A new report on federal student loans from the National Center for Education Statistics came out today, and it is troubling. Much of the media attention is likely to focus on the default rates of borrowers who attended for-profit colleges—and they are atrocious—but the report’s contents condemn the entire system.
Delving into the data reveals that there is a whole lot of defaulting going on—among first-time students who began school in 1995-96 and took out federal loans, 13.7 percent had defaulted on their most recent loan—but there’s been a whole lot of deferring payment, too. The share of borrowers who were deferring or in forbearance stood at 13.3 percent. 31.8 percent were still repaying. And the 41.3 percent listed as “paid or closed without default” hadn’t necessarily fully paid off their loans, either. No, “this includes either loans that are paid off by the borrower or forgiven [italics added].”
As for students attending for-profit schools, yes they had the highest default rate. But, remember that for-profits take on the students with the greatest obstacles to success while receiving essentially no state subsidies or tax-preferred donations. Indeed, they pay taxes. Their default status for students starting in 2003-04—during the for-profit boom—is terrible at 34.8 percent. However, community colleges came in at 15.7 percent, which is also awful, given their low, directly subsidized prices and considering that, while their students often have significant obstacles, students at for-profit schools tend to have bigger ones. Of course, a smaller percentage of community college students borrow. And default rates at about 9 percent at public and private, nonprofit, 4-year colleges is hardly anything with which to be impressed.
The higher education system is flooded with taxpayer money producing oodles of negative results, including skyrocketing costs, credential inflation, and increasingly anemic learning. The federal loan story told by these data reinforces how much draining needs to happen, and a good place to start is phasing out federal loans.
China-based cryptocurrency exchange BTCC suspended all domestic trading in yuan last weekend. The decision came on the heels of a September 5 statement from regulatory authorities in China, which required all domestic cryptocurrency exchanges publish closing announcements, stop registering new users, and establish a schedule to cease yuan-denominated trading by September 15. Huobi and OKCoin — two other exchanges based in China — have announced similar plans to stop trading. To be clear: China has not banned the use of cryptocurrencies. It has banned cryptocurrency exchanges and initial coin offerings (ICOs). Even still, it has prompted some to consider whether a government might ban cryptocurrencies like bitcoin — and, perhaps more importantly, whether such a ban would be effective.
There seems to be no denying that governments can ban cryptocurrencies. Bolivia, Ecuador, Kyrgyzstan, and Bangladesh have already done so. Russia issued a draft bill to ban cryptocurrencies in October 2014 and recent rumors suggest it might follow through. More broadly, governments have taken steps to prevent other alternatives to their preferred monies. Cambodia recently suggested it might ban the dollar. Syria prohibited the use of any foreign currency in 2013. And the U.S. government shut down Liberty Dollar and E-Gold in the mid-2000s. Can governments ban cryptocurrencies? Absolutely. The question is whether — or, to what extent — a ban will actually discourage use.
Some bitcoin proponents have argued that governments cannot really prevent bitcoin use. Jon Matonis once stated that “a government ban on bitcoin would be about as effective as alcohol prohibition was in the 1920s.” How could a government prevent people from using bitcoin? It’s online. It’s pseudonymous. And, as Matonis notes, “demand for an item […] does not simply evaporate in the face of a jurisdictional ban.”
Or, does it? For starters, one must recognize that monetary demand — that is, the demand to use an item as a medium of exchange — is not quite like the demand for most other goods. Monies are subject to network effects. I can enjoy a fine bathtub gin even if no one else does. But the usefulness of a would-be money like bitcoin depends crucially on whether other people are using it. We must coordinate beliefs. If one does not believe others will use bitcoin, she will be less inclined to accept it herself. For cryptocurrencies that lack some non-monetary use, that means demand might fall to zero even if everyone would prefer it to the relevant alternative.
In general, governments might determine the medium of exchange by coordinating beliefs, employing transactions policy, and punishing users of alternatives. By declaring an item legal tender, for example, the government might create an especially salient focal point around which individuals can coordinate on a particular money. Legal tender status might be nothing more than a designation; it need not convey any special privileges under the law (though sometimes it does). Since I want to use the money you are using and you want to use the money I am using, simply stating that the dollar is legal tender and bitcoin is not legal tender might be enough to generate coordination on the dollar. As a large and powerful player in the economy, governments are often in a position to provide such a focal point.
Of course, if the net gains from switching to bitcoin are greater than the costs of coordination, we might establish some competing focal point to coordinate on the superior alternative. In this case, governments might resort to transactions policy — that is, committing to accept and spend its preferred money — in order to prevent bitcoin from gaining widespread acceptance. By collecting taxes and spending dollars (and not bitcoin), the government guarantees some demand for dollars and, correspondingly, limits the potential network size of bitcoin. Some governments will not be able to determine the medium of exchange with transactions policy. But a sufficiently large government can.
Even if a government is not large enough to determine the medium of exchange via transactions policy, it still has one last trick up its sleeve: punishments. By punishing those employing an alternative money, it lowers the expected benefits of the alternative and, hence, the relative demand for its preferred money. Whereas a sufficiently big government is required to determine the medium of exchange with transactions policy, a government of any size can determine the medium of exchange with punishments, so long as it is willing and able to mete out sufficiently severe punishments.
There are some obvious limits to the government’s ability to punish cryptocurrency users. For one, it has to find them. And, as Matonis and others have noted, cryptocurrencies like bitcoin are pseudonymous, making it difficult to tie an individual to his or her balance of bitcoin. Difficult, but not impossible.
Even if one has access to a perfectly anonymous payment mechanism, many transactions inevitably reveal one’s identity. When you purchase a good or service, there is usually some point in the transaction where you actually receive the good or service. And, at that moment, your identity is vulnerable to detection. That you can send or receive an anonymous payment is of little consequence if the other party in the transaction is a government agent. Sting operations are real.
More importantly, exploring the limits also reveals the vast range of transactions that would be relatively trivial to stamp out. A ban with significant punishments for those caught sending and receiving cryptocurrencies would surely see those “bitcoin accepted here” signs disappear. It would make finding a trading partner willing to use bitcoin a lot more cumbersome. Most people prefer to be on the right side of the law most of the time. Most routine transactions do not warrant the added costs of obscuring one’s identity or vetting one’s trading partner. Why risk being caught using a banned currency to buy milk and bread? Privacy is just not that important to most people in most situations. No doubt a government would find it difficult — perhaps even impossible — to eliminate all bitcoin transactions. But a committed government would have little trouble making bitcoin sufficiently unattractive for most users, significantly limiting bitcoin’s potential network size. In such a world, bitcoin would function as a niche currency — or, not at all.
There are exceptional cases, to be sure. If the government’s preferred currency is poorly managed (think: hyperinflation, not two-percent) or black market transactions are the norm, one might not hesitate to use a banned alternative. That seems to be the case in Venezuela at present. If one must operate outside the law just to buy lunch, she might not be too concerned about the risk of additional sanctions for using bitcoin. And, since everyone else is operating outside the law, she might be reasonably confident that others will accept bitcoin as well. But such cases are the exception. When bitcoin proponents maintain that governments cannot prevent bitcoin use, they do not usually limit the claim to such extreme scenarios.
Fortunately, very few countries have taken steps to ban cryptocurrencies to date. But the threat is legitimate. Governments might not be able to prevent all cryptocurrency transactions, but they can significantly discourage their use. With this in mind, we should continue to push for choice in currency. We should continue to explain the benefits of financial privacy and stateless monies. However, we should also support sensible regulation that would preserve most of the benefits from cryptocurrencies while eliminating the major justifications for outright bans. It is a second-best solution. In a world with powerful governments, it might be the best one can hope for.
[Cross-posted from Alt-M.org]
Earlier this afternoon, the House Judiciary Committee circulated its draft FISA Sec. 702 reauthorization bill. This is a preliminary readout of the major problems I see with this legislation.
Mandatory Data Destruction Not Mandatory
One of the biggest vulnerabilities Americans face today is the growing volume of their personal data being stored on servers in the private sector and in government. In the government counterterrorism (CT) context—and CT intelligence collection was the original rationale for this authority—there is simply no reason for the government to continue the collection and storage of the information of innocent U.S. Persons (a legal definition that includes citizens and legal permanent residents).
The bill as drafted would allow the government to do exactly that for at least 90 days for "foreign intelligence purposes" and it allows the Director of the NSA (DIRNSA) to waive that requirement on an individual and specific basis if DIRNSA determines that such waivers are "necessary to protect the national security." All this provision will do is create more paperwork for NSA, but the waiver process could no doubt be largely automated, rendering this alleged reform meaningless. A genuine reform would 1) explicitly prohibit the government from obtaining and maintaining the data of Americans unless said Americans were the actual target of an authorized criminal investigation, and 2) require mandatory external audits (read Government Accountability Office) to confirm said data destruction.
No Penalties for Lying to the FISA Court
In September 2017, Demand Progress issued a report highlighting the number of times the NSA and Department of Justice have been caught violating Sec. 702, FISA Court orders, or both. From the report's executive summary:
The FISC has twice found that certain Section 702 collection violated the Fourth Amendment. In 2011 the government revealed that as part of its “upstream” Section 702 collection it collected non-targeted, entirely domestic communications. When NSA violated the rules that were supposed to make this collection legal, FISC again deemed the practice “a very serious Fourth Amendment issue.”
For almost 12 years, both under Section 702 and other programs before it, NSA was always engaging in or retaining some kind of electronic surveillance the FISC would go on to deem unauthorized, and NSA would only fix the problem when threatened with criminal sanctions.
The draft House Judiciary bill makes no mention of these past violations, much less proposes any remedies. House and Senate members apparently need to be reminded that the Constitution's impeachment function is applicable to all civil officers of government who engage in such violations.
Hiding The Real Numbers
For over six years, Senator Ron Wyden (D-OR) has attempted to get two Administrations to come clean on the actual number of innocent Americans whose communications are swept up by the Sec. 702 program, to no avail. The House Judiciary draft gives the Director of National Intelligence (DNI) a pass on compiling and making public this information if the DNI deems that such a calculation is "not achievable."
If AT&T, Verizon, and the other carriers have no problem finding you and me to give us our monthly cellular bills, there's simply no valid excuse for the DNI to be able to bob and weave on providing the number of innocent Americans being caught in this digital dragnet. The House Judiciary Committee should not be in the business of legalizing the DNI's subterfuge on the issue.
There are a number of other reforms that should be in this bill, but it is important to remember that the underlying premise of the FISA Amendments Act and the PATRIOT Act—that 9/11 happened because we didn't collect enough information on the terrorists prior to the attacks—has been refuted by the Congressional Joint Inquiry and the 9/11 Commission. This draft is proof that when it comes to rolling back unnecesary and ineffective mass surveillance programs, those facts simply don't matter.
The Secretary of Homeland Security shall take such actions as may be necessary (including the removal of obstacles to detection of illegal entrants) to construct, install, deploy, operate, and maintain tactical infrastructure and technology in the vicinity of the United States border to deter, impede, and detect illegal activity in high traffic areas.
Media outlets are describing this as codifying Trump’s “border wall.” I have previously detailed the numerous problems with building a border wall, including the fact that it would require huge amounts of private land along the Southern border. This deprivation of the right to private property is serious, but it’s compounded by the fact that the government seizes the land first and only then, many years later in some cases, provides just compensation. Unfortunately, the Supreme Court has long ago signed off on this procedure. It’s a problem that Congress must fix.
The Problem of Seizing Private Land
Figure 1 is a map of the border that shows the federally owned portions in green. Tribal land, which comes with its own restrictions, is green with black stripes. The existing border fencing is in black and yellow. The yellow portions are vehicular barriers, and the bolded black is the pedestrian or "real" fence. The dotted line in Texas is the Rio Grande River. As you can see, most of Texas is without any barriers and is almost entirely privately owned.
Border Fencing and Federal Land
One reason why Congress built the fences where it did is due to the problems associated with seizing private land. In July 2007, Customs and Border Protection spokesperson Michael Friel explained to The Seattle Times that the fences “were going up first in New Mexico, Arizona and California, where much of the land already belongs to the federal government.” He added, “We realize that in Texas there are folks that own property, that have land on the border. That dynamic is different.”
DHS’s Inspector General (IG) concluded in 2009 that “acquiring non-federal property has delayed the completion of fence construction,” and that “CBP achieved [its] progress primarily in areas where environmental and real estate issues did not cause significant delay.” The IG report again:
For example one landowner in New Mexico refused to allow CBP to acquire his land for the fence. The land ownership predated the Roosevelt easement that provides the federal government with a 60-foot border right-of-way. As a result, construction of fencing was delayed and a 1.2-mile gap in the fence existed for a time in this area. CBP later acquired this land through a negotiated settlement.
The IG found more than 480 cases in which the federal government negotiated the “voluntary” sale of property, and up to 300 cases in which condemnation would be sought through the courts.
Legal Process and Legal Authority to Seize Private Land
Congress has already given the administration authority under a 1996 law and a 2006 law to condemn and seize land using eminent domain to build barriers. One way to address eminent domain along the border is simply to ban it. Rep. Ruben Gallego (D-AZ) has introduced a bill today that would do so. This would be effective, but it may not be politically feasible, given the wall fever that has descended on Congress.
Another approach would address the process. Right now, when Border Patrol wants to take someone’s land, they send them a letter offering them a nominal low sum of money for their land and threatening to file condemnation proceedings against them if they don’t accept it. In 2006, when the Secure Fence Act fences were built, many property owners accepted the low offer because they did not understand their right to negotiate over just compensation in court. Just compensation is a constitutional guarantee. Under the 5th amendment, “private property [cannot] be taken for public use, without just compensation.”
Just compensation refers to the fair market value of the property seized—what you could get for the land if you attempted to sell it—but less than what you would demand to receive in a voluntary sale. But in many cases, the seizure of a single strip of property in the middle of someone’s property can depreciate the value of the entire land. For this reason, it is necessary to present evidence in a court of the total impact of the seizure to the landowner. Other issues that may arise in this process are the exact boundaries of someone’s property and who exactly holds financial interests in the land. These issues also take time to sort out.
Seizures without Just Compensation
Here’s the problem: under the eminent domain statute, the federal government can seize property almost as soon as they file a condemnation proceeding—as soon as the legal authority for the taking is established—then they can haggle over just compensation later. It's called "quick take." Quick take eminent domain creates multiple perverse incentives for the government. 1) They can quickly take land, even when they don’t really need it, and 2) they have no real incentive to compromise or work with the land owner on compensation. The land owner’s bargaining power is significantly diminished. The federal government already possesses the property.
This means that for years, people who are subject to a border wall taking go without just compensation. The government is supposed to compensate the landowner for this time by paying interest on the agreed amount. But in the real world, many people cannot survive for years being deprived of income that they might have from the land. According to an NPR analysis of 300 fence cases, the resolved cases took more than three years to resolve. In other cases, the process took seven, eight, or even 10 years. Some cases are still pending a decade on.
Congress could rectify this injustice by requiring the federal government to work out just compensation before the wall is built or, better yet, before the land is taken. That would give the landowner a fair position to negotiate with the government and give the government a reason to respect their rights. That it would slow up a pointless waste of taxpayer dollars is just an added bonus.
E-Verify is the supposed silver-bullet of immigration enforcement. Despite its serious and unsolvable problems, the House Judiciary Committee was going to have a markup today on the Legal Workforce Act (LWA) that would mandate E-Verify for all new hires in the United States. Although they canceled the markup at the last moment, this is still a wonderful opportunity to explore the main reason why E-Verify is ineffective: employers ignore it.
E-Verify is a government system whereby employers enter the identity information of new hires via an online portal. 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.”
The states of Alabama, Arizona, Mississippi, and South Carolina have mandated E-Verify for all new hires in their states. Arizona was the first to mandate it on January 1, 2008, South Carolina mandated it on July 1, 2010, Mississippi on July 1, 2011, and Alabama on April 1, 2012. In those four states, the law demands that every employer must run every new hire’s identity information through the E-Verify system. The response to a Freedom of Information Act (FOIA) request filed by Cato shows that there are far fewer E-Verify cases or queries than there are new hires in these states, which means less than 100 percent of new hires are actually being run through the system (Table 1).
Percentage of New Hires Run Through E-Verify by State
Sources: Author's Calculations of Longitudinal Employer-Household Dynamics of the U.S. Census and Cato FOIA.
The number of E-Verify cases does differ somewhat per year compared to older data but the conclusion is similar: Only 57.6 percent of all new hires were run through E-Verify in 2015 in states where 100 percent of all new hires were supposed to be verified. The best-performing state was Arizona, which saw marked increases in E-Verify usage since its 2008 implementation. In 2014, Arizona had about 1.9 times as many illegal immigrants as Alabama, Mississippi, and South Carolina combined.
At the very minimum, E-Verify cannot be effective if employers do not use it. And it’s no wonder so many employers ignore E-Verify as it comes with a 17-page memorandum of understanding and 139-page User Manual that employers must understand in order to run the program properly. Few people who are trying to run a business want to take the time to master the details of this complex government system just so they can hire somebody. Even if employers do take the time to master E-Verify, it does not provide a safe harbor from future government audits, as employers across the country have discovered. The 2006 immigration raid of Swift & Company, a Colorado-based meatpacker, found that 10 percent of the firm’s workforce were illegal immigrants even though Swift had used E-Verify since 1998. If E-Verify doesn’t work when it’s used, employers aren’t protected when the system makes errors, and they can still be punished when they rely on that system, the real question is why would any employer would actually use it?
Increasing E-Verify compliance would require worksite visits and remote audits, just like the current I-9 system. If Arizona, Alabama, Mississippi, and South Carolina cannot assure better than 73.6 percent compliance with E-Verify—all states with large political constituencies that demand immigration enforcement—how well will a nation-wide mandate fare in states that don’t have such constituencies? Not well.
The low E-Verify compliance rate in states that have mandated the system indicates that it will fail to demagnetize the wage magnet if Congress ever mandates the LWA or a similar piece of legislation nationally. At that point, policymakers will demand more expensive and intrusive methods to guarantee that employers hire only legal workers, such as a biometric identity card. The major problems with E-Verify are economic, not technical. E-Verify has many serious problems but the low compliance rates should dampen enthusiasm among its supporters.
The other shoe is about to drop in the Boeing-Bombardier trade row. But first, some background...
Last week, smack dab in the middle of the third round of the NAFTA renegotiations taking place in Ottawa, the U.S. Department of Commerce issued a preliminary determination in a countervailing duty case brought by the Boeing Company in May. The Countervailing Duty Law provides “relief” (usually in the form of import duties) to domestic industries that can demonstrate that they are “materially injured” or threatened with material injury by reason of sales of subsidized imports.
In early summer, the U.S. International Trade Commission ruled, preliminarily, that there was a reasonable indication that U.S. manufacturers of large civil aircraft (i.e., Boeing) may be threatened with material injury by reason of prospective sales of aircraft from Bombardier to Delta Airlines, which may be offered at artificially low prices made possible by various government subsidies to the Canadian producer.
Subsequently, Commerce’s investigation turned up 16 different subsidy programs—equity infusions, launch aid, “provision of land for less than adequate remuneration,” various tax credits and incentives, and federal and provincial grants—constituting specific benefits to Bombardier by the governments of Canada, the United Kingdom, and the province of Quebec, which amounted to an aggregate subsidy rate of 219.6 percent ad valorem.
By historical standards, that is a very large number. If finalized at that rate, the duty would put the U.S. market out of reach to Bombardier and—of greater significance to the U.S. economy—put Bombardier airplanes out of reach to U.S. carriers, reinforcing Boeing’s monopoly power, and ensuring higher costs of air travel and air shipping in perpetuity.
Understandably, many on both sides of the border are upset over these findings. Recriminations and demands for retaliation have been swirling. Canadian Prime Minister Justin Trudeau has threatened to cancel his government’s planned purchases of Boeing fighter jets. Even the UK government, concerned about the future of Bombardier’s manufacturing operation in Northern Ireland, has discussed retaliation.
Many analysts are interpreting Commerce’s announcement of these results as a manifestation of Trump’s “America First” worldview, with its timing intended to secure some leverage for U.S. negotiators in the NAFTA talks. But it is in no way apparent how this finding could or would be used to extract concessions from the Canadians somewhere in the negotiations. Meanwhile, the fact is that determination dates in trade cases are set according to statute (there is some scope for extensions), and this prelim was set well before the NAFTA negotiations were scheduled, which brings us to another unfortunate set of circumstances.
Just as passions are subsiding from last week’s tempest, today the Commerce Department will announce its preliminary finding in a companion antidumping case, which was also filed by Boeing in May. The Antidumping Law provides “relief” (usually in the form of import duties) to domestic industries that can demonstrate that they are “materially injured” or threatened with material injury by reason of “less-than-fair-value” imports (sales made at prices in the United States that are lower than “Normal Value.”). This is a very, very, very, very, very, very, very, very, very, very, very, very bad law, deceptively invoked under the guise of ensuring fair trade and level playing fields, which has no economic justification and is used increasingly by U.S. companies as a weapon of domestic commercial warfare to kneecap U.S. competitors and their own U.S. customers. As was the case with respect to the countervailing duty matter, the U.S. International Trade Commission ruled earlier this summer that there was a reasonable indication that domestic industry was threatened with material injury by reason of less-than-fair-value imports.
Based on the unscrupulous analysis that Commerce seems to have teed up in the AD case (the capricious details of which are described here and here), the results are likely to further inflame the situation and threaten progress in the NAFTA talks, if not North American trade relations writ large.
By the end of this year, Commerce will attempt to verify information on the record, accept new information, and modify its results, accordingly, in these companion cases. But it’s rare that Commerce makes changes favorable to the foreign exporter or U.S. importer between the preliminary and final determinations. Ultimately, the question of whether duty orders will be imposed comes down to the final injury determination rendered by the U.S. International Trade Commission. If the ITC finds that Boeing is not threatened with material injury because, for example, it finds that Boeing doesn’t even produce (nor is it capable of producing over the next few years) the kinds of aircraft that Bombardier is hoping to sell to Delta, then the cases will both terminate and all will be well. That decision is due in February 2018.
Or, if duties orders are imposed, the decisions can be challenged by Bombardier, Delta, or other parties in U.S. court or in a NAFTA dispute panel. Although the Canadians seem to have a preference for the NAFTA panels, it is highly likely that the U.S. Court of International Trade would find all sorts of overreach by Commerce, if the Commerce analysis is based on the fictitious sales and incomplete cost data that is on the record.
In the meantime, maybe trade analysts, policymakers, and the public can think more deeply about whether these trade laws really serve U.S. interests. The laws, as written, preclude objective analysis at the ITC, forbid consideration of the effects of these punitive duties on downstream U.S. companies and consumers, and give the Commerce Department vast discretion over administrative matters that dramatically affect the bottom line—the duty rates calculated and applied. Pointing the finger at Trump and his America First policies (an understandable impulse that has been on display this past week) instead of focusing on the disruptive effects of these commercial weapons, which are easy to self-administer and operate on statutory auto pilot, wastes an important opportunity to achieve greater awareness and, possibly, some reforms. Why not put these trade remedy laws on the NAFTA negotiating table? Really, how can one NAFTA country's producers be dumping in another NAFTA country when nearly all tariffs are zero and there is no protected market from which to cross-subsidize cheap exports? Let's make these laws inutile among the NAFTA countries. Or push for a public interest test that could authorize the ITC to actually analyze the adverse impact of duties on downstream industries. Instead of piling on and lazily blaming Trump, let's figure out how to rein in these unbalanced laws that wreaked commercial havoc during the Obama, Bush, Clinton, Bush, and Reagan adminstrations.
CNBC reports that the burger chain Shake Shack is planning to trial a new restaurant in New York which will not have a traditional cashier’s counter. Instead, “guests will use digital kiosks or their mobile phones to place [and pay for] orders.” Their order will be processed immediately to the kitchen and the guest will receive a text message when their food is ready.
Great, you might think. Shake Shack is investing in innovations which could improve the productivity of remaining workers, increasing wages (indeed, they want to pay the lower relative number of staff in this restaurant at least $15 an hour). Such investments might provide a more efficient and desirable service to customers too. This frees resources and excess labor for other more productive pursuits in the economy.
But the kicker for why Shake Shack is undertaking such investments comes later in the article:
it's likely that in the next 15 to 20 months that areas like New York, California and D.C., in which there are many Shake Shacks, will transition to a $15 minimum wage…Adopting this payment policy in Astor Place will give the company a chance to work out the kinks before it rolls out a $15 minimum wage in these locations.
Anyone who has been to a McDonald’s in France will know what’s going on here. Shake Shack suspects that the cost of labor will rise due to an increased minimum wage, and given that projection, it’s become economic to consider investments in labor-saving technologies. Higher minimum wages act in effect as a subsidy to automation.
But these investments for productivity improvements don’t come for free. A recent paper by Grace Lordan and David Neumark finds empirical evidence showing that between 1980 and 2015, increasing the minimum wage by $1 decreased the share of low-skilled automatable jobs by 0.43 percent in general and by 0.99 percent in manufacturing. Other jobs might be created of course, but they may well be more demanding or stressful, such as overseeing the running of multiple machines or having to have the skills to deal with technical problems etc. “Regulating to innovate,” subsidizing the rapid introduction of some technologies before they are actually high quality and cost effective, drives up prices for consumers too.
Perhaps more pertinently, low-skilled workers younger than 25 and older than 40, especially women, tend to be particularly affected by the disemployment effects of automation and can find it very difficult to find replacement work given their productivity levels.
As I concluded in a recent Daily Telegraph article:
If we are moving into a period when technological innovations are speeding up, we could be hiking minimum wages dramatically at just the wrong time. It will prove enough of a policy challenge as it is, to equip people with new skills to adapt in a rapidly changing labor market. Making more low-skilled jobs uneconomic by artificially hiking the cost of labor substantially could exacerbate this change at a time before new investments would otherwise make economic sense.
Being worried about this consequence is not to be anti-technology or anti-innovation. We all recognize that mechanization and technological innovation are the only way to sustainably raise living standards. But encouraging new investments by raising business costs and driving out low-skilled jobs is another matter entirely.
Just because Luddite efforts to destroy machines was economically harmful does not mean that destroying low-skilled employment opportunities would be beneficial.