Cato Op-Eds

Individual Liberty, Free Markets, and Peace
Subscribe to Cato Op-Eds feed

Many Puerto Ricans are still without power after Hurricanes Irma and Maria knocked out the island’s electric grid in the Fall. The federal government has poured in resources, but there are still hundreds of thousands of people without power. The sad episode has highlighted the gross failings of Puerto Rico’s government-owned electricity infrastructure. 

Even before the hurricanes, the power grid in Puerto Rico was in bad shape. A 2016 audit of the state-owned Puerto Rico Electric Power Authority (PREPA), found that “transmission and distribution systems are falling apart quite literally: they are cracking, corroding, and collapsing.” The audit found very old and dilapidated assets, and also that the utility has inexperienced staff, terrible record keeping, and a vastly bloated bureaucracy. PREPA has had a poor environmental record, and its power systems have constantly broken down, causing costly forced outages.

Before his resignation in November, the head of PREPA indicated that a “transformation plan” was in progress. But the bankrupt government company has no way to pay for modernization since it is heavily indebted and its finances are a shambles.

The good news is that Puerto Rico’s Governor, Ricardo Rossello, has unveiled a plan to privatize PREPA. Calling the company “a heavy burden on our people, who are now hostage to its poor service and high cost.” PREPA, he said, “does not work and cannot continue to operate like this.”

Rossello’s plan should be good news for Puerto Ricans, who are sick and tired of high-cost and unreliable power. Vast experience around the world since the 1980s shows that privatizing infrastructure, including electric utilities, increases operational efficiencies, improves capital investment, and enhances customer service. Moving PREPA’s assets to the private sector, within an appropriate regulatory framework, should reduce costs and lead to efficient new investment in Puerto Rico’s electric system.

Puerto Rico’s electric utility is not the only one that should be privatized. President Trump’s new budget proposes privatizing the Tennessee Valley Authority and the federal Power Marketing Administrations. Those reforms are long overdue, as privatization of state-owned businesses has swept the world outside of the United States since the Thatcher reforms of the 1980s.

Read more about privatizing the TVA and PMAs here and here, and more about the general benefits of privatization here.

Cato intern Johnathan Postglione helped assemble this post.

A recent New York Times article discusses an exemption for heavy-duty diesel truck emissions standards. First enacted by the Clinton administration, the standards include a provision that exempts truck engines built before 1999. Some entrepreneurs are taking advantage of this provision to circumvent the regulations by installing rebuilt pre-1999 engines in new truck bodies. According to the article, the rebuilt trucks sell for at least 10 percent less than compliant new trucks and also cost less to operate.

The focus of the article is the lobbying directed towards maintaining the exemption, but I want to focus on the real underlying problem: the exemption of existing sources from more stringent emission standards, also called grandfathering. If emissions and degraded ambient air quality have negative effects on morbidity and mortality, then those effects occur regardless of whether the emissions come from “existing” or “new” sources. So there is no public health or scientific rationale for grandfathering. Emissions are emissions regardless of their origin.

Instead, grandfathering eliminates the expense of retrofitting or scrapping existing durable equipment and imposes the cost of compliance gradually. This reduces the explicit cost of compliance and thus limits political opposition to emission control.

The downside of grandfathering is that it incentivizes the retention and operation of old equipment. In the cover story of the spring 2006 issue of Regulation, Shi-Ling Hsu describes the thirty-year struggle over the imposition of emission controls on coal-fired electricity power plants. When the Clean Air Act was amended by Congress in 1977, existing plants were exempted from the pollution control measures that were required of new power plants. Ever since, much ink has been spilled, and large legal fees and campaign and lobbying expenditures have been spent in the struggle over whether emission controls would ever be imposed on such plants because the right to emit without controls is very valuable.

The emissions loophole for trucks is similarly valuable, and old trucks are also exempt from excise taxes on new trucks and other regulations increasing the value of the exemption. 

What should be done about grandfathering? One possibility is emission taxes rather than regulation, but that works only if the politics of tax exemption differ from the politics of regulatory exemption. A second possibility is to set a certain end date to the grandfathering exemption. But that requires the legislature to resist the temptation for a grandfathering extension in return for political support, the subject of the Times article. Experience suggests this temptation is large.

Assuming that limiting emissions is a worthy goal, emission standards or taxes should be applied equally to all sources of emissions. So good public policy would not grandfather, but instead would require Congress to impose explicit costs on all existing emitters or explicit taxes on the rest of us to pay for emission reduction. And that appears to be as difficult as it sounds.

Written with research assistance from David Kemp.

Ballotpedia asked five redistricting buffs to comment on how the Supreme Court might rule on the two gerrymandering cases it is considering, Gill v. Whitford from Wisconsin and Benisek v. Lamone from Maryland. My response

One clue is timing. Rather than fast-track consideration of Benisek v. Lamone, the Court instead set an oral argument date of March 28. That leisurely pace ensures that any decision will come late enough in the term to wreak havoc on this year’s [election] cycle if it announces the imposition at once of a new constitutional standard. Few if any Justices wish to wreak such havoc avoidably. From which I deduce that the Court either 1.) does not expect to announce a new standard, or 2.) expects to do so only in a staged or prospective way that allows states time for compliance or kicks in with the next Census cycle.

While I believe Justice Kennedy remains open to the development of some constitutional standard in this area, I also think he is looking for a fix that is objective and mechanical enough that 1.) it yields the same results from state to state and from Republican-appointed as from Democratic-appointed lower court judges, and 2.) the clarity of what it calls for and how to comply cuts off a need for continual massive litigation and judicial supervision (compare here the success of the one-person, one-vote revolution). Kennedy’s having joined with the conservatives to stay the Gill ruling leaves me thinking that as of then he wasn’t convinced that the “efficiency gap” standard fit that demanding bill.

I’m also pleased to note that my essay on gerrymandering from a libertarian/classical liberal perspective, which appeared originally in Cato Unbound, is featured in the latest Cato Policy Report. You can read it here

Terrible mass shootings like the one at a Parkland, Florida high school are so shocking that it is easy to get the impression that mass shootings are increasingly common.  The number of deaths from mass shootings has been unusually high since 2007, because of five horrific incidents – Las Vegas (58), the Orlando nightclub (49), Virginia Tech (32), Sandy Hook (27), and the Texas First Baptist Church (26).  Statisticians would never try to fabricate a trend from such a small sample, even though the untrained eye may want to.

Last November, however, a Wall Street Journal essay by Ari Schulman claimed,

It isn’t your imagination: Mass shootings are getting deadlier and more frequent. A recent FBI report on “active shooters” from 2000 to 2015 found that the number of incidents more than doubled from the first to the second half of the period. Four of the five deadliest shootings in American history happened in the past five years, and 2017 already far exceeds any previous year for the number of casualties.

That FBI report “identified 160 active shooter incidents that occurred in the United States between 2000 and 2013,” with 486 people killed. The authors literally drew a straight line between just one incident in 2000 (after many in 1999) and 13 incidents in 2013, and called that a “rising trend.”

It is interesting, however, that schools have been the second-highest risk location. The FBI data show that the largest number of active shooting incidents from 2000 to 20016 were in workplaces and other commercial buildings (43%), followed by education facilities (22%), then open spaces (13%), government buildings (11%), residences (5%), health care facilities (3%) and houses of worship (4%).

The cited FBI data from 2000 to 2015 omit the two biggest mass shootings after 2015 and others before 2000.  In addition to Columbine, there were four other mass shootings in 1999, bringing yearly fatalities to 42 fatalities. We can’t be sure which mass shootings were “the worst in American history,” because (1) history didn’t begin with 2000, and (2) Congress didn’t define mass shootings as 3 killed until 2013, and (3) systematic data about such incidents were not collected until 2012.  

Schulman mentioned a longer time series from Mother Jones, but not any data from it, so I created the graph below from the Mother Jones data. This project began in 2012 and attempted to recreate earlier years from news records going back to 1982. Early years report at most one or two incidents per year, which may indicate “headline bias” – finding only those incidents that were sufficiently sensational to attract national news coverage.  

Importantly, the Mother Jones figures define mass shootings as public attacks in which the shooter and victims were generally unknown to each other, and four or more people were killed.  Unlike the FBI’s “active shooting incidents,” where gangs and drugs are frequently involved, Mother Jones excludes all multiple murders related to drugs, gangs or domestic violence. They do include mass shootings by jihadist terrorists, however, which accounted for only 4 of their 98 incidents by my count.

The Mother Jones writers claim that “A recent analysis of this [Mother Jones] database by researchers at Harvard University, further corroborated by a recent FBI study, determined that mass shootings have been on the rise.”  We already questioned the FBI trend.  What about those “researchers at Harvard University”?   Unlike the FBI, who compared the number of incidents between 2000 and 2013 to suggest such a rise, the trio of Harvard and Northeastern University researchers settled for only three years.  

Rather than counting annual changes in a small number of mass shootings as the FBI did, the Harvard-Northeastern team instead counted the average period of time between incidents, and found them more frequent from 2011 to 2013 than the average from 1982 to 2010 (although the journalists’ count before 2012 is doubtful). 

“The rate of mass shootings in the United States has tripled since 2011, according to an [October 15, 2014] analysis [of Mother Jones’ data] by researchers from Harvard School of Public Health and Northeastern University.”  That press release was not from the Department of Criminology, but from a subsection of the School of Public Health, which specializes in thinly-veiled advocacy of tough gun control laws.  “Since 2008,” they note, “we have received funding from the Joyce Foundation to write dozens of scientific articles on firearms issues, to disseminate findings through press releases, and Bulletins.” 

It seems more transparent to simply examine annual estimates from the graph. Adding a preliminary estimate of 17 deaths from Parkland to the Mother Jones list brings the total number of deaths up to 816 from 98 mass shootings between 1982 and early 2018 – or just 23 deaths per year.  That makes this sort of random mass shooting one of the rarest mortality risks imaginable. Falling or the flu are far more dangerous. Even when it comes to guns, 23 deaths a year pales next to the number of homicides by firearms in 2014 alone, which was 11,208 (69% of all homicides)  and the number of suicides by firearms, which was 21,386 (50% of all suicides).

Every time one of these random mass shootings occurs, journalists and legislators invariably seize on the tragedy to lecture about the need for artfully unspecific changes in federal gun control laws. Of all the risks posed by guns or knives, however, random mass shootings are among the least likely.

The Senate is currently debating many competing proposals that would legalize some Dreamers, enhance border security, and reform legal immigration.  This morning, the Department of Homeland Security (DHS) issued a press release criticizing the bipartisan Rounds-Collins proposal that has more support in the Senate than any other amendment.  This DHS press release was accompanied by a veto threat from President Trump.  Most of DHS’ talking points against Rounds-Collins are either hyperbolic, half-truths, or just inaccurate.  Below, I will respond to the three most egregious sections of the DHS press release.


By halting immigration enforcement for all aliens who will arrive before June 2018, it ignores the lessons of 9/11 and significantly increases the risk of crime and terrorism.


There have been nine terrorists who entered the United States illegally since 1975 who went on to plan or commit an attack on U.S. soil.  One of them, Glen Cusford Francis, managed to kill one person in an assassination on U.S. soil.  That’s one successful murder in a terrorist attack over 43 years, committed by one of the roughly 49 million illegal immigrants who entered during that period (most left the United States).  The annual chance of being murdered by an immigrant who entered illegally was about 1 in 11.6 billion per year during that time.

The Rounds-Collins amendment does not increase the threat from immigrant criminals.  Immigrants are already less likely to be criminals relative to native-born Americans.  Furthermore, Rounds-Collins does not limit the ability of law enforcement to track down immigrant criminals or to deport those convicted of a crime.  In fact, it focuses interior immigration enforcement resources on actual criminal and national security threats rather than dissipating them on raids.  Thus, the new enforcement priorities of Rounds-Collins are more likely to reduce the already low threat posed by illegal immigrant criminals and national security threats.         


It eviscerates the authority of DHS to arrest, detain, and remove the vast majority of aliens illegally in the country by attempting to limit DHS enforcement by codifying a “priorities” scheme that ensures that DHS can only remove criminal aliens, national security threats and those who arrive AFTER June 30, 2018 creating a massive surge at the border for the next four months.


This is simply untrue.  The enforcement priority section of Rounds-Collins just forces DHS to prioritize the removal of illegal immigrants who have committed crimes, as well as national security threats, over immigrants whose only offense was violating immigration law – unless they entered after June 30, 2018.  Furthermore, the relevant section of the bill does not prevent DHS from enforcing immigration laws against non-priorities, but it does remove the open-ended discretion.

New priorities for interior enforcement won’t much affect illegal entries across the border.  Southwest border apprehensions are at very low historical levels and a prioritization of interior immigration enforcement toward what the Obama Administration had in place in his second term certainly doesn’t open the border to illegal entries.  Amnesties have not historically increased the flow of illegal immigrants but appear to actually reduce them, for a time at least.  However, more border enforcement tends to lock illegal immigrants inside of the United States who would otherwise have left, boosting the stock but not the flow. 


Note: After publishing this, I learned that the Senate pulled this back to January 2018.  That change weakens DHS’ argument further.


By keeping chain migration intact, the amendment would expand the total legalized population to potentially ten million new legal aliens – simultaneously leading to undercutting the wages of American workers, threatening public safety and undermining national security.


DHS provides no estimate of how this legalization would expand “the total legalized population to potentially ten million.”  Legalized Dreamers wouldn’t be able to legalize their parents under this bill, which is essentially a restatement of current law.  Given the current long backlogs in the family-sponsored immigrant system and that most of the Dreamers legalized come from countries afflicted by those, there’s really no way that 10 million additional legal immigrants would be able to enter through this.  The best recent evidence is that each new immigrant eventually sponsored about 3.5 new immigrants through family reunification.  Because Dreamers’ ability to sponsor family members is limited and the wait times for many of the green card categories are so long, the number will certainly be lower for the legalized Dreamers.  It’s also unclear why that would be an argument against the Rounds-Collins amendment from the perspective of the agency tasked with enforcing it. 

The effect of immigrants on the wages of American workers is small and, if it is negative, is entirely concentrated on high school dropouts.  There are far better ways to help them than reducing economic and wage gains for the 90 percent of the workforce in educational categories positively impacted.  Furthermore, the effect of immigration on the wages of blue-collar workers is slightly positive.  There is not a good economic argument for reducing legal immigration.  It’s unclear why DHS would even be making this argument, since they aren’t supposed to be economic central planners but enforcers of immigration law. 

Legal immigrants entering on green cards from 1975 through 2017 have murdered 16 people in terrorist attacks on U.S. soil.  Assuming all of those green cards were issued in the family reunification categories or through the diveristy visa lottery, the chance of being murdered in a terrorist attack committed by a chain immigrant or a diversity visa recipient was about 1 in 723 million per year.  Even if that annual rate of deaths in terror attacks were to increase 10-fold, it would still not be a serious threat to national security. 

DHS’ open lobbying against Rounds-Collins via press release is extraordinary.  Past secretaries of DHS have certainly testified in favor of other immigration bills, but press releases by administrative agencies taking a clear side in a current debate in the Senate is a new expression of the power of administrative agencies.  DHS is openly lobbying for more resources that would enhance its power and respect.  That’s not a surprise according to what we know about how bureaucracies behave, but it is rare to see it expressed to nakedly in press releases during a Senate debate.

DHS should either not comment on these issues through press releases while the debate is occurring or only do so to make points about the feasibility of enforcement.  At a very minimum, the bureaucracy should not use policy-based talking points that are hyperbolic, half-truths, or inaccurate.   

George Hodgin’s mission seemed simple: manufacture uncontaminated, chemically consistent cannabis for use in scientific research on marijuana’s medical effects, all while complying with federal regulations surrounding the production of a drug still classified by the Drug Enforcement Administration (DEA) as highly dangerous. Despite new rules the DEA promulgated eighteen months ago, with the stated goal of allowing expanded cultivation of marijuana for scientific research, George Hodgin is still in administrative limbo. 

Hodgin, a former Navy SEAL, approached us recently for advice after encountering numerous regulatory roadblocks.  We have no special knowledge or ability in that direction; but perhaps publicizing his endeavors will nudge public opinion (and regulators) in the right direction. 

Expanding research access to high-quality marijuana is important. The Marijuana Policy Project estimates that roughly 2.5 million patients use medical marijuana – just in states with legal medical marijuana programs. This number is likely an underestimate, as it does not account for individuals obtaining marijuana for medical use through non-medical channels. Marijuana’s illegality at the federal level prevents the collection of much needed data that could help drive future research.

Veterans represent a particularly important category of medical marijuana users. Although Veterans Health Administration physicians are prohibited by federal law from recommending medical marijuana, new guidelines issued in December 2017 revise existing standards to encourage doctors and patients to discuss the use of medical marijuana without fear of recrimination. 

A January 2017 report from the National Academies of Sciences, Engineering, and Medicine acknowledges the difficulty researchers face in acquiring appropriate cannabis products and recommends actions be taken to ameliorate the situation; yet, obstacles remain. 

The major difficulty facing those who want to produce or obtain cannabis products for research purposes is that the DEA currently classifies marijuana as a Schedule I drug, which means the DEA views it as having “no currently accepted medical use.” But thirty states and the District of Columbia have laws permitting the medical use of marijuana. Additionally, the research available suggests potentially widespread medical applications for marijuana. 

Legal restrictions on access have severely hampered medical research into marijuana’s possible effects. One solution would be to reschedule marijuana, yet the DEA has repeatedly denied petitions to do so. In a July 2016 denial, the DEA asserted that marijuana has no currently accepted medical use in the United States because, among other reasons, “the scientific evidence is not widely available.”

This will be true, by definition, so long as the DEA makes it virtually impossible to conduct scientific research. Catch 22.

In August 2016, the DEA changed its rules to allow registered entities to supply researchers with the quality and quantity of marijuana needed for scientific study. Despite the rule change, however, the DEA has yet to approve a single application. In testimony given to the Senate Judiciary Committee last October, Attorney General Jeff Sessions confirmed that 26 applications are outstanding from marijuana suppliers; but the DEA has not approved a single one. 

The DEA’s resistance is part of the long-running, tough on drugs approach it has taken since the 1970s. As recently as 2013, courts upheld the DEA’s monopoly on the production of (federally) legal marijuana. The current administration shows no intention of changing this stance. 

In a recent poll, 91 percent of Americans supported the use of medical marijuana. For years, the DEA’s opposition to legalization or rescheduling has hinged on the argument that no scientific evidence suggests medical benefits. Despite hopes that the August 2016 rule change would usher in a new era of research, the past year and a half has proven that inaction remains the status quo. Until something changes, countless sufferers of pain, PTSD, and other ailments will remain trapped in a legal gray zone. Allowing properly vetted companies to manufacture and distribute research grade marijuana legally is the least the DEA can do.

Research assistant Erin Partin contributed to this blogpost.


Seven GOP Senators have proposed a plan that they claim would fulfill a pledge by President Trump to provide permanent residence (a pathway to citizenship) to 1.8 million young immigrant Dreamers. Sen. Tom Cotton (R-AR) appeared to go further than a mere “pathway” alone, claiming that it would actually provide citizenship itself to 1.8 million. Sen. James Lankford (R-OK) made the same claim, stating that he expects “1.8 million [to] go through naturalization.”

In reality, only an estimated 877,100 people would receive permanent residence under the White House-Senate GOP plan – and only approximately 587,650 should be expected to receive U.S. citizenship. A realistic number of those who may benefit from the White House plan, as embodied in a GOP Senate bill, is important because supporters have used the 1.8 million figure to justify large-scale reductions in the number of legal immigrants entering the country – potentially 22 million fewer legal immigrants over 50 years.

The 1.8 million figure is fiction. Based on the experience of prior documentation efforts and the specifics of this particular proposal, the GOP senators’ Secure and Succeed Act would provide an initial status to about 1.1 million. But of them, only about 877,100 would likely receive permanent residence, sometimes called a “pathway to citizenship,” and only about 587,650 would likely end up receiving citizenship. Table 1 provides the actual enrollment rates and extension rates for DACA compared to estimates for the Secure and Succeed (S&S) Act.

Table 1

Sources: Authors’ calculations based on Migration Policy Institute (DACA Eligibility); Migration Policy Institute (S&S Eligibility, LPR Rates); Pew Research Center (Naturalization rate); U.S. Citizenship and Immigration Services (DACA Enrollees; DACA Extensions); S&S Initial Enrollment Rates Based on Congressional Budget Office. *An individual cannot apply for citizenship from DACA.

The S&S Act creates a four-part framework for potentially receiving permanent residence and later citizenship (see Table 2 at the end). First, Dreamers would need to meet a set of basic criteria to receive a conditional residence status valid for up to 7 years. Second, after 7 years, they could apply for an extension under a second set of stricter criteria. Third, at any time after the extension, they could apply to have the “conditions” removed and receive full permanent residence status with a pathway to citizenship under a third set of criteria. Fourth, they could apply for citizenship after another 7 years and more conditions. Each stage is fraught with obstacles for the about 3.3 million unauthorized immigrant Dreamers who entered the United States as minors several years ago.

Why the Secure and Succeed Act Won’t Provide Even Temporary Relief for 1.8 Million People

Under the Secure and Succeed Act’s initial requirements, applicants must have lived in the United States continuously since June 15, 2012—more than five years and eight months ago—and have entered before the age of 16. They need to have been younger than 31 in June 2012—36 years old today—and have graduated high school or be enrolled in college. According to estimates from the nonpartisan Migration Policy Institute (MPI), under the legalization portions of the S&S Act,* fewer than 1.6 million people could potentially become conditional permanent residents.

Even fewer will actually apply. DACA applicants had similar requirements when President Obama created the program in June 2012 (see Table 2), but only 60 percent of the eligible population ever signed up. While certainly S&S’s promise of permanent residence could entice some more applicants to apply, the factors that led to the low levels of DACA participation are likely to continue under S&S’s legalization program.

Some people who are included in MPI’s eligible population are not actually eligible, because they have committed certain criminal offenses (because these offenses can’t be modeled in the American Community Survey data MPI employs). This population of “eligible ineligibles” will grow significantly under S&S, because the legislation includes a variety of new criminal and non-criminal bars to a successful application, including the inability to support oneself without government benefits, prior deportations, removal orders, falsely claiming to be a U.S. citizen, false statements to obtain immigration benefits, and state or local offenses arising due to a lack of immigration status. No estimates exist of how many Dreamers fall into one of these categories, but it is potentially quite large.

In addition to the “eligible ineligibles,” some Dreamers believe they are ineligible but are actually eligible. This population could explain a major portion of the DACA enrollment-eligibility gap. The Secure and Succeed Act would likely increase the confusion, with its variety of new requirements on top of those from the original DACA program. DACA required enrollment in school of any kind or a high school degree. S&S would increase those initial requirements to require college enrollment or a high school degree.

S&S would create a new category of quasi-“eligible ineligibles”: those who are initially eligible, but could not meet the secondary requirement to extend status or receive permanent residence. The risk of a denial may keep some from taking the risk to apply. Nearly 8 percent of applicants for DACA were rejected. The S&S Act requires applicants to sign away their rights to an immigration hearing before a judge, meaning an agent could remove them quickly without due process for any infraction. If they dropped out of college or lost their job for more than a year, S&S could quickly end up as a pathway to deportation. This actually imposes a new risk that wasn’t present with the DACA program itself.

Applicants also consider the cost. DACA required an application fee of $495. This forces the recipients to have this amount on hand to pay to enter the program. Many DACA recipients cite the fee as a primary challenge. MPI’s analysis also cites family income as a factor “strongly affecting” Dreamers’ ability to apply. S&S would increase the fee by an unknown amount. But various requirements in the law would imply that the fee would increase as much as 100 percent or more. It requires a medical examination and could require an in-person interview—neither of which DACA required. This could make S&S legalization more like applying for adjustment of status to permanent residence, which costs about $1,225.

Fear of deportation counterintuitively affected DACA applications. The more immigrants in a certain community who feared deportation, the more likely they were to apply for DACA. This makes sense, because enforcement makes legal documents more valuable than they would otherwise be. Communities less affected by enforcement are more likely to fear putting themselves on the government’s radar for the first time than those where the government is already targeting them. For this reason, Asian immigrants signed up at the lowest rates, while Mexican immigrants—the most likely to be deported—signed up at the highest rates.

S&S’s impact on this phenomenon is likely mixed. On the one hand, Asian immigrants are more likely to say that green cards are more important than relief from deportation for unauthorized immigrants, making them more likely to apply. On the other hand, S&S doesn’t immediately provide a green card but rather a seven-year conditional residence status subject to a variety conditions. If this population was concerned about bringing attention to themselves under President Obama, there is little reason to believe that concern would decrease under President Trump, whose administration has demonstrated a willingness to deport even people who regularly checked in with immigration enforcement.

Moreover, many Dreamers expressed concern that their application could be used to target their families. Not only does S&S not address this fear, it amplifies it by providing enforcement resources and new legal authorities to the administration to speed up deportations.

The increased benefits of a potential green card may draw out some new applicants who previously didn’t want to take the risk to apply. But overall, the increased costs, greater risks, heighted eligibility requirements, and more frightening political environment would act to depress application rates. According to the Congressional Budget Office (CBO), the last major legalization—the 1986 amnesty—had only a two-thirds participation rate, despite much less stringent requirements than the ones contained in S&S. Ultimately, we chose to use the CBO’s higher rate of 67 percent, rounding it up to 70 percent—10 percentage points higher than DACA’s initial enrollment rate. Based on this analysis, we can conclude that at most 1.1 million Dreamers would receive initial legal status under the Senate GOP proposal.

Why Secure and Succeed Won’t Give a Pathway to Citizenship to 1.8 Million People

The 1.1 million people who are legalized by the Secure and Succeed Act receive (at first) only conditional permanent residence under the bill, not full permanent residence with a right to seek citizenship. For that, S&S recipients would have to reapply for an extension and separately for permanent residency. Under DACA, which had no additional requirements at all to extend status other than maintaining residence in the United States for another two years, just 86 percent of initial enrollees maintained status through the end of the program. Under S&S, applicants for extension and ultimately permanent residency would be required to pay a fee of at least (another) $1,225, have accumulated 7 years of residency, English language literacy, and 62 months of a mix of either employment, military service, or college enrollment.

Only 79 percent of initial enrollees would meet these requirements and move onto the permanent residence phase, according to MPI estimates. That means fewer than 900,000 Dreamers would receive permanent residence – the promised “pathway to citizenship” – under the White House-GOP Senate bill.

Finally, this population will only have the right—after yet another 7 years—to seek citizenship. The actual population that will receive it is much lower. Nearly 90 percent of DACA recipients are from Mexico, Guatemala, Honduras, and El Salvador. According to the Pew Research Center, these nationalities have naturalization rates below 50 percent. Mexicans, which account for 80 percent of all DACA enrollees, have a 42 percent naturalization rate. Given that Dreamers grew up in the United States, however, they are more likely to want citizenship. For this reason, Table 1 above applies the average naturalization rate for all countries of 67 percent. This implies that fewer than 600,000 people would end up receiving citizenship under the White House-Senate GOP proposal.


In the best case scenario, the Senate GOP plan would likely provide a pathway to citizenship to fewer than 900,000 Dreamers—less than half of the president’s promise. Moreover, only an estimated 587,657 would likely naturalize—less than a third of the 1.8 million that some senators have claimed.

If Congress wants to fulfill the president’s promise, it would need to institute a broader legalization program for Dreamers with as few risks and costs, and as little confusion, as possible. Congress would also need to provide legal certainty in some form for their parents to mitigate fear of coming forward. Members of Congress should also stop exaggerating the extent of the legalization of Dreamers as part of a strategy to justify politically questionable policy choices, including imposing large-scale reductions in the annual level of legal immigration and eliminating many current immigration categories.

Various Legalization and Citizenship Tracks for Young Immigrants

Sources: Senate Amendment 1959 to H.R. 2579; S. 1615USCISH.R. 1468

*Migration Policy Institute evaluated the Succeed Act, which contains the same basic criteria for legalization as the Secure and Succeed Act.

The evidence is in. And it’s great news for kids in charter schools. A just-released study by my colleagues at the University of Arkansas and me finds that, overall, public charter schools across eight major U.S. cities are 35 percent more cost-effective and produce a 53 percent higher return-on-investment (ROI) than residentially assigned government schools.

And every single one of the cities examined exhibited a charter school productivity advantage over their district school counterparts. As shown in Figure 1 below, charter schools outperformed district schools in each city on student achievement despite receiving significantly less resources per student. Charter schools in all eight cities studied are getting more bang for the buck. And in places like D.C. and Indianapolis, charter schools are doing more with a lot less.

Figure 1: Charter School Funding and Performance

Our ROI models consider the effect that each schooling sector has on children’s lifetime earnings relative to the total taxpayer investment for children’s K-12 education in each sector. As shown in Figure 2 below, charter schools provide a huge ROI for taxpayers. And D.C. charter schools are knocking it out of the park by producing an 85 percent higher ROI for their taxpayers than district schools.

Let’s make this a bit more concrete. The data show that every thousand dollars spent on education in D.C. district schools translates to around a $4,510 increase in students’ lifetime earnings. That is commendable. But that same thousand-dollar-expenditure produces an estimated $8,340 in students’ lifetime earnings if allocated to a public charter school in the city. And that 85 percent advantage is huge considering that taxpayers spend over $458,000 for each child’s K-12 education in D.C. district schools.

Figure 2: ROI for Charter Schools Relative to TPS (13 Years)

Notably, charter schools in Boston and Indianapolis both produced ROIs that were over 60 percent higher than their neighboring district schools. New York City, San Antonio, and Denver all produced ROIs that were 29 to 32 percent higher than district schools.

But these results shouldn’t surprise anyone. When educational institutions have the incentive to spend money wisely, they do just that. Because residentially assigned government schools do not have to attract their customers, they can spend tons of money on administration and fancy buildings. On the other hand, charter schools must spend money on kids – rather than administrators – if they want to keep their doors open.

From 1983 to 1999, the CBO issued two-year forecasts that added up to a 2.7% growth rate, which would now be widely dismissed as a “rosy” forecast. Yet actual growth averaged 3.7% from 1983 to 1999 – a full percentage point higher – despite a recession in 1991. Today, the CBO forecasts that even 2.7% economic growth is impossible, and claims only 1.9% is within reach. 

The Administration thinks the economy can grow a percentage point faster. The 2019 Budget estimates the economy will grow by 2.9% a year for ten years. The Committee for a Responsible Budget (CFRB) argues that this “strains credulity, especially if interest rates and inflation also remain under control, as the budget predicts they will.” [This appears to suggest higher inflation would be good for growth.]

“Given population aging and other economic fundamentals,” says the CFRB, “the United States is likely to ultimately achieve growth of 2 percent per year or perhaps less – not 3 percent. The Federal Reserve projects long-term sustained growth of 1.8 percent per year [1.7–2.2%], and the Blue Chip average for sustained growth is only slightly higher at 2.1 percent. Prior to the tax deal, CBO projected a long-run growth rate of 1.9 percent.”

Is the OMB unrealistic to estimate the economy can grow by 2.9% a year or is the CBO unrealistic to assume it can’t grow faster than 1.9%?

The real contest here is between current OMB projections after the tax deal, and CBO projections “prior to the tax deal.” Fed and Blue Chip forecasts are unofficial and unpersuasive.

The Federal Reserve does not make official long-term projections. The quarterly FOMC Summary of Economic Projections (SEP) defines  “longer-run projections” as rates of growth to which “a  policymaker expects the economy to converge over time – maybe in five or six years.” Last December 13, the 19 survey participants thought that after five or six years real GDP would settle down to 1.7% to 2.2%, but that is not a ten-year average since growth in the previous five or six years might be rapid.

Blue Chip Indicators collect monthly forecasts from over 50 business economists (I used to be one of them). It constructs a “consensus” by averaging a possibly wide array of different estimates. All such frequently-revised forecasts are most reliable as lagging indicators – becoming pessimistic after economic news turns bad and optimistic after things pick up. The models and techniques used to make monthly or quarterly forecasts have no predictive power at all beyond two quarters, if that.

What about the CBO? Their projection of 1.9% growth is a full percentage point below that of the Trump administration. Could the CBO possibly be that far off? Sure. They’ve done it before.

From 1983 to 2000, the CBO’s two-year forecasts of real GDP growth were exactly one percentage point too low, on average.  A two-year forecast for 1983–84 was made in 1983 and combined estimated growth rates for both years, so the fact that 2-year growth rates kept being underestimated repeatedly in all but two years (1990 and 1991) from 1983 to 2000 was a triumph of theory over experience.

The graph, from “The CBO’s Economic Forecasting Record,” shows the CBO systematically underestimated growth of real GDP after the Reagan tax rate reductions were phased in during 1983–84 and 1988–90 (TRA86), and again after the capital gains tax was slashed from 28% to 20% in 1997. Conversely, the CBO overestimated GDP growth after Bush 41 raised tax rates in 1990, after Obama raised tax rates in 2013, and during the high-tax bracket creep years of 1976–82.   The CBO appears to suffer from a pro-tax estimating bias – assuming higher tax rates do no harm, and lower tax rates do no good.

The CBO argues that its “five-year forecasts of output and inflation are more accurate than its two-year forecasts of those variables, in part because long-term forecasts rest more on underlying trends in the economy than on short-term cyclical movements, which are very difficult to predict.”  Unfortunately, CBO five-year forecasts also underestimated real GDP growth in all but one 5-year period between 1981-1985 and 1999-2003.  The exception was 1987-91, when the CBO estimate proved slightly optimistic (0.28%) thanks to recession on the heels of the ill-fated Bush 41 “tax increase.”  A single mild recession (aggravated by higher taxes) can’t explain why the CBO consistently underestimated the 4.4% rate of real GDP growth for seven years from 1983 to 1989, or the same 4.4% pace for four years from 1996 to 2000.

A recent paper, “How CBO Produces Its 10-Year Economic Forecast” explains that “CBO projects potential TFP on the basis of historical trends in TFP growth. However, projecting trends in TFP is particularly challenging because it is, by definition, a measure of unexplained growth in output.” On the contrary, extrapolating “historical trends” is just lazy, not “challenging.”

The so-called “economic fundamentals” the CFRB mentioned essentially consist of projecting recent trends into the future.  That means assuming output per hour (productivity) keeps growing at the anemic 1.3% pace of 2006 to 2015, and that hours worked grow at half that rate because labor force participation is assumed (“projected”) to remain extremely depressed and most part-timers are assumed to reject longer hours.

If you add 1.3% projected growth of productivity to 0.6% projected growth of the labor force, you end up with a 1.9% limit on potential economic growth (slower than 2010–2017 when the economy grew at a 2.2%).  But productivity and labor force participation depend on incentives, not past trends. And incentives to invest and work just changed dramatically.

We know that the CBO is perfectly capable of underestimating economic growth by a full percentage point over a 10-year period since it already managed to do that over a 17-year period. If history is any guide, the CBO’s 1.9% long-term forecast is once again much too low, as it has been whenever the highest tax rates on income and/or capital gains were reduced.

As a native Texan, I make an effort to stay current on the latest happenings in my home state. And even though the announcement by the Consumer Financial Protection Bureau that it will reconsider new federal rules that would regulate payday lending is national in scope, the nature of the affected industry means that the particular impact will inevitably vary from state to state. Accordingly, the recently published editorial by the San Antonio Express-News addressing the topic calls for a state-specific response.

The Editorial Board must have viewed its argument as a common sense, self-evident proposal: in order to cure the payday malady, we need more laws! But the argument that “Texas lawmakers need to step up their game next session” in the event these federal regulations are rescinded gets it exactly backward; what Texas needs is not more fix-one-problem-while-causing-two-more statutes. Instead, an epinephrine injection of vigorously enforcing good laws should be combined with the surgical removal of bad ones.

Texas has gone down the “just pass another law and fix it” road before on this issue, and this approach has consistently made things worse, not better. After the passage of the federal Fair Credit Reporting Act in 1970, an industry offering “debt repair” services emerged. Unfortunately, many debt repair organizations engaged in disreputable practices and, in order to combat the excesses of this industry, the Texas Credit Services Organizations Act was enacted in 1987. But the organizations (“CSO’s”) created and defined under this Act not only included businesses paid to improve a consumer’s credit rating, but also those involved in “obtaining an extension of consumer credit for a consumer.” After the FDIC issued new guidelines on payday lending in 2005, Texas payday lenders sought to avoid these and other restrictions by registering and operating as CSO’s. And now, in an effort to fix the problem caused by the CSO statute, which itself was designed to fix a supposed problem in the Fair Credit Reporting Act, we are told that yet another statute must be passed. Who is actually gullible enough to think that this new “fix” will not again create at least as many new problems as it supposedly solves? It’s deja vu all over again.

No legislative body, no matter how powerful or well-intentioned, can repeal the laws of economics. In 2008, congressional mandates for Freddie and Fannie combined with the passage of the Community Reinvestment Act to mandate lending to those who could not afford to pay the loans back, thereby injecting systemic risk into the market. Similarly, the unintended consequences of severely restricting or eliminating the ability of desperate people facing financial emergencies to take out payday loans will only drive the market underground, resulting in less competition and more harm to consumers.

The real problem is not the existence of payday loans per se, but rather the unseemly entanglement of government enforcers with payday lenders. When borrowers default on credit cards or fail to pay back a signature loan from their bank, they face a denial of future credit from that institution, negative credit reporting making it more difficult to obtain credit with other institutions, and execution on civil judgments that can be satisfied against their nonexempt assets. These consequences work well to both constrain irresponsible behavior by consumers and allow institutions to properly assess the risk of lending. But the payday lending industry commonly eschews such reasonable remedial measures in favor of employing state actors to do their dirty work.

The process goes something like this. A payday lender requires the borrower to provide a post-dated check in order to receive the loan. Unsurprisingly, on the appointed date these checks often bounce due to insufficient funds. Lenders then take advantage of unsophisticated borrowers by threatening prosecution for check fraud unless they either pay up or roll over the loan. If these threats don’t do the trick, the lenders then refer the matter to the local district attorney’s office for potential prosecution.

These threats from collectors are not legally supportable under any fair interpretation of the penal code, and thus should constitute a violation of the Texas Debt Collection Act’s provisions against falsely accusing consumers of crimes or threatening them with arrest. Unfortunately, not only are such collection actions rarely punished, but many district attorney’s offices are often all too willing to countenance such charges. In fact, some district attorneys not only send out legally required notices on behalf of merchants using official government letterhead, but they have also established fast-filing programs that allow these lenders to expedite the process.

Taking a ding on your credit report is one thing; facing jail time is quite another. It is true that these pseudo-crimes are rarely prosecuted (presumably, because many recipients are suitably terrified into immediate payment), and that claims of modern-day debtors’ prisons lurking just around the corner are a bit hyperbolic. Even so, hijacking the government—the entity that by definition has a monopoly on the legitimate use of force—by transforming prosecutors into private debt collecting muscle is simply unconscionable.

Rather than pass another statute, the Texas Legislature should start by repealing the wrongheaded provisions of the CSO that allow payday lenders to avoid the laws intended to regulate their industry. The Consumer Protection Division of the Texas Attorney General’s Office should more vigorously enforce provisions of the Texas Debt Collection Act prohibiting fraudulent collection practices. And prosecutors should cease threatening to break borrower’s financial legs unless they pay up. These are the sort of solutions needed to combat the most pernicious aspects of the industry.

Payday lending, as currently constituted, is indeed a boil on the skin of the financial system. But the “medicine” of passing a new state statute that significantly limits these loans will not only fail to cure the patient, it will both exacerbate the current illness and produce a whole litany of unwanted side effects. The Texas Legislature should observe the Hippocratic Oath instead; first, do no harm.

Today millions of Americans will celebrate Valentine’s Day by purchasing roses for their loved ones and, in so doing, will participate in one of the everyday miracles of capitalism which too often escape our notice. As a recent Washington Post article points out, these roses will most likely have been grown thousands of miles away in Colombia, flown to the United States aboard cargo jets, and then delivered to florists and other retailers at the cost of a mere $1.50 per stem. That this is possible is not only a tribute to the magical powers of capitalism, but—as the newspaper notes—free trade and a 2012 agreement between the United States and Colombia which permanently lifted U.S. import tariffs on Colombian flowers. Indeed, a close reading of the piece reveals some of the many advantages of free trade and the benefits it confers.

Among them: 

Imports save consumers money: One of the most straightforward benefits of free trade is the reduced cost to consumers for the goods they purchase as a result of reduced tariffs and the ability of businesses to develop more cost-effective supply chains. By importing flowers from Colombia, according to the article, the price of roses has been kept almost unchanged for decades with a dozen red roses often available this week for less than $20.

Imports create jobs: Unsurprisingly, the growth of Colombia’s flower industry has helped provide jobs in that country—to the tune of 130,000 according to The Washington Post (including, it seems, for thousands of Venezuelans fleeing from that country’s experiment in socialism). Often overlooked, however, is that imports also create new employment opportunities for Americans, both through the actual process of importation and as an intermediate good. The numerous planes full of flowers, for example, require logistics personnel to offload, store, and transport them to their final destination. Cheaper flowers, meanwhile, mean increased sales and more workers at the distributors and retailers which carry them. Indeed, the article cites the example of the USA Bouquet Company in Doral, Florida which employs 75 workers to put “imported red roses into vases and then carefully [pack] them in boxes for a Valentine’s shipment to Walgreens.”

Gains from specialization and comparative advantage: Free trade between countries allows for specialization and an increased focus on those areas in which each country enjoys a comparative advantage. In this example, Colombia enjoys a comparative advantage in the growing of roses and other flowers which, as the article points out, is a major factor behind the decline of U.S. rose production. Unmentioned, however, is that increased trade between the United States and Colombia—up 95 percent since 2006—has led to billions of dollars in exports in sectors where Americans enjoy a comparative advantage, such as agriculture, machinery, and computer software. In addition, competition from Colombian growers has forced American businesses to adapt and move higher up the value chain to areas where they might have a comparative advantage within the flower sector. As a result, the article notes that the price of U.S.-grown roses has “ticked up in recent years because U.S. growers have focused primarily on higher-end roses that are designed for weddings and special events.”

If heartache is to be found in this free trade valentine it is that the trade agreement signed with Colombia is the last to have been approved by Congress (along with free trade agreements with Panama and South Korea, all of which were passed on October 12, 2011). While President Trump has promised the conclusion of additional bilateral agreements, new negotiations have yet to be initiated. That’s unfortunate, and we should hope that 2018 will see a new push on this front. Colombian roses are but one beautiful example of the gains to be had from tariff-free access to the world’s offerings, and Americans deserve access to all of them.

Today, President Trump hosted several members of Congress to discuss his possible plans for imposing new restrictions on steel and aluminum imports under “Section 232” of US trade law. Amidst that discussion comes this nugget, via Politico Pro[$] (emphasis mine):

The president was equally dismissive when Sen. Lamar Alexander (R-Tenn.) brought up the negative consequences of former President George W. Bush’s decision to restrict steel imports in 2002.

“The effect was it raised the price of almost all steel in the United States,” leading to job losses in the auto-parts sector and among other manufacturers, Alexander said. “There are ten times as many people in the steel-using industry than steel-making.”

Trump shrugged off the complaint. “It didn’t work for Bush, but it worked for others,” he said. It was not clear, based on a pool report of the closed-door conversation, whether he explained that point.

I guess it was good of the President to acknowledge the widely-known costs that the Bush Administration’s 2002 steel safeguards imposed on American consumers and companies.  However, the President is sorely mistaken to assume that other instances of US steel protectionism turned out much better.  Indeed, as I wrote in my 2017 paper on the historical failures of American protectionism, the US steel industry has for decades gone to the government trough for new restrictions on its foreign competition, and the results these import measures are always the same: immense consumer costs and very few, if any benefits to the industry and its workers.

For example, multiple academic studies in the 1980s showed that efforts to restrict imports of various steel products annually cost American consumers between $200,000 and $2.3 million (2017 Dollars) for every US steel industry job protected:

These and other steel imports restrictions also didn’t fix the US steel industry or save its workers (emphasis mine):

A 1986 report from the CBO considered the effects of trade protection in revitalizing domestic firms in four cases— textiles and apparel, steel, footwear, and automobiles—between the 1950s and 1970s. The authors found that, indeed, U.S. trade barriers limited imports and increased protected firms’ output, employment, and profits above what they would have been without protection. However, “[i]n none of the cases studied was protection sufficient to revitalize the affected industry.” The steel, footwear, and textile and apparel industries repeatedly sought new import protection after their previous protection expired, and the automobile industry succumbed to imports of the small cars that were the source of their “competitive difficulties.” Furthermore, the protection did not significantly increase the companies’ incentive to invest in cost-saving technologies that would improve their long-term competitiveness— even when the protected companies had the capital available to make such investments. The authors concluded that, because the system of trade restraints was unable to save protected industries, the U.S. government should consider policies other than import protection—such as encouraging investment in less labor intensive industries, aiding displaced workers, or ending “special treatment for trade-impacted industries”—when crafting new U.S. trade policies.

In 1986, Robert Lawrence and Paula DeMasi examined the effect of escape clause relief (tariffs, quotas, or “orderly marketing arrangements”) on 16 U.S. manufacturing industries [including steel], representing nearly all industries that secured such protection between 1950 and 1983. The results were damning: only 1 of the 16 industries—again, the bicycle industry—expanded after the protection lapsed, 11 contracted, and the remaining 4 were inconclusive at the time. Furthermore, 8 of the 12 industries whose “temporary” escape clause relief had expired actually went back to the government for more protection.

A 1987 study from Robert Crandall on “The Effects of U.S. Trade Protection for Autos and Steel” found that this protection limited imports but actually harmed the industries’ long-term position. For example, it discouraged improvements in product quality and encouraged cannibalistic overinvestment and too-rich labor contracts that “simply postpone[d] part of the necessary adjustment to the loss of competitiveness.” As a result, Crandall concluded, “The experience with the auto and steel industries raises serious questions about the effectiveness of quotas as a means to revitalize an industry.”

The results of American steel protectionism in the 1990s and 2000s were no different: more American consumer pain and no industrial revival (emphasis, again, mine).

No U.S. industry has benefited more from protection than the steel industry. In the 1990s and early 2000s, for example, approximately 150 steel antidumping orders were in place, covering almost 80 percent of all steel imports during the period. … As with other bouts of protectionism, the costs of these actions were massive, while gains were minimal. Estimates of the economic costs of U.S. trade remedy protection against steel imports range from approximately $60 million per year to well over $2.7 billion—or $450,000 per job saved in 2001 ($596,000 in 2017 dollars). And given that U.S. steel-consuming industries employ between 40 and 60 workers for every 1 steelworker, every job allegedly saved in the steel industry was far outnumbered by job losses in steel-using industries….

[D]espite decades of protection and billions of dollars in government subsidies, Barfield noted in 2003 that “the U.S. steel industry has dramatically shrunk, with employment down by two-thirds … capitalization only one tenth its former valuation,” and 12 different bankruptcies between 1998 and 2000 alone. These exact same dynamics continue today: As of October 2016, 90 191 of 373 antidumping and countervailing duty orders in place were on iron and steel products. Yet the industry continued to clamor for more import protection and government assistance. Both points testify to the duties’ failures (and the steel industry’s political connections).

The steel industry, of course, is not alone.  In fact, as shown in my paper and the many academic studies cited therein, the poor results above are indicative of the vast majority of cases of American protectionism.  It just doesn’t work.

Maybe it’s time the President and his advisers stop “shrugging off” this simple fact.

The views expressed herein are those of Scott Lincicome alone and do not necessarily reflect the views of his employers.

The Trump Administration FY 2019 budget was released yesterday. Among other reductions to spending, the Department of Housing and Urban Development (HUD) would be cut by more than 14 percent if Congress implements the administration’s recommendations.

That’s a very big “if” indeed. Congressional Republicans are not in the mood to make difficult budget cuts these days. Still, howls of anger have gone up across Washington, D.C. at the mere suggestion of a cut to HUD. Advocates maintain that HUD needs more money, certainly not less.

That’s because many housing activists misunderstand housing policy. In reality, “subsidize-your-way-to-affordability” has driven HUD’s approach to housing for decades and hasn’t meaningfully improved the housing affordability landscape. In fact, by HUD’s own measure the proportion of households spending more than 30 percent of household income (also called cost-burdened households) is rising over time.

Figure 1: U.S. share of cost-burdened households is not improving


Worse yet, HUD’s subsidies have arguably actively undermined housing affordability and will continue to do so. That’s partly because subsidies don’t address root causes of the affordable housing shortage.

Zoning regulations contribute to housing shortages in major cities, and this drives up housing costs by an estimated 30-50 percent in some locales. But HUD rewards cities that have more restrictive zoning and land use regulations with greater housing subsidies. For example, HUD provides around 2x the housing subsidy dollars to the most restrictively zoned states as compared with the least restrictively zoned states.

Figure 2: Federal housing affordability spending is highest in the most-regulated states

Source: “2017 Budget State-By-State Tables,” Office of Management and Budget, 2017. This pairs the author’s 2015 state rank with 2015 budget numbers. Dollar values include Section 8 housing vouchers, Public Housing Operating Fund, and Public Housing Capital Fund spending.

The Trump administration’s proposed cuts would reduce existing counterproductive incentives for states. As it stands, housing subsidies act as a convenient political distraction for local politicians. With reduced HUD subsidies, states and local municipalities will be forced to confront the natural consequences of restrictive local regulations. 

In an op-ed published yesterday at The Hill, Alex Nowrasteh wrote about why a bill working its way through the U.S. House of Representatives may be one of the most anti-immigration bills in decades. With both chambers of Congress now looking to make good on promises to provide a legal framework for the roughly 700,000 DREAMers impacted by President Trump’s decision last year to rescind DACA, many House Republicans are pledging to support the Securing America’s Future (“SAF”) Act. 

Democrats may not support the SAF Act, but Nowrasteh notes that it will likely represent a line in the sand for many Republicans as negotiations proceed. The problem? “As a so-called DACA fix,” he writes, “the SAF Act barely measures up.”

It would provide DREAMers with temporary and renewable residency permits—in other words, short-term reprieves. And in return, DREAMers would face a new set of restrictions, including the requirement that they maintain an income 125 percent higher than the poverty line.

The SAF Act also makes major cuts to legal immigration categories:

[It] inexplicably cuts legal immigration, reducing the number of immigrants by as much as half after 10 years. Among the categories cut are the diversity green card, which is completely eliminated, as well as most family-sponsored immigrants. Asylum seekers will also get a significant chop under the bill.

Under the new SAF Act status quo, immigration would allow fewer skill-based immigrants, due to the move away from the green card system’s growing tendency to select educated workers. It also means that immigrants might risk separation from their family—the SAF Act would make it almost impossible for green card recipients to sponsor their spouse or children if their marriage or the child’s birth occur after the green card is conferred.

And as a cherry on top, the SAF Act allocates to border security approximately $124 billion over five years. This is dozens of times more money than Border Patrol spent last year, and at a time when illegal crossings at the border are at a nadir.

There’s no way to sugarcoat the SAF Act as any kind of concession or compromise. It is give and take, with an emphasis on take. 

As Nowrasteh concludes, there are better ways for Republicans in Congress to do something about the plight of DREAMers. Lawmakers should propose DACA fixes that don’t drastically reduce the number of legal immigrants.

You can read the full piece here.

President Trump’s new budget for 2019 proposes privatizing federal assets such as airports, air traffic control, and electricity facilities.

The budget’s infrastructure section suggests that Congress “Authorize Federal Divestiture of Assets that Would Be Better Managed by State, Local, or Private Entities.”

It continues:

The Federal Government owns and operates certain infrastructure that would be more appropriately owned by State, local, or private entities.

For example, the vast majority of the Nation’s electricity needs are met through for-profit investor-owned utilities. Federal ownership of these assets can result in sub-optimal investment decisions and create risk for taxpayers.

Providing Federal agencies authority to divest of Federal assets where the agencies can demonstrate an increase in value from the sale would optimize the taxpayer value for Federal assets. To utilize this authority, an agency would delineate how proceeds would be spent and identify appropriate conditions under which sales would be made. An agency also would conduct a study or analysis to show the increase in value from divestiture. Examples of assets for potential divestiture include—

  • Southwestern Power Administration’s transmission assets;
  • Western Area Power Administration’s transmission assets;
  • Ronald Reagan Washington National and Dulles International Airports;
  • George Washington and Baltimore Washington Parkways;
  • Tennessee Valley Authority transmission assets;
  • Bonneville Power Administration’s transmission assets; and
  • Washington Aqueduct.

These reform ideas are straight out of Cato’s playbook. The administration also renews its call to privatize the air traffic control system. That is, “shift the air traffic control function of the Federal Aviation Administration to a non-governmental, independent air traffic services cooperative.”

These studies provide background to the proposed reforms:

The budget also proposes reforms to allow for the “disposition of federal real property,” and you can read about that topic in this detailed study of federal privatization.

So regarding all those “sub-optimal investment decisions,” I say enough! We should move assets into the private sector and unleash entrepreneurs on America’s transportation challenges.

The White House released President Trump’s infrastructure plan today, which calls for spending $200 billion federal dollars as seed money to stimulate a total of $1.5 trillion on “gleaming new infrastructure.” Almost lost in the dozens of pages of documents issued by the administration is that the reason why the federal government supposedly needs a new infrastructure program is that our existing infrastructure is crumbling, and the reason it is crumbling is that politicians would rather spend money on gleaming new projects than on maintaining the old ones.

The White House proposes several new funding programs. The administration could have dedicated one or more of these programs to maintenance and repair of worn-out infrastructure. Instead, all $200 billion can be spent on new projects, and knowing politicians, most of it will be. To make matters worse, funds for most of the programs would be distributed in the form of competitive grants, but experience has proven that competitive grants are highly politicized. 

“In the past, the Federal Government politically allocated funds for projects, leading to waste, mismanagement, and misplaced priorities,” agrees White House economic advisor Gary Cohn. The administration’s solution, Cohn continues, is to “stimulate State, local, and private investment.” In other words, instead of most decisions being made by Washington politicians, they will be made by local politicians. But if local politicians were any better at maintaining infrastructure, then we wouldn’t have tens of thousands of local bridges classed as “structurally deficient” and the New York, Washington, Boston, and other subway systems wouldn’t be falling apart.

The White House says that the federal funds it proposes to allocate to infrastructure may be spent on either new construction or maintenance, which is an advantage over some existing federal programs that can only be spent on new construction. But just because they can be spent on maintenance, doesn’t mean they will be.

The New York subway system is falling apart because the city doesn’t have enough money to maintain it. Yet it has enough money to spend $10 billion on a tunnel between Penn Station and Grand Central Terminal for Long Island Railroad trains, which the New York Times has called “the most expensive subway in the world.” It also has enough money to build the eight-mile Second Avenue subway, which at $2.1 billion a mile must be the second-most expensive subway in the world.

The Washington Metro system is falling apart because the region is short $10 billion to maintain it. Yet Virginia was able to find enough money to build the Silver Line and Maryland to build the Purple Line; the costs of these two projects would have been enough to fix most of the existing Metro system.

The Boston rail transit network is also falling apart because the region can’t find the $470 million a year needed to maintain it. Yet is was able to find $2.3 billion to build a 4.3-mile light-rail extension. In all of these cases, local politicians decided to spend money on new projects even though 50 to 80 percent of the money in each case could have been spent on maintenance.

The truth is that our infrastructure isn’t in as bad shape as some claim. State highways are in good condition; local roads less so. Rail freight systems are in good condition; rail transit systems less so. In general, infrastructure paid for out of user fees are in good condition; infrastructure paid for out of tax dollars less so. 

This means there is a simple way to take the politics out of infrastructure and make sure that managers maintain it: fund it out of user fees, not tax dollars. Managers of user-fee-funded infrastructure tend to do the best job of maintenance because they know users will stop paying to use their facility if it becomes unreliable.

To its credit, the White House program does make some token movements in the direction of more user fees. For example, it would allow states to charge tolls for all interstate highway, something that is now, for the most part, prescribed by Congressional edict. But even this depends on state politicians asking their constituents to pay tolls for something they have been getting for “free,” which is unlikely to happen.

In the end, however, the major impact of the Trump plan is to drop $200 billion new dollars onto state and local governments on top of existing federal spending programs. This will merely provide more insulation for state and local politicians from having to ask users to pay for the infrastructure they build. The result will be that most, if not all, of that $200 billion will go to build new infrastructure that we probably don’t need and can’t afford to maintain rather than maintaining what we have.


Quickly reading through the overviews of President Trump’s proposed FY 2019 budget, the good news is that funding coming through the U.S. Department of Education would be cut. The bad news is that the budget would potentially include up to $1 billion applicable to private school choice, which would threaten centralized regulation of choice, rendering such choice far less meaningful. Think Common Core for all!

Overall it appears spending by the U.S. Department of Education would decrease by around $3.6 billion—or about 5.4 percent—from 2017, based on quick calculations using the Department’s budget summary and data in an addendum that alters the summary due to the budget legislation enacted last week. As I’ve noted before, eliminating such ineffective—and unconstitutional—undertakings as the $1.2 billion 21st Century Community Learning Centers would be solid policy, and frankly the evidence is compelling that the overall K-12 and higher education federal endeavor has been an expensive mistake.

The bad news is that overall cuts would not be greater, while the budget would create a new, $1 billion Opportunity Grants program that would include money for private school voucher programs. The program would be broken into two pieces—Scholarships for Private Schools and Open Enrollment Grants—with only the former open to private choice programs. No specific funding split between the sub-programs is identified in any of the budget materials I’ve seen, so it is unknown how much of the funding would go to private choice. But even a small amount of money relative to overall education spending can be a powerful lever to get states and schools to open themselves to regulation—it just needs to look like a lot in news stories or ledgers—and that is the huge danger of federal school choice. Of course, the Constitution no more authorizes federal choice programs than it does other education undertakings.

The budget is likely dead on arrival, and there are certainly things I missed in a quick once-over. But at the very least it reveals an administration that has sort of the right inclination on education—shrink the federal footprint—but that will curb that inclination when it comes to school choice.

The Trump administration has released its federal budget for 2019. The document lays out various reform proposals and provides projections of revenues and spending through 2028. The projections are a little outdated already given the budget-busting spending deal reached last week, but it is still interesting to take a look.

The chart compares Trump’s proposed revenues and spending to the most recent CBO baseline projections from last June (which run through 2027).

Despite the large tax cut in December, the Trump administration is projecting federal revenues to be higher in years after 2023 than did the CBO before the cuts were enacted. The administration is assuming strong economic growth of about 3 percent in coming years, partly based on the benefits of tax reform and its deregulatory initiatives.

On spending, the administration projection assumes that it can move various cut proposals through Congress, generating large savings down the road. But with a spendthrift attitude currently prevailing in Congress, such reforms seem optimistic. Note that the CBO projections do not incorporate last week’s discretionary spending increases, whereas Trump’s figures include large defense increases and large nondefense cuts in coming years.

As I was preparing for a Demand Progress-sponsored panel on Congressional oversight of intelligence matters on the afternoon of February 9, Demand Progress Policy Director Daniel Schuman and I agreed that if President Trump was going to refuse to “declassify” the House Intelligence Committee Democrats rebuttal to the “Nunes Memo,” he would wait until the late Friday news cycle to do it. We didn’t have to wait long for that prediction to come true

In a moment, I’ll get to the issue of whether Trump actually has the authority under the Constitution to do what he did, but I want to start with is this paragraph from the New York Times story referenced above:

But Donald F. McGahn II, the president’s lawyer, said in a letter to the committee on Friday night that the Democratic memo could not be released because it “contains numerous properly classified and especially sensitive passages.” He said the president would again consider making the memo public if the committee, which had approved its release on Monday, revised it to “mitigate the risks.”

In that same NYT story, House Intelligence Committee ranking member Adam Schiff provided further context:

In a statement on Friday night, Mr. Schiff said that Democrats had provided their memo to the F.B.I. and the Justice Department for vetting before it was approved for release by the committee. The Democratic memo was drawn from the same underlying documents as the Republican one.

“We will be reviewing the recommended redactions from D.O.J. and F.B.I., which these agencies shared with the White House,” Mr. Schiff said, “and look forward to conferring with the agencies to determine how we can properly inform the American people about the misleading attack on law enforcement by the G.O.P. and address any concerns over sources and methods.”

So if Schiff is to be believed, House Intelligence Committee Democrats ran their memo by Justice Department and FBI officials prior to the unanimous committee vote to release his memo, then sent the memo over to the White House for reaction. Trump and his team then demanded still more redactions. If the above account is correct, the same Justice Department or FBI officials who reviewed the original “Schiff Memo” apparently demanded still more redactions once it got to Trump’s desk.

It’s this sequence of events which brings me to the question of whether Trump has the authority under the Constitution to censor or rewrite Congressional work product, with or without Congressional assent, if it contains references to Executive branch information asserted as being classified, in part or in whole. The short answer is no. The longer answer is still no, but with some caveats.

Congress and Secrecy

The word “secrecy” appears only once in the Constitution, specifically in Article I, Section 5, which contains the following clause:

Each House shall keep a Journal of its Proceedings, and from time to time publish the same, excepting such Parts as may in their Judgment require Secrecy; and the Yeas and Nays of the Members of either House on any question shall, at the Desire of one fifth of those Present, be entered on the Journal.

From the beginning of the Continental Congress in 1774 through the adoption of the Constitution in 1789, the Congress was responsible for keeping such matters of state secret as they deemed necessary. It was not until 1818 that Congress passed a resolution directing President Monroe to publish previously secret material from the nation’s earliest history.

The ability of the Executive branch to keep military and foreign policy secrets was not something Congress would get involved in legislating until well into the 20th century. Examples include the Atomic Energy Act, the Freedom of Information Act, and the Classified Information Procedures Act

It’s worth noting that in none of these statutes did Congress renounce its authority to make materials covered by these laws public if it chose to do so—including material deemed classified by the Executive branch. The first real confrontation over this principle occurred in the aftermath of Congressional investigations of Executive branch domestic spying scandals that first surfaced in 1971.

The Pike and Church Committees

In 1975, the House and Senate each created Select Committees to investigate domestic spying and political repression operations carried out by the NSA, FBI, CIA, and military intelligence elements. The two committees became known by the names of their respective chairmen: Frank Church (D-ID) in the Senate and Otis Pike (D-NY) in the House. Both committees encountered deliberate efforts by Ford administration officials to block access to relevant agency or department records, resulting in months of often heated confrontations with CIA, NSA, FBI, and White House officials over committee demands for documents. 

As I noted in a recent piece in The Hill, the slightly differing approaches of the two committees led to very different outcomes:

In a now-infamous incident known as the “September compromise,” Pike agreed to allow the Ford administration to make the call about what executive branch documents could or could not be made public. When Pike moved to finalize his committee’s report and make it public in January 1976, Ford persuaded the House to block publication on the grounds that the entire Pike Committee report was a classified document. In contrast, the Senate refused to submit the Church Committee report for Ford’s review and published its findings in April 1976. Constitutionally, Church and his colleagues made the right call, Pike’s House colleagues the wrong one.

Indeed, in its preamble to its final report, the Church Committee made its position abundantly clear (from the committee’s final report, Vol. 1, p. 13):


It should be noted that the Ford administration made no attempt to challenge the Church Committee’s publication of the report in federal court—a de facto admission that the Congress did indeed have the authority to release—and thus simultaneously declassify—information previously deemed secret by Ford and his predecessors. 

The Caveats

The Pike Committee experience proved that the Executive branch could, under the right circumstances, still thwart an official Congressional release of information President Ford and his agency heads considered classified. However, he needed help to do it—in this case, Pike’s House colleagues, who failed to authorize the release of the Pike Committee report.

The subsequent legislation passed by the House and Senate creating standing Select Intelligence Committee’s largely adopted the system that had been used by the House in the Pike Committee episode. Molly Reynolds of Brookings recently wrote a good overview of this process at Lawfare

What the “War of the Memos” reveals is that the original Church Committee approach to this kind of Executive-Legislative confrontation was the right one. The process now being used—and more than likely abused—by the House Intelligence Committee GOP majority and the Trump White House makes it far less likely that the public will learn the truth about whether or not the FBI and Justice Department misled the FISA Court in the Carter Page episode. The larger and more long-term ramifications of this episode are that a new and dangerous precedent is being set by not just the House Intelligence Committee, but the Congress as a whole.

Letting any President dictate what Congress can or cannot publish is a clear assault on the Separation of Powers. It’s also possible for Congress to voluntarily breach the Separation of Powers and compromise its oversight of intelligence matters. That was the error committed by Senator Diane Feinstein (D-CA) when she voluntarily submitted the Senate Intelligence Committee’s full report on the Bush 43-era CIA torture program to President Obama for a “declassification” review. Instead, Obama—a fellow Democrat—sat on the full committee report. To date, Feinstein has failed to get her Senate colleagues to hold a vote to make the full report public.

Trump’s quashing (at least for now) of the “Schiff Memo” only underscores why the House and Senate must modify their respective chamber rules to make it clear, as the Church Committee did, that while the House and Senate Intelligence Committees will consider Executive branch concerns and arguments against making something public (and thus declassifying it), the final call will remain with the committees. 


The Trump administration has rolled out its 2019 federal budget, which includes a plan to boost investment in the nation’s infrastructure.

I see five components to President Trump’s infrastructure policies so far.

First, Trump approved business tax cuts in December, which will boost private capital investment in pipelines, broadband, factories, and much else. The private sector owns more than three-quarters of U.S. infrastructure, so the tax cut will create wide-ranging benefits.

Second, the administration is taking steps to reduce costly regulations and speed permitting on infrastructure projects. The new budget proposes to “shorten the process for approving projects to 2 years or less.” The number of federal rules creating roadblocks to construction has increased over time. Kudos to Trump for removing some of the barriers.

Third, the budget proposes spending an additional $200 billion over 10 years of federal dollars on infrastructure, as summarized here. There is a new “Incentives” program, a “Transformative Projects” program, and a “Rural Infrastructure” program.

These new programs are unaffordable, especially given that last week’s budget bill exploded annual deficits to more than $1 trillion. Creating a new $50 billion program for rural areas is particularly ridiculous given that the government already has a range of wasteful rural subsidies.

In general, new federal subsidies for infrastructure are not needed. Any state wanting to improve its infrastructure can do so with its own funding. States can raise funds through taxes, debt, user charges, public-private partnerships (PPPs), and privatization. Federal infrastructure subsidies are often counterproductive, as I discuss here.

Giving Trump’s new subsides fancy titles such as “Transformative Projects” program” will not make them more efficient than current bureaucratic federal efforts. All federal spending for infrastructure comes from taxpayers who live in the 50 states, and there is no “transformative” magic that happens when cash is sent through Washington.  

Fourth, the Trump plan would boost incentives for the states to pursue PPP deals, including expanding various loan programs and broadening eligibility for private activity bonds. Here the end goal is a good one, but a better way to encourage PPPs and privatization would be to end federal subsidies altogether and repeal the tax exemption for municipal bond interest. That exemption dissuades the states from privatizing facilities such as airports.

Fifth, the Trump plan would make reforms to federal control over lands and structures. It would create a “capital revolving fund” to allow agencies to buy real property that they currently lease. That may save money, but the administration should instead focus on reducing the government’s size and its need for office space in the first place. (The government currently owns or leases 275,000 buildings). The Trump plan would also allow agencies to generate more revenues from public lands, and allow “for the sale or lease of federally owned assets.” That sounds positive.

In sum, Trump’s infrastructure plan includes both big government and small government policies. The plan states, “President Trump’s proposal will return decision-making authority to State and local governments, which know the needs of their communities.” But the best way to do that would be to end federal subsidies and all the related string-pulling from Washington.