riddled with redundancies rather than a coherent whole.

Finally, the history of prosecutions under the law tends to support this understanding. Again, if discussing classified matters with other cleared persons on an insecure channel—as opposed to “removing” and transmitting classified documents—were understood to be a federal crime, you’d expect it wouldn’t be hard to find instances of that crime being prosecuted. But in fact, there don’t seem to be any such cases. As Ben Wittes, a former national security reporter turned think tank scholar, wrote this summer:

People simply don’t get indicted for accidental, non-malicious mishandling of classified material. I have followed leak cases for a very long time, both at the Washington Post and since starting Lawfare. I have never seen a criminal matter proceed without even an allegation of something more than mere mishandling of senstive information. Hillary Clinton is not above the law, but to indict her on these facts, she’d have to be significantly below the law.

A common refrain in discussions of the Clinton case has been that “anyone else” would have been charged in similar circumstances. Yet if you look closely at the supposedly parallel cases where lower-ranking individuals have been prosecuted for mishandling, you find that invariably the circumstances aren’t similar. You’ll find prosecutions involving classified material knowingly and intentionally provided to uncleared persons (as in the case of Gen. David Petraeus), or where large quantities of documents were literally removed from secure facilities, but I haven’t turned up any cases where conversations about classified subjects on an insecure line have been treated as a criminal matter. In a country where literally millions of individuals hold top secret clearances, we may safely assume this is not because such indiscretions never occur, but because they had not been understood to be criminal acts, and were dealt with as matters of administrative discipline.

There are, it’s worth noting, a small number of e-mail threads in which a small portion of an actual classified document does appear to have been sent to Clinton’s private account. These were apparently marked (C) for confidential, though Clinton told the FBI that, at least in the context of an e-mail (as opposed to a classified document with appropriate headers), she did not recognize that marking as a classification marking. Many have found that claim dubious, but if she testified in court that she either failed to notice it or interpreted it as a section marker (as in, text copied from a document in which it had been preceded by paragraphs A and B), I’m not sure how one would prove otherwise beyond reasonable doubt. At best, then, one would be left with a potential, albeit unprecedented, case against the staffer rather than Clinton.

At last we come to the constitutional barriers to prosecution. As Comey made clear in his testimony to Congress, while the statutory standard in §793(f) is “gross negligence,” the Justice Department has historically been reluctant to prosecute—indictments under §793(f) are vanishingly rare—without something at least approaching evidence of intent. One reason for that may be that the Supreme Court’s opinion in Gorin v. United States (1941), which suggests that the Espionage Act’s intent requirements are an important feature that save it from unconstitutional vagueness. Due process requires that the law give citizens clear notice of what conduct is criminally prohibited, and in a country that (in large part for First Amendment reasons) has never had a British-style Official Secrets Act, there would be real reason to fear it would be too easy to unwittingly commit a crime absent relatively strict intent requirements. After all, classified information is routinely (and lawfully) printed on the front pages of newspapers—it would be perverse if forwarding a New York Times article were an indictable offense—and the sheer volume of classified material means it’s often difficult for officials to keep track of whether a particular fact is classified at a given time. Add in the historical paucity of prosecutions for mere insecure discussion of classified facts and it’s easy to imagine Clinton’s attorneys successfully sinking any case with a due process argument.

At this point you may be asking the same question a conservative friend posed to me when I sketched this argument: If it’s so clear that Comey was correct, wouldn’t he and the Clinton campaign have been shouting all this from the rooftops? That did give me some pause, but there are reasons each might have incentives to be circumspect. If, as Vladeck argues in the piece linked above, the Espionage Act simply doesn’t cover a whole category of “mishandling” classified information, intelligence officials like Comey, who expect cleared persons to take their obligations seriously, might not want to advertise the fact too loudly—and may want to leave their options open for the sake of future prosecutions or threatened prosecutions. The Clinton campaign, on the other hand, probably understands that a lawyerly argument to the effect that her conduct was not technically illegal doesn’t sell terribly well on the campaign trail. Whatever the technicalities, this whole affair reflects terribly on her judgement, and they’ve likely concluded there’s no advantage to giving the issue any additional exposure, even if it’s to mount a defense. The old campaign adage “if you’re explaining, you’re losing” applies in spades here.

So do the additional e-mails recovered from Huma Abedin’s laptop change any of this? It’s certainly possible to imagine scenarios in which it might: If those e-mails contained clear evidence of intent to obstruct, or unambiguous instructions from Clinton that classified documents should be sent to her personal account. But then, one could as easily speculate about finding a confession to Vince Foster’s murder. If, instead, the e-mails include more of the same sort of correspondence the FBI has already reviewed—even if it includes more threads in which classified topics were discussed—all of the above analysis would apply.

To sum up, there’s little question that, as Clinton herself now acknowledges, the use of a private e-mail server to conduct official correspondence was a serious lapse in judgement. Individual voters can determine for themselves whether they find it disqualifying in a commander in chief. But we don’t need to resort to theorizing about political chicanery to explain why she wasn’t prosecuted for it: The simple answer is that—fortunately for all politicians—not every act of stupidity is a federal crime.

In the search for ways to reform the flawed current welfare system, some form of basic income guarantee has received more attention. My colleague Michael Tanner has reviewed some of the related pros and cons, but most of those studies confined themselves to the immediate to short-run impact on work, leaving many important questions unanswered. A new paper from Daniel Price co-authored with Jae Song offers one of the first studies to analyze the long-term impact of cash assistance from the negative income experiments that took place last century. Their findings suggest “unintended and unexpected long-term consequences for recipients” and ambiguous effects on their children. It’s important to understand these effects when considering ways to try to address the many problems with the troubled status quo.

In the 1960s and 1970s, when interest in a basic income was at a peak, there were large-scale evaluations of the idea of unconditional cash transfers in the Seattle and Denver Income Maintenance Experiments (SIME and DIME, respectively). Recipients were given an unconditional cash transfer that was phased out as earned income rose, for a period of either three or five years. There were a host of studies analyzing the impact on poverty, work impact, and other measures, but these were more focused on the short-term effects.

This paper is so important because, as the authors explain, “virtually no other research has been conducted on the impact of cash assistance—or, indeed, any other type of government assistance—on beneficiaries themselves long after the assistance has ended.” Some of this is due to data limitations, and the authors are able to combine SIME/DIME date with administrative records from the Social Security Administration and the Washington State Department of Health.

The impact on poverty and material hardship is more straightforward and easier to measure, but some of the effects on work effort and other measures are still not as well understood. In this paper the authors find some evidence that participation did seem to have an impact on both work and earnings later in life for adult recipients. Participation reduced the probability a worker was active in any given year by 3.3 percentage points, and decreased average earnings by $1,800 (about 7.4 percent of mean annual earnings).

Another way to put it: for each $1 in additional government transfers, the authors find that an individual’s discounted lifetime earnings are $4.50 lower. 

Impact on Propensity to Work and Earnings for Parents, by Age

Source: Price and Song (2016).

Notes: Each data point represents the estimate and 95% confidence interval of the coefficient on a dummy for financial treatment status in one regression, limiting the sample to data from individuals when they are a certain age. Earnings variables are based on one observation per year for all years between 1978 and 2013.

These effects are mostly concentrated towards the end of a person’s working life between ages 50 and 60. Song and Price suggest three channels that might explain this pattern of some reduced work effort during the experiment, minimal effect immediately after, and then larger effects decades later as former participants near retirement age. If these adults saved some of those transfers, they might be more able to reduce work hours or retire earlier. If they worked less, the wages they were able to command might be lower because they had less time to develop their skills and human capital. Or it could change their preferences, by making them place more importance on leisure time that they were able to increase during the experiment.

This could have significant implications for discussions about how to address some of the worst shortcomings of the current welfare system. The substantial reductions in earnings and work effort should be incorporated into how we think about the ramifications of these programs.

The parents weren’t the only ones affected by these programs, their children were too. There is a wide range of plausible effects on the children: parental benefit receipt could the probability of their children also receiving benefits and reducing work effort. It’s also possible that the additional income from transfers will allow parents to invest more in their children and increase their human capital and earning potential later in life.

In this long-term analysis, the authors find “little evidence of an effect on children for any variable studied.” These include probability of applying for disability, propensity to work, and impact on annual earned income. The results for children should be considered with caution due to the number of tests that they run on child outcomes. These findings cut both ways, in a sense: it doesn’t seem to be the case that parents getting sizable cash assistance during the program duration discouraged work for their children, but there also weren’t substantial positive effects on children’s eventual work outcomes due to more investment or other effects of having higher household income. In their words:

Taken as a whole, our results suggest that cash assistance could have unintended and unexpected long-term consequences for recipients without significantly improving their children’s earning potential or decreasing their propensity to use government benefits. On the other hand, in our context, we can rule out the idea that cash assistance creates a welfare culture that decreases children’s earned incomes or their dependency on disability benefits by a large amount.

Impact on Propensity to Work and Earnings for Children, by Age

Source: Price and Song (2016).

Notes: Data from Denver families only. Each data point represents the estimate and 95% confidence interval of the coefficient on a dummy for financial treatment status in one regression, limiting the sample to data from individuals when they are a certain age. Earnings variables are based on one observation per year for all years between 1978 and 2013.

These findings highlight why experiments with rigorous evaluation are so important because there are significant effects that aren’t yet fully understood. Studies like this can produce surprising results and new evidence that forces policymakers and researchers to update how they think about these issues. There are some aspects of cash assistance and a basic income guarantee that the authors can’t analyze here, for example these experiments treated a small share of people within their respective cities and the impact of a community wide experiment could be significantly different. The one thing we do know is that much more research in this sphere is needed to understand the limitations and unintended consequences involved.

Pages