Feed aggregator

“… equally efficacious, and equally a hoax.” – Benjamin Disraeli, 1848[1]


One of the highlights of the U.S. summer for Fed watchers is the annual ritual in which the Fed’s economic soothsayers peer into their crystal balls, a.k.a. their stress tests, to reassure us that the U.S. banking system is robust and getting stronger all the time.

You see, while the future is uncertain, the results of the stress tests are not. Praise be that the news is always good and getting better.

This year, the news is particularly good. As usual, the key capital metrics across the system are better than ever. And whereas in previous years there were always dunces who failed, the latest set of stress tests are the first in which all the banks passed and this year’s class laggard, Capital One, got only the mildest of slaps on the wrist.

As James Ferguson of The MacroStrategy Partnership notes in a recent commentary on the latest stress tests:

… everywhere you look, the Fed now seems to be bending the rules  in the banks’ favour. … This [stress test] appears to be a test that has been designed to be passed.”[2]

In fact, the Fed is so pleased with the performance of its stress-test examinees that it decided to reward them (or, more precisely, their shareholders) with a big dividend/buyback party that will give them a big windfall.[3] The Fed provides the punchbowl which will be paid for by other bank stakeholders including taxpayers — yes, the same taxpayers who are still being compelled to subsidize the banks (via Too Big to Fail, deposit insurance, and such like) to take excessive risks and overleverage themselves, and who stand to pay the bill if there is another crisis and the banks get bailed out again.[4]

It is curious that these capital distributions are being welcomed by many of the same people who have argued vociferously against higher capital requirements. Advocates of low capital requirements say that high capital requirements would limit banks’ lending capacity, but they fail to note that this is what dividend payouts do too.

Nor is there any sign that the Fed is inclined to take away the punchbowl any time soon. Former Fed chairman William McChesney Martin must be turning in his grave.

Indeed, plans are afoot to make future stress tests even less demanding: the days when banks felt challenged by the Fed’s stress tests are well and truly over.

So this year’s stress tests are great news for bank shareholders, but bad news for everyone else.

Headline results

 The results for the Fed’s stress tests were published in two stages. The results of the first stage – formally known as the Dodd-Frank Act Stress Tests (DFAST) 2017  — were published on June 22. These tests give results showing how the banks involved would fare under an “adverse scenario” and a “severely adverse scenario” projected over the period 2017:Q1 to 2019:Q1. The 34 bank holding companies (BHCs) tested — generally those with $50 billion or more in total consolidated assets — represent more than 75 percent of the assets of all domestic BHCs.

These scenarios project the steep decline in real U.S. GDP shown in Figure 3 of the report.

The “severe adverse scenario” suggests a shock (in real GDP terms) that is more severe than the Global Financial Crisis (GFC). It also posits the U.S. unemployment rate rising by approximately 5.25 percentage points to 10 percent, accompanied by heightened stress in corporate loan markets and commercial real estate. Equity prices fall by 50 percent through the end of 2017, accompanied by a surge in equity market volatility, which approaches the levels attained in 2008. House prices and commercial real estate prices also experience large declines, with house prices and commercial real estate prices falling by 25 percent and 35 percent, respectively, through the first quarter of 2019. Total projected losses across the 34 banks come out to $493 billion.

The results for the second stage — the Comprehensive Capital Analysis and Review (CCAR) 2017  — were published on June 28. The CCAR provides a quantitative assessment of a firm’s capital adequacy and planned capital distributions (or “capital plan”), such as any dividend payments and stock buybacks, and provides a qualitative assessment of the bigger and/or more complex firms. The principal difference between the DFAST and CCAR stress tests is that the former uses a standardized capital plan mandated by the Dodd-Frank Act, whereas the CCAR uses a firm’s planned capital actions under its BHC baseline scenario.

The significant point about this year’s CCAR was that the Fed proved to be in a remarkably generous mood: it approved the capital plans of 33 out of the 34 banks in the stress test exercise. The one exception, Capital One, was given a conditional non-objection to its capital plan, conditional on its plan making certain improvements by the end of the year.

“This year’s results show that, even during a severe recession, our large banks would remain well capitalized,” said the official in charge of the stress tests, Governor Jerome H. Powell, on June 22. “This would allow them to lend throughout the economic cycle, and support households and businesses when times are tough.”

I disagree.

Generic Weaknesses of the Stress Tests

My doubts start with the observation that the primary purpose of the stress tests program is to reassure the public that the banking system is safe. So when the Fed says the U.S. banking system is in good shape, well, they would say that, wouldn’t they? That is what central banks always say and they can realistically do no other. The stress tests are not some independent assessment of the financial strength of the banking system carried out by experts who are free to arrive at conclusions that might not suit the Fed; instead, the stress tests are part of a publicity campaign by a public agency with is own interests and agenda. Therefore, the credibility of the exercise is compromised before it has even started.

A second problem is that the stress tests place far too much emphasis on a single scenario, its “severe adverse scenario.” Recall that stress tests originated as tools for financial risk management in the private sector, and the financial risk management literature recommends against stress tests that rely on a single stress event. The reason is obvious: a firm will face a range of risk scenarios, and if its risk managers assess its vulnerability to only one of these, then their stress tests cannot give them any reassurance that it will be safe in the event of any major stress scenarios that it did not consider.

The same should apply to stress tests carried out by a central bank. Now admittedly, the Fed’s “severe adverse stress scenario” is, indeed, a severe one. However, it is still only one scenario (if one ignores the pointless “adverse scenario”), and there are other scenarios that could have a significant adverse impact on the U.S. banking system or economy more generally.[5] Examples would be a collapse in the Chinese or eurozone banking systems, a major trade war, or a major shock in the oil market. Do the Fed’s stress tests give us any assurance that the U.S. banking system could withstand such shocks and still be in good shape? No. And why not? Because the Fed’s stress tests did not even consider them.

Turning now to the modeling itself, it seems to me that the stress test projections fail a basic reality check: the Fed’s projected losses seem to be low. First, the projected loss of $493 billion from a scenario that in GDP growth or unemployment rate terms is about as severe as the GFC is barely half the losses accumulated so far since the GFC.[6]

More tellingly, Table 2 on p. 25 of the 2017 DFAST Report indicates that the stressed portfolio loan-loss rate is 5.8 percent. This is a very low loss rate for a major stress. Loan-loss rates in crises can easily be double that (e.g., the loan-loss rate in the United States in the period since the GFC is over 12 percent[7]) and sometimes even higher. These low projected loss rates suggest that the link between the severe adverse scenario and projected losses is too low to be plausible.

The modeling of the Fed’s stress scenarios over the years has been subject to a number of other major problems, some of which are set out in Morris Goldstein’s new book, Banking’s Final Exam.[8] The modeling is much  too orderly to capture key features of real-world financial crises. It understates the fat tails and nonlinearities, does not capture the amplification effects, and ignores the chaos, confusion, contagion, funding and firesale problems involved.

In addition to underestimating the impact of real-sector effects on the financial sector, it also underestimates the impact of financial-sector effects on the real sector. These factors make financial crises much more costly than normal recessions and the stress tests greatly understate them. Goldstein also provides a neat example that shows the near-impossibility of getting empirically plausible models of real-world financial crises:

(a) Note that when former Federal Reserve Chairman Ben Bernanke testified to Congress in 2007 about the subprime crisis, he estimated that it would generate total losses in the neighborhood of $50 billion to $100 billion … (b) But … when Bernanke gave testimony in an AIG court case … he explained that, by September and October of 2008, 12 of 13 of the most important financial institutions in the United States were at risk of failure within a period of a week or two. The question for stress test architects and modelmakers is, How do you make your models generate a transition from (a) to (b) in the course of, say, a year or two? This is not a technical sideshow. In stress modeling, it is the main event [emphasis added].[9]

I would also note several other problems with the Fed’s modeling. First, the Fed’s own history suggests that the Fed itself is a major risk to the banking system, e.g., through its own erratic monetary policy, and yet the Fed’s stress tests take no account of Fed risk. Second, the Fed’s prudential policies, including the stress tests, have the effect of standardizing banks’ risk management practices, thereby exposing the whole system to the weaknesses in the Fed’s own risk management models and creating the potential for hidden systemic risks to which the Fed’s and the banks’ risk models are blind.[10]

There is also a sense in which the stress tests are aiming to achieve the impossible. Frank Partnoy and Jesse Eissinger have made the compelling case that it is no longer possible even for a qualified expert to infer a bank’s true financial state from its audited accounts. But in that case one has to wonder how anyone – the Fed included – can possibly work out a bank’s financial condition from exercises such as the stress tests. It strains credibility to suggest that some questionable spreadsheet model of an arbitrary hypothetical shock can be relied upon to compensate for a major weakness in a bank’s accounts. If a bank has a large hidden (e.g., off-balance-sheet) loss that does not show up on its accounts, then how can we expect a stress test to identify that hidden loss? The answer is that we can’t. If the accounting numbers are not right, then the stress tests cannot correct for them.

Biggest Problems with the 2017 Stress Tests: Low Capital Standards and Regulatory Capture

My biggest concerns with the 2017 stress tests, however, relate to the low pass standards and to the evidence that the longstanding tug of war between the banks and the Fed over the stress tests has finally ended with the victory of the banks over the Fed.

Take the headline capital adequacy metric, the CET1 (Common Equity Tier 1) ratio —that is, the ratio of CET1 capital to Risk-Weighted Assets (RWA). I would immediately dismiss these numbers because the RWA metric is unreliable to the point of being discredited.[11]

The more reliable capital ratio metric is the leverage ratio, defined as the ratio of core capital to some measure of total exposure such as un-risk-weighed assets.

In its 2017 stress tests, the Fed uses two different leverage measures each with its own pass standard: the plain “leverage ratio,” defined as the ratio of Tier 1 capital to average assets, with a pass standard of 4 percent; and a newly introduced and more demanding “supplementary leverage ratio” with the same numerator but an expanded denominator equal to average assets plus (some) off-balance-sheet exposures, and which has a pass standard of 3 percent. These pass standards are not high, however. I find it difficult to believe that a banking system with leverage ratios close to these pass standards would be robust and able to service the real economy as we would wish it to in a subsequent stress scenario—especially bearing in mind the unreliability and gameability of the regulatory and accounting numbers, the off-balance-sheet risks that are not captured in those numbers, and the enormous scope for errors in the stress test modeling.

Moreover, both leverage ratios use Tier 1 capital in the numerator, but Tier 1 is an unreliable capital measure because it includes hybrid securities that are unreliable as core capital in a crisis.[12]

Then there is the question of what the ideal minimum required leverage ratio should be. To start with, many economists have called for higher capital requirements and/or criticized inadequate regulatory capital standards as a contributory factor to the severity of the GFC. These include: James Barth and Matteo Miller, Sheila Bair, Charles Calomiris, James Grant, Thomas M. Hoenig, Simon Johnson, Ed Kane, Norbert Michel, Gerald P. O’Driscoll, Jr.[13], The Systemic Risk Council and Sir John Vickers. In his book, The End of Alchemy, former Bank of England Governor Mervyn King wrote that a “minimum ratio of equity to total assets of 10 per cent would be a good start.”[14] Others would suggest higher minimum ratios. A famous example is an important letter drafted by Anat Admati in the Financial Times in 2010, in which no less than 20 renowned experts recommended a minimum ratio of equity to total assets of at least 15 percent. Independently, Morris Goldstein has recommended a leverage ratio of around 15 percent, John Allison, Martin Hutchinson and yours truly have called for minimum capital to asset ratios of at least 15 percent, and Allan Meltzer[15] and Walker Todd[16] have recommended a minimum of 20 percent for the largest banks.

The pass standards in the stress tests (and those in the leverage ratio tests in particular) are far from being an academic issue. Had the latter been higher, then some or all of the banks would have failed to meet the minimum capital requirements and the Fed would have been obliged to object to their capital plans — in other words, the banks would have been required to further build up their capital.

This issue is particularly acute for the 8 big systemic banks. The following table shows their DFAST and CCAR supplementary leverage ratios under the severe adverse scenario:

Table: Minimum Stressed Supplementary Leverage Ratios for U.S. Systemic Banks (Percent) Bank DFAST CCAR Bank of America 5.4 4.3 Bank of New York Mellon 5.5 4.8 Citigroup 5.5 4.5 Goldman Sachs 4.1 3.1 JP Morgan Chase 5.0 3.9 Morgan Stanley 3.8 3.2 State Street 4.2 3.6 Wells Fargo 6.1 5.3 Weighted average 5.4 4.4 Note: The pass standard is 3 percent. Numbers based on Board of Governors of the Federal Reserve System, Dodd-Frank Act Stress Test 2017: Supervisory Stress Test Methodology and Results, Tables 2 and 4; and Comprehensive Capital Analysis and Review 2017: Assessment Framework and Results, Tables 1 and 6.A. Both reports were published in June 2017.


Several points jump out.

First, the stressed supplementary leverage ratios in the DFAST column are low even before the bank-specific distributions in the CCAR column. Had the Fed applied a higher minimum pass standard for CCAR plans then most if not all of the banks’ capital plans would have been rejected: at a 4 percent minimum, the capital plans of only four banks (BOA, NY Mellon, Citi and Wells Fargo) would have been approved; at a 5 percent minimum, only Wells Fargo’s plan would have been approved; and at a 6 percent minimum none of the big banks’ capital plans would have been approved.

In this context, it is interesting to note that the Fed is in the process of imposing a 5 percent minimum “enhanced supplementary leverage ratio” on the 8 big systemic BHCs, and a 6 percent minimum on their federally insured subsidiaries, to become effective on January 1, 2018. One must wonder then why the Fed chose to impose a 3 percent supplementary leverage ratio on them in the 2017 stress tests. As a general rule, the pass standards in any central bank stress tests should be at least the minimum required standards under the capital adequacy rules, otherwise banks can be (and in the case of the 2017 CCAR seemingly were) deemed to have passed the stress tests even though their stressed leverage ratios appear to fall below the regulatory minima.

Second, the difference between the two columns — the distributions approved by the CCAR vs. those assumed in the DFAST — are remarkably high, with most of them being at least 100 basis points. Not reported in the table are the banks’ plans to increase their payout ratios — their ratios of distributions to net earnings — to nearly 100 percent on average, with some banks planning even higher payout ratios. For the Fed to approve distributions on such a scale, when the big systemic banks have such low leverage ratios and are being subsidized to run down their capital and take excessive risks too, seems reckless in the extreme.

Goldman Sachs gamed the system to near perfection: with a CCAR supplementary leverage ratio of 3.1 percent, it managed to persuade the Fed to approve a capital plan that left it with 10 basis points to spare over the Fed’s regulatory minimum. We can now presumably expect that in future CCARs, the banks’ stressed supplementary leverage ratios will converge to the minimum possible, 3 percent.

Why is the Fed approving such generous distributions when the banks’ leverage ratios are so low? Two words: regulatory capture.

Or consider the following live blog feed from MarketWatch on the Fed’s press conference on June 28th announcing the results of the CCAR:

4:30 pm EDT

One thing that has emerged from the stress tests is that banks are running very close to the minimum thresholds on capital and leverage. A Fed official didn’t deny that. That means the banks are getting good at calculating the maximum payouts they can make to investors without failing the stress tests.

4:30 pm EDT

A Federal Reserve official said banks have substantially increased their payouts, as they’ll be paying out close to 100% of projected net income over the coming four quarters, compared to 65% last year. The Fed sees that as a sign of health. “I’m pleased that the CCAR process has motivated all of the largest banks to achieve healthy capital levels and most to substantially improve their capital planning processes,” said [Governor] Powell in a prepared statement.”

I hope that I am not the only one to sense a disconnect here.

Is the U.S. Banking System Really in Good Shape?

So the Fed is confident that the U.S. banking system is in good shape – and, so much so, that on June 25 this year, Chair Yellen said that she did not expect to see another financial crisis in her lifetime. Let’s wish her a long life and hope that her prediction does not go the way of Keynes’s “We will not have any more crashes in our time” (1926) or Irving Fisher’s “Stock prices have reached what looks like a permanently high plateau” twelve days before the October ’29 crash.[17] But still the doubts creep in. For a start, she has often gotten her forecasts wrong before, even with a large staff of economists from top schools grinding them out. Consider this little beauty from January 22, 2007:

While the decline in housing activity has been significant and will probably continue for a while longer, I think the concerns we used to hear about the possibility of a devastating collapse—one that might be big enough to cause a recession in the U.S. economy—have been largely allayed.

History suggests that crises always recur and that each crisis always catches the central bank off-guard. “This time is different,” they always say, but it never is.

We should be concerned at the obvious blindspots in the Fed’s stress tests: the scenarios not considered, the low pass standards, the danger of hidden systemic risks, and the worry that the stress tests have now become an easy-to-pass compliance ritual that provides little useful information about the true state of the banks. One cannot dismiss the possibility that the good performance of U.S. banks in the latest stress tests merely indicates that banks have mastered the now routine annual stress test game. They have had plenty of practice.

We should also be concerned about how the Fed is going to achieve monetary normalization without inadvertently bursting the “everything bubble” that it has blown so lovingly since 2008. Elementary back-of-the-envelope calculations suggest that losses from a collapse in asset prices could run into trillions, i.e., multiples of the $493 billion losses in the Fed’s “severely adverse scenario” which point also suggests that the Fed’s projected losses might be well on the low side.[18]  Even a natural optimist like Robert Shiller is now warning that his CAPE index is higher than at any time since 2000, the eve of the dotcom crash, and 1929.

The Fed may regard the U.S. banking system as well nigh unsinkable but the Fed’s capital adequacy metrics and stress test pass standards on which this view is based are clearly inadequate. The U.S. banking system is also sailing in iceberg-infested waters and one worries that the Fed’s stress test radar system does not detect even those icebergs that are in plain view.


I thank Anat Admati, Jim Dorn, James Ferguson, Morris Goldstein, Martin Hutchinson, Gordon Kerr, Jerry O’Driscoll, Sir John Vickers and Basil Zafiriou for helpful feedback.

[1] M. Hutchinson, “St Januarius’ blood,” The Bear’s Lair, January 6th 2010.

[2] J. Ferguson, “Annual Fed Stress Test: Turning the dividend ATM on full,” The MacroStrategy Partnership, June 26th 2017, p. 2.

[3] As an aside, one has to ask why share repurchases by deposit-insured banks are even legal? Given bankers’ incentive to game the system by running their capital down to the lowest possible level, share repurchases are highly undesirable because they give bankers an additional and easy-to-implement means of decapitalizing their own banks. Buybacks also destabilize the banking system by making it more pro-cyclical. Banks will repurchase shares at the top, and do emergency issues or lobby for bailouts at the bottom, thereby also helping to destroy their shareholders’ wealth and inflict losses on taxpayers as well. Share buybacks may be justifiable for a company under laissez-faire conditions, but when banks are incentivized to take excessive risk and run down their capital, they become a self-dealing handout to stock-optioned management that undermines the stability of the banking system.

[4] The definitive reference on this issue is A. Admati and M. Hellwig, The Bankers’ New Clothes: What’s Wrong with Banking and What to Do about It, Princeton University Press, 2013. See also L. Gambacorta and H. S. Shin “Why bank capital matters for monetary policy,” BIS Working Papers No. 558, April 2016.

[5] The Fed also applied a global market shock to the trading portfolios of six BHCs with large trading and private equity exposures, and it applied a counterparty default component, which assumes the default of a BHC’s largest counterparty under the global market shock, to the same six BHCs and two other BHCs with substantial trading, processing, or custodial operations. However, these components are add-ons to the economic conditions and financial market environment specified in the adverse and severely adverse scenarios, and the former is merely a watered-down version of the latter. So in essence, the stress tests are still dependent on one key scenario, the severe adverse one.

[6] Ferguson, op. cit., p. 5 estimates that the cumulative loss inflicted on U.S. banks by the GFC was as much as $880 billion, excluding fines. I suppose a defender of the stress tests might argue that the Fed’s projected losses are so low because of the relatively short horizon assumed in the stress test, but this is hardly an adequate defense: the logical implication of this position is that the Fed should use a longer horizon. Remember that many of the losses associated with the GFC took much longer than 2 years to feed through into publicly disclosed realized losses.

[7] Ferguson, loc. cit. estimates that U.S. banks’ loss rate was equivalent to 12.4  percent of peak 2008:Q3 loans of $7.1 trillion.

[8] M. Goldstein, Banking’s Final Exam: Stress Testing and Bank-Capital Reform. Washington DC: Petersen Institute for International Economics, May 2017. See also here and here.

[9] Goldstein, op. cit., p. 251.

[10] Indeed, concerns about the stress test program have recently been voiced by the GAO  and even by the Fed itself.

[11] See, e.g., here, here or here.

[12] Sir John Vickers, “Response to the Treasury Select Committee’s Capital Inquiry: Recovery and Resolution,” March 3rd 2017.

[13] Gerald P. O’Driscoll, Jr., “The Financial Crisis: Causes and Consequences.” The Intercollegiate Review (Fall 2009): 4-12.

[14] Mervyn King, The End of Alchemy: Money, Banking and the Future of the Global Economy, London: Little, Brown, 2016, p. 280.

[15] Cited in A. Admati and M. Hellwig, op. cit., p. 311.

[16] “Start with 20 percent on a leverage basis, not risk adjusted, for the big boys, and then we’ll talk.” Personal correspondence, May 21st 2017.

[17] New York Times, October 17th 1929.

[18] See, e.g. K. Dowd and M. Hutchinson, “From Excess Stimulus to Monetary Mayhem,” Cato Journal Vol. 37, Number 2, Spring/Summer 2017, p. 316, note 16.

[Cross-posted from Alt-M.org]

In 2012, various properties in Van Buren County, Michigan became subject to foreclosure for property tax delinquencies. In 2014, the properties were subject to an order of foreclosure and were auctioned off to satisfy the delinquencies. Wayside Church owed $16,750 in back taxes on a parcel it used as a youth camp. When the property was sold for $206,000, Van Buren County kept the $189,250 in surplus as required by Michigan’s General Property Tax Act. Other taxpayers were similarly situated. For example, Myron Stahl and Henderson Hodgens had their properties auctioned for $68,750 to pay a $25,000 debt and $47,750 to pay a $5,900 debt, respectively.

Michigan law doesn’t recognize a right to surplus proceeds from tax sales, so the property owners sued in federal court, alleging that the county violated the Fifth Amendment’s Takings Clause when it kept the surplus proceeds from the sale of their properties. The district court dismissed the suit, precisely because Michigan law doesn’t recognize a right to surplus proceeds in such cases. On appeal, a divided Sixth Circuit dismissed the case for lack of jurisdiction. Citing the Supreme Court’s ruling in Williamson County Regional Planning Commission v. Hamilton Bank of Johnson City (1985), the court held that plaintiffs’ failure to first pursue avenues of relief in state court barred the door to federal court.

Wayside Church and the other property owners filed a petition asking the Supreme Court to take the case and clarify takings law. Along with the National Federation of Independent Business, Southeastern Legal Foundation, and Prof. Ilya Somin, Cato has filed an amicus brief supporting that petition. We argue that this case provides an excellent opportunity to preferably overrule, but at least reconsider, Williamson County’s requirement that a property owner must first sue in state court to ripen a federal takings claim.

The reality is that Williamson County’s state-remedies requirement results in constitutional absurdity: the very state court decision that a property owner must receive in order to ripen their claim simultaneously bars the owner from (re)litigating the issue in federal court. The Williamson County rule has also proven to be a potent weapon in the hands of manipulative defendants. Since the Supreme Court ruled in 1997 that a takings claim filed in state court could be “removed” to federal court (because of the federal constitutional issue), governmental defendants have removed claims to federal court, and then argued that they should be dismissed as unripe!

The state-remedies rule has no doctrinal basis and is antithetical to the Fourteenth Amendment, which was ratified to secure constitutional rights against the states and was seen as necessary to curb state government abuses. Fearing state courts could not be trusted to enforce the U.S. Constitution against their own state governments, a federal civil rights law 42 U.S.C. § 1983 was then enacted to ensure a federal forum for vindicating federal rights. Yet Williamson County has effectively gutted the protections of both of these Reconstruction-era reforms.

Before Williamson County, there was no rule that required a property owner to resort to litigation in order to ripen a takings claim, and nothing in the text of the Fifth Amendment suggests that litigation in state court is necessary to ripen a takings claim. Instead, the text should be read to recognize a ripened claim the moment property is taken if there isn’t a readily available administrative procedure for obtaining just compensation. 

The Supreme Court will decide this fall whether to take up Wayside Church v. Van Buren County.

The North American Free Trade Agreement has been a source of controversy since well before its implementation in 1994.  It was the first trade agreement involving the United States and a “developing” country, so it raised concerns that a giant sucking sound from south of the border would hoover up U.S. investment and jobs.  Ross Perot, Pat Buchanan, and most Democratic presidential candidates beginning with John Kerry all lamented the imminent or unfolding devastation wrought by NAFTA.

Even though the U.S. manufacturing sector has continued to attract more investment than every other countries’ manufacturing sectors ever since NAFTA was implemented, and even though that implementation did not accelerate the trend of U.S. manufacturing job decline, which had been underway for 14 years since employment peaked at 19.4 million in 1979 (2.6 million decline between 1979 and 1993; 2.7 million decline between 1993 and 2007; 600,000 increase between 1993 and 1999), NAFTA became a symbol of corporate excess and a rallying cry for organized labor, environmental organizations, and other anti-business groups over the years.  It also made it nearly impossible for Democrats in Congress to support trade liberalization in the ensuing decades.

During the 2008 presidential election campaign, Democratic candidates John Edwards, Hillary Clinton, and Barack Obama all vowed to re-open NAFTA to make it less unfair for U.S. workers.  Within a few weeks of assuming office, President Obama let the president of Mexico and the prime minister of Canada know that he wasn’t about to follow through on his NAFTA pledges and risk disrupting North American production and supply chains that have enabled regional producers to compete more effectively against Asian and European rivals, while delivering better goods and services at more affordable prices to consumers.

Probably owing to the anti-trade agreement fervor that brewed during the debates over Trade Promotion Authority and the Trans-Pacific Partnership over the last few years, killing NAFTA (and the TPP) became a central plank in Donald Trump’s presidential campaign. Although, regrettably, he withdrew the United States from the TPP, Trump seems to have been talked off the ledge about jettisoning NAFTA , which (as of this morning) is being renegotiated.

As a guide to better understanding what’s on the table and what’s at stake, my colleagues Simon Lester, Inu Manak, and I produced this working paper: Negotiating NAFTA in the Era of Trump: Keeping the Trade Liberalization In and the Protectionism Out.


  • When NAFTA came into force in January 1994, it was a groundbreaking achievement. It eliminated nearly all tariffs among three significant trading partners and achieved liberalization on a wide range of other issues (some of which had never before been included in trade agreements).
  • Total U.S. trade in goods and services with Canada and Mexico reached over $1 trillion in 2016, growing 125.2% in real terms since 1993.
  • NAFTA’s liberalization not only encouraged more trade, but more cross-border investment leading to deeper integration of production networks.  
  • On a historical cost basis, foreign direct investment in the United States from Canada increased from $40.4 billion in 1993 to $269 billion in 2015, and from Mexico it increased from $1.2 billion to $16.6 billion.
  • Today, 40 cents of every dollar of U.S. goods imports from Mexico and 25 cents of every dollar imported from Canada is U.S. value added (material, labor, and overhead contributed in the United States).

Still, the agreement can afford some updating to incorporate issues that were not included originally, to fix some of the flaws that have become evident over nearly a quarter of a century, and to reduce trade barriers in sectors that were originally untouched by liberalization.



NAFTA was drafted in the pre-Internet era, when E-Commerce was virtually non-existent.  Since then, we have seen the growth of online product sales, the conversion of certain products into services delivered electronically, and the development of various online “platforms.”  E-Commerce is now a standard way of doing business for many companies, in both domestic and international trade.  The free flow of information is essential to free trade in electronic commerce, as well as to the industries for which data are crucial components of the product or service.  While governments have grappled with various responses to the growth of E-Commerce, trade policy has been slow in keeping pace with the changes.  But in recent years, U.S. trade negotiators have been making some progress on this issue. 

State-Owned Enterprises

State-owned enterprises are commercial enterprises in which the state has majority ownership, controlling ownership, or the ability to appoint a majority of the board of directors. SOEs have become more prominent actors in the global economy in recent years. Concerns about trade distortions and unfair competition have grown, as SOEs that had previously operated almost exclusively within their own territories are increasingly engaged in international trade and cross-border investment. According to the Office of the U.S. Trade Representative, there was only one SOE on the list of the Fortune Global 50 largest companies in the world in 2000.  By 2015, there were close to a dozen.  The purpose of trade agreement rules on SOEs is to curtail their market distorting effects.

Regulatory Cooperation

Since NAFTA reduced almost all tariffs to zero, some of the largest remaining barriers to North American trade take the form of non-tariff barriers, or so-called “behind the border” measures. Differences in standards and regulations are frequently cited as key obstacles to trade, and many of these regulatory divergences are often the result of regulation occurring in domestic silos, without giving much thought to how those rules may impede trade.

Trade Facilitation

In broad terms, trade facilitation includes reforms aimed at improving the chain of administrative and physical procedures involved in the transport of goods and services across international borders. Countries with inadequate trade infrastructure, burdensome administrative processes, or limited competition in trade logistics services are less capable of benefiting from the opportunities of expanding global trade.  Like tariff cuts, improvements in trade facilitation procedures can help reduce the cost of trade and increase its flow. 

Dispute Settlement

The enforceability of the rules of a trade agreement is of great importance because if governments cannot be held to honor their commitments, the value of those commitments is significantly diminished. Inevitably, disputes arise over the meaning of the obligations in trade agreements. Typically, dispute settlement involves ad hoc panels of experts who hear claims from either governments or private actors, and then issue rulings on those claims. Essentially, these panels act as the judicial arm of the trading system.  If trade agreements are to be reliable and effective, they must include enforcement mechanisms. The degree of enforceability varies depending on the design of the particular system.  In this section of the working paper, we discuss how Chapters 20, 19, and 11 may be addressed in the NAFTA renegotiation.

Rules of Origin

Rules of Origin (RoO) establish the parameters used to determine whether an imported good “originates” within the region, thereby qualifying for the preferential treatment accorded under the agreement. Generally, a product is considered originating if it was wholly made within the region (in the countries party to the agreement), if it was significantly transformed within the region from imported materials and components, or if the relative value of originating materials and manufacturing performed in the region meets a specific threshold.

RoO are necessary components of preferential trade agreements. When products from different countries receive different tariff treatment, importers, exporters, and customs officials must have a way to determine which tariff rates apply.  Rules that permit greater use of non-originating inputs or broader definitions of what constitutes product transformation tend to be more trade liberalizing than more proscriptive rules, which impose greater restrictions on qualification for the agreement’s preferential tariff rates. In today’s globalized economy, strict rules of origin impede the evolution and operation of more efficient supply chains and can be used to privilege existing producers by limiting competition.

Canada’s Supply Management System

Canada’s supply management system for dairy, poultry, and eggs was excluded from NAFTA. The renegotiation offers an opportunity to bring these highly-protected sectors into the agreement, so as to increase market access for U.S. producers, but also to improve purchasing choices for Canadian consumers. For instance, Canada maintains a 270 percent tariff on milk, and the above-quota tariffs on cheese and butter can reach as high as 245 percent and 298 percent, respectively.

Softwood Lumber

The 35-year old lumber dispute between the United States and Canada must be resolved, somehow.  U.S. producers have long complained about Canadian land-management practices that, they claim, provided unfair subsidies to Canadian producers.  Three and a half decades of litigation, temporary supply management agreements, and, frankly, defiance of the trade rules by the U.S. government have strained trade relations and burdened U.S. lumber users and home buyers.  Whether an agreement can be reached within NAFTA is unclear – given that it probably would require the U.S. industry to disavow use of the antidumping and countervailing laws.


NAFTA was the first international trade agreement with significant commitments to liberalize trade in services, and one of the most prominent NAFTA disputes concerns the provision of trucking services in the United States by Mexican firms.  However, in the ensuing years, services liberalization in other trade agreements has evolved to include broader obligations covering more areas of trade, leaving NAFTA’s rules insufficient and outdated.  In the NAFTA renegotiation, there is as an opportunity to take a fresh look at services, such as legal, medical, and educational services, that are more easily traded across borders than was the case in the early 1990s. With growing concerns over the absence of sufficient competition in the provision of U.S. healthcare services, the United States could benefit by allowing more foreign competition in the medical services and health insurance markets.


Trade Deficits

President Trump and members of his administration seem to believe that trade is a zero sum game played between Team USA and – in the case of NAFTA – Team Canada and Team Mexico.  They consider exports to be Team USA’s points, imports to be the foreigners’ points, and the trade account to be the scoreboard.  Since the United States runs a persistent deficit with Mexico (and an occasional deficit with Canada), they conclude that America is losing at trade.  And it’s losing because of poorly negotiated trade deals or outright cheating on the part of the foreign teams.

Secretary of Commerce Wilbur Ross and National Trade Council advisor Peter Navarro have suggested that our trade agreements should include some sort of trigger mechanism, so that if trade balance or some other objectives related to reducing deficits or increasing surpluses are not reached, the deal would be reopened and renegotiated.  That idea, frankly, is absurd.  What would motivate businesses to invest in cross-border relationships or to commit to capital investments if the terms of the underlying trade agreement were prone to such uncertainty and potential upheaval?

Moreover, the bilateral trade balance is not a measure of the success or failure of a trade agreement.  It is a meaningless statistic.  A better measure of the success of a trade agreement is its effect on total trade and investment, and its effect on economic growth.  By those metrics, NAFTA has been an unbridled success.  Furthermore, “despite” 41 straight years of registering annual trade deficits, the U.S. economy has grown significantly, in part because it has benefitted from the capital account surplus (foreign purchases of U.S. assets exceeding U.S. purchases of foreign assets) that goes along with a current account deficit.

The Trump administration’s obsession with trade deficits—especially bilateral trade deficits—is misguided.  The point of trade agreements is to reduce artificial impediments to trade and to provide greater certainty.  It would create vast uncertainty and be a serious mistake to push for any provisions in NAFTA that use the trade account as a trigger of some future action.

Border Taxes

Another problematic suggestion from the Trump trade team is that they would like to use trade agreements to take on the “border adjustment taxes” associated with value-added taxes used by many countries.  Trump advisors have expressed concerns over alleged advantages bestowed upon Mexican producers by way of the rebate of value-added taxes upon export and the assessment of taxes on U.S. products upon import.

A proper understanding of these border adjustment taxes is that their impact on trade is much more benign. The general understanding, according to WTO rules, is that value-added taxes are a consumption tax, similar, in nature, to a sales taxes, and thus a non-discriminatory application of such taxes to domestic and imported products is permitted.  Not that it would necessarily be a good idea, the U.S. government is free to adopt a similar VAT system with rebates upon export or some variation, such as a destination based cash flow tax, which was under consideration in Congress (although such a measure would, of course, have to be applied in a non-discriminatory way).  As of this writing, that controversial proposal seems to have been abandoned.

Regardless, this is not a NAFTA problem or a Mexico problem. Much of the world uses value-added taxes or some similar form of consumption taxes. If the U.S. government has concerns, it can address them at the WTO. There is a long history of disagreement over this issue in the GATT, and finding a way to raise it again at the WTO would not be out of line. Raising it in NAFTA, on the other hand, would be a pointless distraction.

Currency Manipulation

Another controversial issue the Trump administration is likely to press is currency manipulation. While economists are divided in their opinions about how to define currency manipulation and whether it constitutes an especially pernicious offense worthy of counteractions, even many of those who think currency manipulation should be discouraged are skeptical of the wisdom of including punitive provisions in trade agreements.  Even though Canada and Mexico do not generate much concern in the United States as likely currency manipulators, the Trump administration sees NAFTA as an opportunity to set a precedent for future agreements by crafting and incorporating such provisions without much resistance in this agreement.

China and Japan are the countries most often cited as the reasons for including currency manipulation provisions in trade agreements.  But we should be skeptical about the assumptions underlying the relationship between currency manipulation and trade flows.  Given that intermediate goods trade predominates total trade, as a result of the proliferation of global supply chains,  the relationship between currency values and trade is not straightforward.

If only 50 percent of the value of a country’s exports is domestic content and labor (and the other 50 percent is foreign value), as is the approximate case with China, the impact of currency values on trade flows is mitigated.  This helps explain why, despite a 38 percent appreciation of the renminbi against the dollar between 2005 and 2013, the bilateral U.S. trade deficit with China didn’t decrease. Rather, it increased by 46 percent. 

Even if deemed desirable, crafting the right currency manipulation provision would be more difficult than the administration thinks.  There is vast disagreement among economists as to how to identify and measure the effects of currency manipulation. Meanwhile, U.S. government agencies have conflicting views about incorporating currency provisions in trade agreements.  While the Commerce Department—with its mission to protect U.S. industries in the global market place—might favor another weapon in its trade remedy arsenal, the Treasury Department, which has long held sway over financial and economic issues related to currency, is less inclined to take actions in currency markets to achieve trade objectives.

And while the Canadian and Mexican governments might not be overly concerned that their actions would be the targets of such provisions, they will certainly have their own opinions about whether and how such rules should be crafted. Thus, balancing domestic and international views on this issue will take time. If the Trump administration chooses to push this issue, the NAFTA renegotiation may not go as quickly as it wishes.

Buy America

Trump and his trade team have spent a lot of time promoting the idea of Buy America, and have criticized trade agreements that restrict the ability of the U.S. government to buy exclusively from Americans. In the leaked draft memo about NAFTA, it was suggested that the NAFTA renegotiation would provide more flexibility for such Buy America programs. What those pushing for expanded Buy America do not seem to realize is that this issue works both ways.

Cordoning off more of the estimated $1.7 trillion U.S. government procurement market to U.S. suppliers would mean higher price tags, fewer projects funded, and fewer people hired. In today’s globalized economy, where supply chains are transnational and direct investment crosses borders, finding products that meet the U.S.-made definition is no easy task, as many consist of components made in multiple countries. And by precluding foreign suppliers from bidding, any short-term increases in U.S. economic activity and jobs likely would be offset by lost export sales – and the jobs that go with them – on account of copycat protectionism abroad. If Americans are compelled by law to buy more U.S. product from U.S. firms, then the Canadians and Mexicans will insist on the same.

Buy America may sound good in a campaign speech or on a bumper sticker, but in practice it just means higher price tags for taxpayer funded projects and fewer opportunities for U.S. exporters.

On the campaign trail, Donald Trump claimed that he would cancel President Obama’s Deferred Action for Childhood Arrivals (DACA) program, which allows young unauthorized immigrants—known as “Dreamers”—to live temporarily without fear of removal and work legally. To many people’s surprise (including mine), President Trump decided in January to maintain the program, issue renewals, and even allow new applicants into the program. But this will likely change soon.

Now several states led by Texas are attempting to force the president’s hand, requesting in a letter that he terminate the program by September 5 or face a lawsuit. Texas already successfully challenged President Obama’s 2014 attempt to expand DACA to a broader range of immigrants who came here as children and to create a new program for undocumented parents of U.S. citizens called DAPA. This makes a lawsuit very likely to succeed. Even if it were possible for the Trump administration to defend DACA legally, it is not clear that Attorney General Jeff Sessions would want to defend a policy that he has called constitutionally “very questionable.” In July, he refused to say he would defend it.

After the election, President Trump promised to “work something out” with the Dreamers, and in July, President Trump said that he alone will decide the future of DACA. But the politics just became much more difficult. Texas has essentially forced him to defend in court something that he had characterized as illegal amnesty on the campaign trail and something his biggest supporters hate. Moreover, Trump’s record with the courts is already making him appear inept, so he would likely not want to take the political hits when he could easily lose the case anyway.

If the president does rescind the DACA memorandum on September 5, the program will likely not disappear overnight. Rather, it will slowly wind down over the next two years.

Will Dreamers become priorities for removal?

DACA has three different aspects—deprioritization for removal, “deferred action,” and employment authorization. First, the DACA memo tells agents to prevent Dreamers who may be eligible for DACA “from being placed into removal proceedings or removed from the United States.” This provision deprioritizes their removal. Under new expansive enforcement priorities laid out in a February 20th memo from then-Department of Homeland Security (DHS) Secretary John Kelly, many DACA recipients would be targets for removal if Kelly’s successor rescinds the DACA memo. That’s because the Kelly memo creates “priorities” so expansive that they include nearly all unauthorized immigrants, while specifically rescinding all “conflicting” memos except for the DACA memo and the memo that expanded DACA and provided for the never-implemented DAPA program. 

Like the DACA memo, the DAPA memo instructed Immigration and Customs Enforcement (ICE) “to… prevent the further expenditure of enforcement resources with regard to these individuals” (i.e., undocumented parents of U.S. citizens and DACA children). After initially leaving it in place, Secretary Kelly rescinded the DAPA memo on June 15. Since then, DAPA-eligible immigrants have been eligible for removal, though it appears that ICE was already ignoring those provisions of the DAPA memo. If the DACA memo were rescinded without further DHS guidance, the same thing would happen to the Dreamers that it protects.

Texas and the other states threatening a lawsuit claim that their “request does not require the Secretary to alter the immigration enforcement priorities contained in his separate February 20, 2017 memorandum.” Nor does it “require the federal government to remove any alien.” But their letter simultaneously calls for the full rescission of the DACA memo, which would effectively alter the enforcement priorities in the February 20th memo and, absent any additional guidance, require the federal government to target Dreamers for removal.

Will Dreamers immediately lose deferred action?

The second aspect of DACA is “deferred action.” Deferred action provides protection against arrest by formalizing the decision not to remove them. If an ICE agent encounters a DACA recipient and enters their name into the agency’s database, they will find their status listed as “lawfully present.” DACA applicants submit a form I-821D application to U.S. Citizenship and Immigration Services with a $495 application fee, which covers the cost of a biometric background check. As of March 2017, 886,814 young immigrants had received an approved I-797 Notice of Action form under DACA. Each I-797 form is valid “unless terminated” for two years.

Figure 1: DACA Form I-797 Notice of Action

Source: Imgur

If DHS cancels the DACA memo, it’s unclear whether the underlying deferred action grants would disappear. The threat letter from the states only requests that “the Executive Branch will not renew or issue any new DACA or Expanded DACA permits in the future.” The form itself states that it will “remain in effect for 2 years” unless “terminated.” Does this require an individualized termination for some violation of the program’s rules or would the termination of the program under which the form was issued suffice?

When Secretary Kelly rescinded the DAPA and Expanded DACA memo, he specifically stated that the memo would not “alter the remaining periods of deferred action under the Expanded DACA policy granted between the issuance of the November 20, 2014 Memorandum and the February 16, 2015 preliminary injunction order in the Texas litigation.” This refers to the 3-year renewals issued to original DACA recipients who, both before the November 20th DAPA/Expanded DACA memo and after the injunction, received only 2-year renewals. It is possible that the administration could duplicate this language in the DACA repeal itself.

Another option would be for the administration to leave it unclear, and if it does so, I suspect that individual ICE agents will end up making the determinations. According to the Feb. 20th Kelly memo, “the individual, case-by-case decisions of immigration officers” will have the final say over whether to make arrests.

Will Dreamers immediately lose employment authorization?

The final aspect of DACA is employment authorization. Every DACA recipient receives a 2-year (or 3-year, if they renewed in the period mentioned above) employment authorization document (EAD). DACA applicants must separately submit a Form I-765 to receive an EAD. You can see an approved DACA EAD in Figure 2. The card contains a two-year validity period.

Figure 2: DACA-based Employment Authorization Document (EAD)

Source: WangLawOffice

Because employment authorization is dependent on a grant of deferred action, canceling part 2 of DACA would legally end part 3 as well. However, the memo that rescinded the DAPA/Expanded DACA memo also clarified that it does not “affect the validity of [3-year] Employment Authorization Documents” granted before the injunction in the Texas case. Again, it’s possible that the administration will repeat this clarification in the DACA rescission. In any case, the EADs contain no indication that they were issued under DACA, and employers must accept any facially valid, non-expired ID when screening applicants for employment authorization. I have previously written about the Obama administration’s efforts to cancel certain DACA EADs and how difficult it was in my pre-inauguration post explaining how DACA could end. Given that this circumstance seems quite unlikely, I will not repeat myself here.

How DACA will end

We know roughly how DACA will wind down because we know about when DHS approved the DACA applications. DHS publishes quarterly approval figures, and we know that it issued 108,000 3-year renewals from November 20, 2014 to February 16, 2015. Figure 3 shows the approval schedule for DACA as well as the expected expiration schedule if DHS cancels the program on September 5, 2017, as requested in the states’ letter. As it shows, the DACA grants would expire over a two-year period ending in September 2019.

Figure 3: DACA Approvals—Initial and Renewals—and Projected Expirations

Source: USCIS Data Set: Form I-821D Deferred Action for Childhood Arrivals. April 2017 to September 2017 figures are not published yet and are projections based on the number of two-year renewals in those months in 2015.

Unfortunately, the September 5 deadline will occur just before the expiration of the 108,000 three-year permits. Had President Obama issued only two-year renewals, these individuals would have already renewed their status a year ago, receiving an additional year of protection. Figure 4 provides the expirations by year. More than 60 percent of DACA recipients will still have protection and lawful employment through June 2018. More than a third will still be participating in DACA through January 2019—well over a year after the rescission of the DACA memo.

Figure 4: Projected DACA Expirations by Year

Source: Author’s calculation based on USCIS DACA Approval Figures.

The most likely scenario for DACA’s demise is that it will slowly wind down. President Trump has shown no willingness to move more aggressively against DACA recipients than is necessary, although certain ICE agents seem zealous about targeting them. If the president does determine that DACA should end, or the states receive an injunction against the program, Congress will need to act quickly as the permits expire. I suggest that they look seriously at the Recognizing America’s Children (RAC) Act, which would provide legal status and a pathway to citizenship for these immigrants brought to the U.S. as children.

Add yesterday’s rage-spasm of a press conference to the growing list of reasons reasonable people are inclined to worry about Donald Trump’s proximity to nuclear weapons. In addition to what it suggested about Trump’s moral compass (“Very fine people” aren’t attracted to posters that look like this), his performance also highlighted questions about the judgment, temperament, and impulse control of the man entrusted with the world’s most fearsome arsenal.

Last week, recall, Trump threatened North Korea with nuclear annihilation: “North Korea best not make any more threats to the United States…. They will be met with fire, fury and frankly power the likes of which this world has never seen.” “Fire and fury” was ad-libbed, apparently, but on Thursday, he upped the ante: “if anything, that statement may not be tough enough.” (For a cooldown lap, on Friday, Trump warned he was “not going to rule out a military option” in Venezuela.)

When you’re faced with a president who has weekly meltdowns on Twitter and likes to “wing it” with nuclear threats, it tends to concentrate the mind painfully on the legal and practical restraints to presidential power. Does the president have the constitutional power to launch a nuclear first strike on a country for “mak[ing] threats”? If he decides to act on that impulse, is there anything Congress can do to stop him?

The first question’s the easy one: the answer is no. In the absence of an imminent attack, the president has no constitutional power to rain down “fire and fury” on North Korea. As Ilya Somin explains here, “the Constitution very clearly reserves to Congress the power to start a war.”

The president retains some independent power to act defensively: to “repel and not to commence war” or “repel sudden attacks,” as Madison’s notes from the Convention put it. We can argue about whether a second strike—launch under attack—is included within this power. But the constitutional power to “repel sudden attacks” doesn’t include the power to launch them.

The whole point, as James Wilson told the Pennsylvania ratifying convention in 1787, was to design a “system [that] will not hurry us into war…. It will not be in the power of a single man… to involve us in such distress; for the important power in declaring war is vested in the legislature at large.”

Is there anything Congress can do to prevent a trigger-happy president from hurrying us into nuclear war?  Congressman Ted Lieu has drafted a bill that he hopes will do just that.  HR 669, the “Restricting First Use of Nuclear Weapons Act,” provides that “The President may not use the Armed Forces of the United States to conduct a first-use nuclear strike unless such strike is conducted pursuant to a declaration of war by Congress that expressly authorizes such strike.”

Someone from the John Yoo school of constitutionalism might argue that the law encroaches on presidential prerogatives by “micromanaging” the means available to protect national security. But Lieu’s bill is clearly constitutional: if Congress can tell the president not to use ground combat troops in a particular war, it has the legal authority to bar him from launching an unauthorized nuclear first strike.

Would Lieu’s bill work, though? Here, I have my doubts.

In principle, the military is not supposed to obey an illegal order by the president, but in practice, they follow the chain of command. In all but the most extreme cases—e.g., orders to commit war crimes—it’s good that they do. Civilian control of the military is a core principle of free government, and it’s almost always better preserved by erring on the side of military obedience to the CINC, even where his authority may be in doubt.

As a practical matter, though, that means that simply passing a law won’t necessarily restrain presidential warmaking. Our experience with the War Powers Resolution makes that clear. Put aside the question of how the law applies to limited, short-term military operations: if the WPR does anything, it bars the continuation of hostilities beyond 60 days unless Congress has declared war or passed a specific statutory authorization for war. In fact, “the view that the 60-day clock is constitutional has remained consistent executive branch law throughout the Reagan, Bush-Quayle, Clinton, Bush-Cheney, and Obama administrations.” But twice in recent decades, presidents have blown by that limit: first in Kosovo in 1999; second in Libya in 2011. Each time, the clear language of the law didn’t matter: the president ordered the military to keep bombing, and they did.

Why would Lieu’s “no-first-use act” fare any differently? The nuclear launch process is built for speed—designed to give the president the power to launch a second strike in mere minutes, while enemy missiles are still in the air. That also makes it difficult to stop a president who decides to launch first. The guys in the bunkers who turn the keys have surely been vetted to weed out “question authority” types. In the ‘70s, there was actually an Air Force major who was training for the job, and when he asked, “how can I know that an order I receive to launch my missiles came from a sane president?”—that was the end of his career.

Trump can’t literally “push the button” and have Pyongyang obliterated. But once the order is authenticated with the codes on the “biscuit,” and it goes out from the Pentagon, it’s likely already too late.

The one effect Lieu’s bill might have, and this is speculative, is it might embolden the president’s secretary of defense or the chairman of the Joint Chiefs to put up roadblocks before the order is transmitted. (I say “or the Chairman of the Joint Chiefs” because it appears that the president has the option to issue the order without going through the Secretary of Defense).

There’s a story about Richard Nixon, whose behavior during the Final Days—frequently drunk and raving—made people worry about his access to nukes. In a meeting with congressmen, Nixon blurted out that: “I can go in my office and pick up a telephone, and in 25 minutes, millions of people will be dead.” One of the attendees, Senator Alan Cranston called then-Secretary of Defense James Schlesinger, worried about “the need for keeping a berserk president from plunging us into a holocaust.” Schlesinger apparently told the top military brass that if the president issued any unusual orders, they needed to check with him first.

The practical effect Lieu’s bill could have, in the unlikely event it passed, might be to buttress a future Schlesinger. By giving the Secretary of Defense or the Chairman added cover to get in the way of a decision he knows to be crazy, it could slow or stop the delivery of the order from the Pentagon to the launch officers. For that reason alone, it’s worth considering. But when you’re depending on a guy whose nickname is “Mad Dog” to preserve sanity within the executive branch, you’re already in a pretty dangerous place.

I am pleased to announce that this week the Cato Institute published the Encyclopedia of Libertarianism online, unabridged, and unpaywalled

The Encyclopedia was my first project at Cato, and if you ask me its current format is exactly where it belongs. Some of the articles are a bit dated by now, and there are some regrettable gaps in the first edition. We will be working on those in the coming months to create a new and improved Encyclopedia. In the meantime, though, feel free to browse. I still think the original is a pretty good statement of what libertarianism is all about.

When the Department of Labor (DoL) rolled out its fiduciary duty rule last year, I (and others) noted that its likely effect would be to harm the very people it purports to protect.  Unfortunately, it seems I was right.

The rule is intended to help individuals make good choices when saving for retirement.  Under the rule, brokers who sell retirement investments are to be held to a “fiduciary duty” standard.  This is often expressed in the legal world as the care a prudent person would take in managing his or her own affairs.  Those who hold positions of great trust, such as those who are given authority to act for another, are often designated fiduciaries under the law.  For example, corporate board members are fiduciaries of the company they serve, and lawyers owe a fiduciary duty to their clients. 

As I’ve discussed previously, the legal reality of being subject to a fiduciary duty standard is much more than simply deciding to offer good customer service.  And “more” in this case means more liability and more cost.  The risk is therefore that brokers may find it’s just not worth the risk or the cost to serve clients with only moderate amounts of money to invest.

It seems a recent poll of the industry shows my predictions may be correct.  According to a letter submitted to the DoL by the Financial Services Roundtable, its members have reported the following trends as a result of the rule:

 (1) less guidance and support to IRA owners and small plans; (2) increases in minimum account size; (3) limited product shelf; (4) shift to fee-based accounts; (5) moving clients with smaller accounts to self-service or robo-advice; (6) orphaning of smaller, less profitable accounts due to heightened risks; (7) reduced willingness to discuss or consider unmanaged assets with clients due to risks; (8) poor client service due to the time required to perform comparative analysis on the proposed account to the existing account; (9) disinclination to sell annuity products because of uncertainty surrounding the Rule and inability to launch new products because resources are tied up with Rule implementation; (10) additional disclosure documents and other changes to sales process make the sales process markedly longer in each client appointment; (11) less discretion on small accounts and compensation changes make working with small accounts more challenging and less cost-effective for financial professionals;  (12) higher manufacturing and distribution costs; and (13) new liability concerns.

Specifically, the poll found that 68 percent of respondents would be taking on fewer small accounts due to increased compliance costs and legal risks.  It also found that 63 percent expected that the new rule would limit the investment options or products the firms could provide to their clients.  And that 52 percent expected that higher compliance costs would be passed on to clients in the form of additional fees. Only 12 percent of respondents said the rule “is helping me to serve my clients’ best interests.”  Most notably, the poll report highlights the following: “Even advisors who say that the rule is ‘helping me to serve my clients’ best interests’ or has had ‘no impact on my ability’ say that there will be more complicated paperwork and fewer small accounts.”

An anonymous source sent an advanced copy of S.1757, otherwise known as the “Building America’s Trust Act,” to Ars Technica. If passed as written, the bill would dramatically expanded surveillance at the border and ports of entry, putting the privacy of immigrants and citizens alike at risk.

The bill, sponsored by Sen. John Cornyn (R-TX) and co-sponsored by six of his Republican colleagues, mandates increased border drone surveillance and the collection of more biometric information, including but not limited to voice prints and facial scans.

The drone provisions of the bill are consistent with President Trump’s campaign rhetoric. During his campaign, he said drones should patrol the border 24/7. Cornyn’s bill doesn’t quite go that far, requiring Customs and Border Protection (CBP) to fly drones at least 24/5.

Drone surveillance at the border isn’t new, nor is it effective. In 2014 the Department of Homeland Security’s Office of Inspector General (OIG) released a report on CBP’s drone operations. It found that the drone program, which includes Predator B drones originally designed for military use, did not achieve expected results and contributed to very few apprehensions of illegal border crossers and marijuana. The report also found that the drone program cost $12,255 per flight hour. In FY2013, CBP’s drones flew for 5,102 hours for a combined cost of around $62.5 million. 

This was a large expense for an inefficient border security tool. Aside from the fiscal costs, the increased use of drones on the border will worsen the militarization of the border, with American citizens being under the ever-snooping eye of border patrol surveillance equipment. In 2013, Americans on the border were already regularly seeing military-grade surveillance tools in the air. From a 2013 New York Times report:

The United States-Mexico border has become a war zone. It is also a transfer station for sophisticated American military technology and weapons. As our country’s foreign wars have begun to wind down, defense contractors look here, on the southern border, to make money.

Lately it has become entirely normal to look up into the Arizona sky and to see Blackhawk helicopters and fixed-wing jets flying by. On a clear day, you can sometimes hear Predator B drones buzzing over the Sonoran border. These drones are equipped with the same kind of “man-hunting” Vehicle and Dismount Exploitation Radar (Vader) that flew over the Dashti Margo desert region in Afghanistan.

CBP drones do not have to be large, military-grade predator drones. Earlier this year I noted that The Department of Homeland Security (DHS), CBP’s parent agency, is interested in small, portable drones outfitted with facial recognition technology.

Facial recognition is also mentioned in S.1757 as part of the bill’s passport screening section, requiring that CBP “utilize facial recognition technology or other biometric technology” to “inspect travelers at United States airports of entry.”

DHS stated earlier this year in a privacy impact assessment concerning facial recognition, “the only way for an individual to ensure he or she is not subject to collection of biometric information when traveling internationally is to refrain from traveling.” There are already facial recognition trials at airports in six American cities, but the technology will become more prevalent if S.1757 is passed.

Unlike other biometrics such as fingerprints and DNA, facial images can be obtained from individuals who haven’t come into contact with the criminal justice system, the primary source of most biometrics. About half of American adults are already in a facial recognition network thanks to these networks’ access to DMV and passport data. If American citizens’ passport and driver’s license facial images are going to be included in airport facial recognition systems there should be strict limits on sharing these images with other law enforcement agencies and how long facial scans are stored. Although CBP states that facial scans at airports are deleted after 14 days that commitment is policy, not law, and could change.

S.1757 mandates that a biometric exit program will be implemented in America’s 15 busiest airports, seaports, and land ports of entry within two years of enactment. One section of S.1757 requires DHS to upgrade or build a biometric system that enables DHS to store alien iris scans and voice prints that federal, state, and local law enforcement can access.

As if the surveillance technology outlined in S.1757 wasn’t concerning enough, the legislation also includes an increase in the number of border patrol agents (at least 26,370 from around 20,000) and more federal judges in southern border states. Under the Border Patrol Strategic Plan included in S.1757 DHS will consider using military technology.

More drones and biometric technology tools on the border and at ports of entry should concern anyone who values privacy. CBP has statutory authority to stop vehicles within 100 miles of the border, so even those people who don’t live near the border risk being exposed to CBP surveillance tools. In addition, the policies outlined in S.1757 will require DHS to collect more information about American citizens and foreigners alike, information that could be used for reasons other than border security.

Writing at the Niskanen Center, Samuel Hammond has some harsh words for libertarians. It’s a short step, he says, from anti-statism to some particularly ugly forms of nationalism:

The appeal of white nationalism to libertarian anti-statists should not be surprising. After all, nationalist and revanchist movements have historically represented powerful tools for mobilizing secession and other forms of political resistance to “the state.” Their common cause is all the stronger in multicultural, liberal democracies where ethnic grievances can be called upon to portray “the state” less as a political compact between competing groups, and more as tyrannical sovereign infringing on some sub-group’s right to self-determination.

To the extent that he’s right about this, that’s pretty embarrassing. Hammond cites AnCap YouTube to argue that there have been all too many who took this path. I’m not sure that it’s fair to judge anyone else by AnCap YouTube, although his judgment on some of them is certainly correct.

Other parts of his essay I think are quite wrong: It’s not necessarily crazy or evil to think that the state should be at least somewhat congruent to the nation. That proposition does not necessarily entail ethnonationalism, and certainly doesn’t when I assert it. A nation, as an imagined community, need not be ethnic at all. A pluralist nation may include people of many different ethnicities, religions, and other affiliations. The American nation has always been pluralist in its aspirations. Throughout our history we have increasingly delivered on the promise of pluralism, not just to favored groups, but to all. That work should continue, and if saying “you too are a part of this nation” can help with the task, then we should say it loudly and often.

Hammond also claims that “liberty needs the state.” On this point I am sure that the Niskanen Center will get the usual howls of protest from exactly the people who should be the least surprised. Of course the Niskanen Center would say something like this. But is it true?

It’s clearly correct to say, with Hammond, that in many cases “state sabotage automatically empowers the most dominant and dominating subgroups in our otherwise open society,” it’s much less clear that this must always be the case.

The way forward for radical libertarians and others who dream of a stateless (or just a less state-dominated) society consists of figuring out how to manage these tendencies toward domination, so that when the state does retreat, it is individual that liberty advances, rather than some other form of unjust domination.

I don’t know quite to what extent the project can succeed. But I think it’s reasonable to expect that we can enjoy a much smaller state than the one we have right now. Reasonably as well, this development could leave the vast majority of citizens, and particularly the least well off, better off by a range of widely acceptable criteria. What seems in order is not a broad declaration for or against the state, but a constant and relentless tinkering on the margins, with the aim of delivering less arbitrary domination of one person or group by another. Racial groups most certainly included.

To be blunt, Republicans are heading in the wrong direction on fiscal policy. They have full control of the executive and legislative branches, but instead of using their power to promote Reaganomics, it looks like we’re getting a reincarnation of the big-government Bush years.

As Yogi Berra might have said, “it’s déjà vu all over again.”

Let’s look at the evidence. According to The Hill, the Keynesian virus has infected GOP thinking on tax cuts.

Republicans are debating whether parts of their tax-reform package should be retroactive in order to boost the economy by quickly putting more money in people’s wallets.

That is nonsense. Just as giving people a check and calling it “stimulus” didn’t help the economy under Obama, giving people a check and calling it a tax cut won’t help the economy under Trump.

Tax cuts boost growth when they reduce the marginal tax rate on productive behavior such as work, saving, investment, or entrepreneurship. When that happens, people have an incentive to generate more income. And that leads to more national income, a.k.a., economic growth.

Borrowing money from the economy’s left pocket and then stuffing checks (oops, I mean retroactive tax cuts) in the economy’s right pocket, by contrast, simply reallocates national income.

Indeed, this is one of the reasons why the economy didn’t get much benefit from the 2001 Bush tax cut, especially when compared to the growth-oriented 2003 tax cut. Unfortunately, Republicans haven’t learned that lesson.

Republicans have taken steps in the past to ensure that taxpayers directly felt the benefits of tax cuts. As part of the 2001 tax cuts enacted by President George W. Bush, taxpayers received rebate checks.

The article does include some analysis from people who understand that retroactive tax cuts aren’t economically beneficial.

…there are also drawbacks to making tax changes retroactive. …such changes would add to the cost of the bill, but would not be an effective way to encourage new spending and investments. “It has all of the costs of the tax cuts but none of the economic benefits,” said Committee for a Responsible Federal Budget President Maya MacGuineas, who added that “you don’t make investments in the rear-view mirror.”

I’m not always on the same side as Maya, but she’s right on this issue. You can’t encourage people to generate more income in the past. If you want more growth, you have to reduce marginal tax rates on future activity.

By the way, I’m not arguing that there is no political benefit to retroactive tax cuts. If Republicans simply stated that they were going to send rebate checks to curry favor with voters, I’d roll my eyes and shrug my shoulders.

But when they make Keynesian arguments to justify such a policy, I can’t help but get upset about the economic illiteracy.

Speaking of bad economic policy, GOPers also are pursuing bad spending policy.

Politico has a report on a potential budget deal where everyone wins…except taxpayers.

The White House is pushing a deal on Capitol Hill to head off a government shutdown that would lift strict spending caps long opposed by Democrats in exchange for money for President Donald Trump’s border wall with Mexico, multiple sources said.

So much for Trump’s promise to get tough on the budget, even if it meant a shutdown.

Instead, the back-room negotiations are leading to more spending for all interest groups.

Marc Short, the White House’s director of legislative affairs, …also lobbied for a big budget increase for the Pentagon, another priority for Trump. …The White House is offering Democrats more funding for their own pet projects.

The only good news is that Democrats are so upset about the symbolism of the fence that they may not go for the deal.

Democrats show no sign of yielding on the issue. They have already blocked the project once.

Unfortunately, I expect this is just posturing. When the dust settles, I expect the desire for more spending (from both parties) will produce a deal that is bad news. At least for those of us who don’t want America to become Greece (any faster than already scheduled).

Republican and Democratic congressional aides have predicted for months that both sides will come together on a spending agreement to raise spending caps for the Pentagon as well as for nondefense domestic programs.

So let’s check our scorecard. On the tax side of the equation, we’ll hopefully still get some good policy, such as a lower corporate tax rate, but it probably will be accompanied by some gimmicky Keynesian policy.

On the spending side of the equation, it appears my fears about Trump may have been correct and he’s going to be a typical big-government Republican.

It’s possible, of course, that I’m being needlessly pessimistic and we’ll get the kinds of policies I fantasized about in early 2016. But I wouldn’t bet money on a positive outcome.

The annual Education Next gauge of public opinion on numerous education issues is out, and as always it offers lots to contemplate, including special questions this year on the “Trump effect.” I won’t hit everything, just what I see as the highlights.

School Choice

The poll’s headline grabber is a big drop in support for charter schools, public schools run by ostensibly private entities but subject to many public school controls, especially state standards and testing. When people with neutral opinions were removed, 52 percent of respondents approved of “formation” of charters—that word likely made some difference—down from a peak of 73 percent in 2012. With neutral answers included, only 39 percent of the general public supported charters.

The good news is that support for private school choice programs—superior to charters because they offer access to far wider options, including religious schools—saw upticks. Scholarship tax credits remain the choice champ, with support (absent neutral respondents) rising from 65 percent to 69 percent. With neutrals, support stood at 55 percent of the general public. For vouchers, a lot depends on question wording, but without a loaded emphasis on “government funds,” support (minus neutrals) stood at 55 percent, up from 50 percent the previous year. With neutrals, support was at 45 percent, with 37 percent opposing. Education savings accounts—basically, money parents can use not just for tuition, but other education expenses like tutoring or buying standalone courses—garnered only 37 support from the general public, but the concept is pretty new and people may just not have wrapped their heads around it yet.

Why the big drop in charter support but improved backing of private school choice? As always, wording, question order, and other artifacts of the poll itself matter, but assuming those aren’t the major causes of the results, perhaps the answer is that charters, as a compromise between empowering parents and maintaining government control, have traditionally tended to have the highest profile bipartisan support of the various choice mechanisms. As a result of Trump-driven polarization, perhaps they have also had the most visible schisms, maybe casting a more negative light on them. Or maybe people have started to perceive, as Education Secretary Betsy DeVos borrowed from Rick Hess to warn, charters are becoming “the Man” they were supposed to replace.


It is always instructive to see how much people think we spend on public schools, how much we actually spend, and whether the truth sets people free of the assumption that we need to spend more. As in the past, the poll finds that people greatly underestimate how much their districts spend per-pupil, with respondents guessing only about 69 percent of the actual expenditure. The average respondent thought their local district spent $8,877 per student, when in fact their district shelled out $12,899. When actual spending is not provided, 54 percent of respondents think their districts should increase outlays. When it is furnished, that drops to 39 percent. Of course, none of this actually says what the “right” amount of spending is, because we do not know what that is—we don’t have competitive markets that let us see how efficiently education can be provided.

Common Core

With the passage of the Every Student Succeeds Act, which at least in spirit puts states back in charge of their own standards and tests, and with many states likely loathe to quickly revisit the process of overhauling their systems, this once burning issue has dwindled to smoldering. Perhaps that is why after Core support cliff-dove between 2012 and 2016—with neutral responses removed, it plummeted from 90 percent to 50 percent—it caught a bit of an updraft this year, with support lifting to 52 percent. Add the neutral answers, though, and support sits at just 41 percent, and that’s with a loaded question that states the Core will be used “to hold public schools accountable for their performance.” (Who’s against “accountability”?) Much more disturbing, asked generically about uniform standards, federal tests, and letting parents opt their kids out of testing, the public overwhelmingly supports imposition. For educational freedom fans, perhaps the angel is in the details; once you move to real standards with concrete content, and real imposition, generic support turns to real world disapproval.

Agency Fees

From depressing news about imposition, we move to somewhat gratifying news. Agency fees are charges that teachers have to pay a union for bargaining on their behalf, whether they like that bargaining and union or not. These fees likely would have lost in the US Supreme Court in 2016 had it not been for the death of Justice Antonin Scalia. (A new case is on its way that might still strike down these enemies of free speech and association.) The public, it seems, might recognize the injustice of compelled support of hyper-politicized teacher unions, with 44 percent opposing agency fees, 19 percent neutral, and only 37 percent in support.

Affirmative Action

This became a hot topic a couple of weeks ago when it appeared that the Trump Justice Department would undertake investigations of colleges to see if they were discriminating against white applicants. The Trump effort appears now to be specifically connected to one suit by Asian Americans—not a broad hunt for discrimination against whites—but that could always change. Education Next only asked about minority preferences for hiring professors, not admitting students, and as I wrote recently I would encourage private colleges to undertake some affirmative action. That said, I still find the overwhelming result here—people want professors hired on merit, not goosed by considerations of race, gender, or political opinions—encouraging. It suggests that the public shares a crucial ideal: that people be treated as unique individuals, not members of a group.


As with many polls there is good news and bad news aplenty in this one, and lots to debate about question wording, order, and the why of the results. But the news is not bad for the best form of choice—private school choice—and some other important matters, so I’ll take it. Now go ahead and pore over the whole thing yourself!

Kenneth S. Rogoff stands out as the advocate of restricting hand-to-hand currency who has argued the case most comprehensively and probably the most cautiously. I critically reviewed his recent book, The Curse of Cash, in the July 2017 issue of Econ Journal Watch. Here I summarize highlights from my review, taking some passages verbatim. I encourage anyone who is interested to read the review in full. But in this piece I also comment on Rogoff’s response to my review that appeared in the same issue of EJW. When I provide page numbers for quotations, they reference Rogoff’s book; otherwise Rogoff quotations are from his response to my review.

A Case Against Cash

Rogoff does not propose eliminating all cash. In developed countries, he would phase out over a decade or more only large-denomination notes: in the United States, for instance, first $100 and $50 bills and then $20 bills and perhaps $10 bills. For small transactions, he would leave in circulation smaller-denomination notes, although he considers eventually replacing even these with “equivalent-denomination coins of substantial weight” to make them “burdensome to carry around and conceal large amounts” (p. 96). To put this in perspective, $1, $2, and $5 notes comprise less than 2 percent of the value of U.S. notes, or a little over 3 percent if we add in $10 bills. For less developed countries, Rogoff concedes that it is “far too soon” to “contemplate phasing out their own currencies” (p. 205).

Rogoff hedges his case with caveats and tries to address obvious objections. Yet like most proponents of phasing out cash, his argument is two pronged. Because cash is widely used in the underground economy, he believes that the elimination of large-denomination notes would help to significantly diminish such criminal activities as tax evasion, the drug trade, illegal immigration, money laundering, human trafficking, bribery of government officials, and even possibly terrorism. He contends that suppressing such activities would have the additional advantage of increasing tax revenue. The second prong relates to monetary policy. Rogoff believes that future macroeconomic stability requires that central banks have the ability to impose negative interest rates not only on bank reserves but on the public’s money holdings as well.

The Underground Economy

With respect to the underground economy, Rogoff’s Curse of Cash offers no genuine welfare analysis, considering the benefits as well as costs of the underground sector.[1] He gives little attention to the potential deadweight loss from forcing what is productive but unreported activity from a marginal tax rate of zero into marginal rates as high as 30 to 40 percent. Indeed, he disregards any gains to consumers of illegal drugs (except for an offhanded admission that legalization of marijuana may be a simpler approach for at least that part of the illegal drug trade) and exhibits scant concern for the welfare of illegal immigrants (despite his favoring increased legal immigration). The one transition cost that Rogoff tries to quantify is the impact on low-income individuals, recommending that the government provide at the very least about 80 million free, basic electronic-currency accounts, with a total bill of $32 billion per year.

When Rogoff gets to bona fide predatory acts within the underground economy, such as extortion, human trafficking, and violence associated with the drug trade, he descends primarily into lurid anecdotes. He fails to give even crude quantitative estimates to buttress his claim that eliminating cash would curtail these activities. As for corruption and bribery, Rogoff admits that they are really serious only in poorer countries—precisely where he also concedes that a premature elimination of cash would have dire economic consequences. In his discussion of terrorism, he admits that eliminating cash would have at best trivial impacts.

Approximately 50 percent of United States currency is held abroad. Yet Rogoff simply ignores negative effects on the nearly dozen countries that have completely dollarized (including Panama, Ecuador, El Salvador, several island countries in the Caribbean and Pacific) or another dozen partially dollarized (including Uruguay, Costa Rica, Honduras, Bermuda, the Bahamas, Iraq, Lebanon, Liberia, Cambodia, and Somalia). This is just another instance of Rogoff’s avoiding a complete cost-benefit analysis. As Pierre Lemieux, in his review of The Curse of Cash, succinctly puts it: “the economist venturing into normative matters would normally attach the same weight to a foreigner’s welfare as to a national’s.”[2]

Nor can Rogoff demonstrate any increased revenue for the U.S. government from phasing out large denomination notes. Relying on IRS estimates of the legally earned but unreported taxes in 2006 and extrapolating forward to 2015, he puts the potential gains to the national government at $50 billion annually (or less than 0.3 percent of GDP), along with approximately another $20 billion gain for state and local taxation. Yet his most comprehensive estimate of the seigniorage the government would lose from phasing out cash is $98 billion, or over 0.5 percent of GDP. Add to that the $32 billion annual cost of free electronic accounts for the poor, and Rogoff has failed to make a credible case that his proposal would create a net gain for the U.S. government, much less a net benefit for society overall.

True, other developed countries without a foreign demand for their currency have much lower rates of seigniorage than the U.S., and therefore government losses if those countries eliminated most cash would be less severe. The relative size of the underground economy in other countries, whether rich or poor, is also almost universally larger than in the United States. Indeed his cited estimates of the underground sector as a percentage of reported GDP for some of these countries—including developed countries such as Greece (25 percent), Italy (22.3 percent), Spain (19.6 percent), and Portugal (19.5 percent)—suggest that this is where a large fraction of their ordinary citizens live and survive. Rogoff states that the GDP “share of Europe’s shadow economy is more than double” (p. 63) that of the U.S., and he admits that this probably stems from higher tax levels and more burdensome regulation in Europe. But rather than reaching the obvious economic conclusion that the deadweight loss in Europe from inhibiting these activities would therefore be considerably larger than in the U.S., Rogoff merely touts “the benefits of phasing out paper currency” in Europe “in terms of higher tax revenues” (p. 89).

Negative Interest Rates

The second prong of Rogoff’s argument is that it would facilitate imposition of negative interest rates. The reasoning is as follows. When an economy sinks into a depression, the central bank should stimulate aggregate demand by lowering interest rates. But if interest rates are already extremely low, in what is alternately termed the ‘zero lower bound’ or a ‘liquidity trap,’ central banks are constrained in their ability to do this. Central banks can charge a negative interest rate on the reserves that commercial banks and other financial institutions hold as deposits at the central bank, and some are already doing so. If the monetary authorities push negative rates too far, however, the public can just flee into cash with its zero nominal return. Banks can also do the same by replacing their deposits at the central bank with vault cash. Elimination of cash would close off this way of avoiding negative rates, making negative rates truly comprehensive and effective.

The term ‘negative interest rates’ actually obscures somewhat the nature of what Rogoff contemplates. If negative rates can be extended to the general public, they in effect represent a direct tax on the public’s monetary balances, since most cash would be gone. Negative rates thus reverse the causal chain of traditional monetary theory, which focuses on the money stock. To the extent monetary expansion increases spending, it causes higher inflation with its implicit tax on money. Negative interest rates, in contrast, would explicitly tax money in order to cause increased spending with higher inflation.

I will not repeat here my extended discussion of why I believe that a policy of negative rates is not needed and would not work. Readers can find that in my review. Nor will I delve into one of the best parts of Rogoff’s book: his penetrating criticisms of fashionable alternatives for dealing with the zero lower bound, including a higher inflation target, forward guidance, fiscal policy, and dual-currency schemes. I simply point out that the problem vanishes once one thinks about monetary policy in terms of money rather than interest rates. Milton Friedman’s well-known thought experiment about a helicopter drop of money demonstrates this, as does Ben Bernanke’s writings on Japan’s experience with low interest rates, before he became chairman of the Federal Reserve.[3] If the economy needs monetary stimulation, the central bank through merely buying assets that genuinely increase the base of outside money can ultimately end up owning everything in the entire economy—except that sometime before it has done so, people will certainly start spending and drive up inflation.

Rogoff rejects this solution to the zero lower bound because he assumes it requires coordination with fiscal policy. But this assumption is simply wrong. Although we can imagine circumstances in which a desired monetary expansion would exceed the supply of Treasury securities available for open-market purchases, central banks can purchase and have purchased other financial assets or made other types of loans. The Fed has already purchased mortgage-backed securities, and other central banks have extended their acquisitions still more broadly, some even purchasing equities. Though far from ideal, such rare, limited, and temporary expansion of central-bank involvement in credit markets, if needed, would be less invasive than an untested, all-embracing money tax.[4]

The Public-Choice Dimension of a War on Cash

Not only are the positive arguments that Rogoff makes for confining currency to small denominations extremely weak, but his proposal also raises serious political-economy concerns that he hardly addresses and seems largely oblivious to. Even if the phasing out of all but small-denomination notes would accomplish what Rogoff has failed to convincingly substantiate, a marked reduction in crime, would it still be desirable? Not necessarily. Even when gains appear to be greater than losses, we should still hesitate about policies that punish or severely inconvenience the perfectly innocent. Lemieux has most trenchantly made this point:

Criminals are probably more likely than blameless citizens to invoke the Fifth Amendment against self-incrimination, or the Fourth Amendment against ‘unreasonable search and seizures.’ … But that is not a valid reason to abolish these constitutional rights.

These are necessary institutional constraints on State power, not just protecting the innocent but proscribing barriers that protect a free society more generally.

Indeed, one could argue that the underground economy is often a more effective check on government abuses than voting itself. Would alcohol prohibition in the U.S. have been repealed without widespread evasion by countless Americans? Would the U.S. be belatedly moving to marijuana legalization without the escape mechanism of the underground economy? Obviously none of these considerations excuse human trafficking and other forms of violence or brutality that are also within the underground economy. But one should be very cautious about drastic government impositions that indiscriminately impinge on almost the entire population, no matter how deplorable the outrages they are intended to diminish.

Rogoff’s faith in government is so strong that he evinces no discernible unease about possible abuse of his proposed interventions. Consider the battery of ancillary coercive regulations that he thinks might be vital to ensure the success of his proposal:

  • aggressive inducements to get people to turn in their cash (expiration date on large notes, restrictions on the maximum size of cash payments, and charges for very large deposits of small bills);
  • strict regulation of cryptocurrencies, such as Bitcoin and Etherium;
  • “more forceful steps…to pull the plug on money market funds” (p. 86);
  • lowering cash limits on anti-money-laundering regulations;
  • redoubling of “efforts to discourage” prepaid cards “as an alternative for moving large sums anonymously” (p. 97); and
  • banning “large-scale currency storage” or imposing “a tax on storage over a certain amount” (pp.160–161).

He goes on to predict that “[t]o the extent that new approaches to financial transactions are developed to evade government efforts to root out their sources, they will be met with a stiff hand” (p. 214). After all, “it is hard to stay on top of the government indefinitely in a game where the latter can keep adjusting the rules until it wins” (208). Rather than considering this government capacity a chilling concern, Rogoff enthusiastically embraces it.

Rogoff’s Response to My Review

Rogoff’s response to my review is quite respectful. He clearly wishes to encourage a civil dialogue on this question. Indeed, much of his response consists of amplifying details of his proposal. He does accuse me of “polemic exaggeration” because I titled my review “The War on Cash,” but that hardly seems unwarranted given that the title of his book is The Curse of Cash. More important, Rogoff’s response exhibits a shift in emphases in order to make his proposal appear still more tentative than in his book. Thus, he includes “many years of discussion and analysis” before any “advanced democracy is likely to start down the less cash-road.” And he pushes the “ultimate move to coins only (which I throw out as a very long-run idea …)” to “a time frame on the order of half a century or more.” He also shifts his geographical emphasis by conceding that

the case for pushing back on wholesale cash use is weaker for the United States than for most other countries, first because perhaps 40 to 50 percent of all U.S. dollars bills are held abroad, and second because the U.S. is a relatively high tax-compliance economy thanks to its reliance on income taxes for government revenue.

However these shifts introduce some additional tensions into Rogoff’s case. By admitting that phasing out cash is less of a priority for the U.S. than for other countries, especially those with high levels of tax evasion, he in essence is saying that his scheme is least needed where it is least onerous to implement and most needed where it is premature or dangerous to impose. After all, the most serious levels of tax evasion occur in less developed countries, such as Brazil and India. To be fair, Rogoff’s response still confines his focus to relatively advanced economies. He specifically mentions Greece and Italy, where he reports the underground economy as equaling about 25 percent of GDP. But phasing out large denominations in an economy in which unreported cash transactions lift the economy’s total output by as much as one-fourth strikes me as obviously drastic, even if the transition is slow.

By adding emphasis to how slowly he is willing to implement his proposal, Rogoff also undercuts the urgency he has attached to overcoming the zero lower bound, which in his book he characterized as having “essentially crippled monetary policy across the advanced world for much of the past 8 years” (p. 4). Indeed, if he is really willing to wait “at least a couple decades” for the phasing out of large denomination notes, why not just rely on market processes and technological innovations already in play to achieve a less coerced transition? Rogoff even predicts in his response that “the use of cash in the U.S. in legal tax-compliant transactions will be well under 5 percent ten years from now and probably only 1-2 percent twenty years from now, and that is assuming no change in government policy on cash [emphasis mine].”

Rogoff attempts to answer my charge that his welfare analysis fails to consider the economic benefits of productive underground activity by reiterating his speculation that “if the government is able to collect more revenue from tax evaders, it will … collect less taxes from everyone else” (p. 217). As he explained in his book, “if taxes can be avoided more easily in cash-intensive businesses, then too much investment will go to them, compared to other business that have higher pre-tax returns” (p. 59). This is of course correct as far as it goes. But it depends entirely on Rogoff’s expectation that any tax changes from eliminating large-denomination notes will be absolutely revenue neutral. This is a strong assumption seemingly at odds with the politics of taxation. Does he really believe that if phasing out cash brings in more tax revenue, governments are going to graciously reduce tax rates in order to ensure that the total bite out of the economy remains constant? This represents yet another instance of Rogoff assuming the perspective of a central planner and naively ignoring any public-choice considerations.

The simplistic assumption of revenue neutrality ignores a host of other complications as well. Recall that phasing out cash will reduce government seigniorage, so it is not clear how large the tax gains would be, if any at all, even for countries less reliant on that source of revenue. As for the Eurozone, The Curse of Cash (p. 84) cites estimate of seigniorage as high as for the U.S. Moreover, seigniorage, arising from an implicit tax on cash balances, already bears more heavily on underground, cash-intensive businesses. Phasing out cash not only changes both the level and type of taxation that these unreported, productive activities would pay, but also could subject them to burdensome regulation that imposes costs without generating revenue. This concern is particularly acute for countries like Greece and Italy. A genuine welfare analysis should carefully weigh all of these complications.

Regarding the 50 percent of U.S. currency held abroad, Rogoff repeats an assertion that he made in his book:

while there are many reasonable uses of the $100 bill abroad, it is indisputably popular with Russian oligarchs, Mexican drug lords, illegal arms dealers, Latin American rebels, corrupt officials, human traffickers, etc., and of course North Korean counterfeiters. In the book, I argue (conservatively) that foreign welfare should be thought of as a wash.

My review points out that Rogoff offers no quantitative evidence for this bold claim, and even if it were remotely close to accurate, it would still ignore the poor outside the United States who rely on dollars.

Given that we have only guesses based on anecdotes about alternative uses of dollars abroad, it certainly is appropriate for me to quote a contrasting view from a friend who read both my review and Rogoff’s response:

Based on my experience with overseas relatives $100 bills are also favored by ordinary citizens seeking a refuge from their own country’s unstable currency. They have no use for smaller bills, as they don’t use dollars for ordinary transactions. Dollars, for them, are a way of protecting their savings from the vagaries of the local currency. They aren’t familiar with all the denominations of US currency and would not be confident that smaller bills were genuine but they know what a $100 bill looks like and are comfortable with it.

Rogoff, however, remains willing to overlook welfare impacts on foreigners, whether they be potential illegal immigrants or overseas users of dollars. He declares:

The Federal Reserve and U.S. Treasury, not to mention Congressional decisionmakers, certainly do not directly take into account foreign welfare. The long-established approach to studying international trade and finance issues has always assumed that national authorities take into account national welfare, and that coordination and cooperation are needed to achieve a global social optimum. This is the right way to think about the problem, and my discussion is completely consistent with it.

Maybe for a politician pandering to voters but for an economist? This nationalistic bias is one of the critical flaws in Rogoff’s overall approach.


Rogoff raises many other interesting issues in his response, and trying to cover them all would make this article much too lengthy. His arguments are generally sophisticated and sometimes challenging, even when I disagree with him or believe he hasn’t adequately addressed my concerns. Our most fundamental difference remains our analysis of the State. Rogoff unreflectively adopts what Harold Demsetz characterizes as the “nirvana” approach to public policy. This makes him far more optimistic than is justified about the overall benevolence and competence of governments, particularly in developed countries. He thus oversells any advantages from his scheme and ignores or understates the myriad disadvantages. And it is he who bears the burden of proof for such an extensive reshaping of monetary systems, no matter how cautiously or slowly implemented.


[1] Larry White made this point with respect to advocates of eliminating cash generally in a previous Alt-M article.

[2] Pierre Lemieux, “Banning Cash: This Time is Not Different,” Regulation, 39 (Winter 2016-2017): 51

[3] Milton Friedman, “The Optimum Quantity of Money” in The Optimum Quantity of Money and Other Essays (Chicago: Aldine, 1970), pp. 1-67; Ben S. Bernanke, “Japanese Monetary Policy: A Case of Self-Induced Paralysis,” in Japan’s Financial Crisis and Its Parallels to U.S. Experience, ed. by Ryoichi Mikitani and Adam S. Posen, (Washington, D.C.: Institute for International Economics, 2000), pp. 149-66; and “Deflation: Making Sure ‘It’ Doesn’t Happen Here,” remarks before the National Economists Club, Washington, D.C., November 21, 2002; George Selgin anticipated Bernanke’s argument in a 1999 unpublished paper, “Japan: The Way Out,” reprinted at Alt-M.

[4] George Selgin has written on ways to make open-market operations more flexible and efficient using competitive auctions, see here and here.

[Cross-posted from Alt-M.org]

While same-sex couples ought to be able to get marriage licenses—if the state is involved in marriage at all—a commitment to equality under the law can’t justify the restriction of private parties’ constitutionally protected rights like freedom of speech or association.

Arlene’s Flowers, a flower shop in Richland, Washington, declined to provide the floral arrangements for the wedding of Robert Ingersoll and Curt Freed. Mr. Ingersoll was a long-time customer of Arlene’s Flowers and the shop’s owner Barronelle Stutzman considered him a friend. But when he asked her to use her artistic abilities to beautify his ceremony, Mrs. Stutzman felt that her Christian convictions compelled her to decline. She gently explained why she could not do what he asked, and Mr. Ingersoll seemed to understand.

Later, however, he and his now-husband, and ultimately the state of Washington, sued Mrs. Stutzman for violating the state’s laws prohibiting discrimination in public accommodations. The trial court ruled against Arlene’s Flowers on summary judgment. The Washington Supreme Court affirmed, holding that Mrs. Stutzman’s floral design did not constitute artistic expression worthy of First Amendment protection. Now the case is on the U.S. Supreme Court’s doorstep and Cato, joined by the Reason Foundation and Individual Rights Foundation, has filed an amicus brief urging the Court to take up the case and consolidate it with Masterpiece Cakeshop, the case of the similarly situated Colorado baker that the Court has already agreed to hear.

Although floristry may not initially appear to be speech to some, it’s a form of artistic expression that’s constitutionally protected. There are numerous floristry schools throughout the world that teach students how to express themselves through their work, and even the Arts Council of Great Britain has recognized the significance of the Royal Horticultural Society’s library, which documents the history, art, and writing of gardening. The Supreme Court has long recognized that the First Amendment protects artistic as well as verbal expression, and that protection should likewise extend to floristry—even if it’s not ideological and even if it’s done for commercial purposes.

The Court declared more than 70 years ago that “[i]f there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion, or force citizens to confess by word or act their faith therein.” W.Va. Board of Education v. Barnette (1943). And the Court ruled in Wooley v. Maynard—the 1977 “Live Free or Die” license-plate case out of New Hampshire—that forcing people to speak is just as unconstitutional as preventing or censoring speech. The First Amendment “includes both the right to speak freely and the right to refrain from speaking at all” and the Supreme Court has never held that the compelled-speech doctrine is only applicable when an individual is forced to serve as a courier for the message of another (as in Wooley).

Instead, the justices have said repeatedly that what the First Amendment protects is a “freedom of the individual mind,” which the government violates whenever it tells a person what she must or must not say. Forcing a florist to create a unique piece of art violates that freedom of mind. Moreover, unlike true cases of public accommodation, there are abundant opportunities to choose other florists in the same area.

Finally, granting First Amendment protection to florists would not mean that public-accommodation laws could provide no protection to same-sex couples. The First Amendment protects expression, which should include floristry but would not include many other businesses like caterers, hotels, and limousine drivers who are not in the business of creating artistic expression. These sorts of businesses may have other defenses available, constitutional or statutory, but that’s a different legal matter.

My recent blog post on the deaths and injuries caused by terrorists according to their motivating ideologies sheds some light on how frequent attacks like Charlottesville occur. I found that there were 3,350 total murders on U.S. soil caused by terrorists from 1992 through August 12, 2017. Of those, Islamists were responsible for 92 percent, Nationalist and Right Wing terrorists for about 7 percent, and Left Wing terrorists for less than one percent. The most common query after reading my post was: “What happens if you exclude deaths from the outlier attacks of 9/11 and the Oklahoma City bombing?” 

I originally did not exclude the deaths in these outlier attacks in my first post because I merely sought to describe who was killed and by whom. In response to that common question, I decided to post the results that exclude the outlier 9/11 and Oklahoma City attacks. Doing so changes the ratio of murders by ideology but it does not change which terrorism-inspired ideologies are the deadliest.

Table 1 subtracts the 2,983 deaths and 14,842 injuries caused by Islamist terrorists on 9/11 and the 168 deaths and 650 injuries caused by a Nationalist/Right Wing terrorist in the Oklahoma City bombing. Excluding the outliers reduces the total number of deaths by 94 percent from 3,350 to 199. The number of injuries also falls by 89 percent. Just two attacks account for nearly all of the deaths and injuries, though 9/11 was the bigger contributor. After removing the outlier deaths and injuries, Islamist-inspired terrorists are responsible for 52 percent of the murders and 78 percent of the injuries. That is a decline from my original post where I included 9/11 and found that Islamists are responsible for 92 percent of deaths and 94 percent of injuries. The relative percentage of murders committed by Nationalist and Right Wing terrorists rises from about 7 percent in my original post to 30 percent when the 9/11 and the Oklahoma City attacks are excluded. The deaths by Left Wing terrorists also grow in importance from less than 1 percent to 10 percent. 

Table 1

Deaths and Injuries in Terrorist Attacks by the Ideology of the Attacker Excluding 9/11 and Oklahoma City, 1992-2017.

LINK Excel.Sheet.12 "\\\\vmfile40\\cato$\\desktop\\anowrasteh\\Desktop\\Charlottesvile
Blog\\terror1.xlsx" Findings!R1C11:R6C15 \a \f 5 \h  \* MERGEFORMAT  










Nationalist and Right Wing





Left Wing










Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, and author’s calculations.

Figure 1 shows that Islamist terrorists killed 103 people while Nationalists and Right Wing terrorists killed 60. The number killed by Left Wing and Unknown/Other terrorists remained unchanged at 19 and 17, respectively. The injuries also drop dramatically (Figure 2).

Figure 1

Deaths in Terrorist Attacks by the Ideology of the Attacker Excluding 9/11 and Oklahoma City, 1992-2017.


Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, and author’s calculations.

Figure 2

Injuries in Terrorist Attacks by the Ideology of the Attacker Excluding 9/11 and Oklahoma City, 1992-2017.


 Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, and author’s calculations.

Two years ago, researchers at Duke University, drawing on a survey they conducted with police departments around the country through the Police Executive Research Forum, published a study on police perceptions of the domestic terrorist threat. It’s worth recounting the key findings:

Law enforcement agencies in the United States consider anti-government violent extremists, not radicalized Muslims, to be the most severe threat of political violence that they face.

They perceive violent extremism to be a much more severe threat nationally than the threat of violent extremism in their own jurisdictions.

And a large majority of law enforcement agencies rank the threat of all forms of violent extremism in their own jurisdictions as moderate or lower (3 or less on a 1-5 scale). 

The study looks at post-9/11 incidents and comes to conclusions comparable to a GAO study on the topic, commissioned by the bipartisan leadership of the Senate Homeland Security and Government Affairs Committee, earlier this year. Nearly a decade ago, a then-controversial DHS report on domestic extremism highlighted the potential danger for violent acts by white supremacist or neo-Nazi groups. 

My colleague Alex Nowrasteh has a very interesting and informative piece out today that goes into some depth about the relative threat from terrorists compared to other forms of violence. One point I would make is that the 9/11 attacks represent an anomaly in the overall picture because of the magnitude of the intelligence failure involved. As I’ve written previously, that foreign terrorist attack on America was entirely preventable. That’s not to suggest that Salafist terrorism does not pose a domestic threat; clearly it does. But the on-the-ground daily reality—as the studies cited above show—is that in post-9/11 America, the threat from white supremacists, “sovereign citizens,” and those professing similar views and acting on them is at least as great a threat as Salafist-inspired killers.

In the wake of the Charlottesville tragedy, the phrase “anti-government group” is likely to get tossed around rather carelessly, both in the media and by some in the advocacy community. Calling for a smaller federal government whose powers—especially surveillance powers—are reduced and properly controlled does not make one an “extremist.” Spewing racial hatred and committing acts of murder is the very manifestation of violent extremism, something all of us should condemn and oppose.


Last week, the Trump administration filed its merits brief in the Supreme Court case over his executive order suspending all travel and immigration from six African and Middle Eastern countries. On Twitter, President Trump has been insistent that the executive order is a “travel ban,” not some “politically correct term.” The statement shows that, while he is often difficult to understand, the president is actually very interested in how he brands his proposal. This fact matters because the constitutional case against the ban depends, in part, on Trump’s statements about it—specifically, the fact that he has repeatedly equated his current policy with his original proposal for a “Muslim ban.”

Beyond the lawsuit, however, it matters why the president has chosen to carry out certain proposals. If the president believes his travel ban will improve security by reducing Muslim immigration, then this is an important consideration for voters or anyone interested in influencing his policies in the future.

Trump’s Statements Equating the Muslim and Travel Bans

I reviewed the president’s comments about the ban—a list of which you can find below with fuller context—and found at least 12 statements where Donald Trump equated his plan to suspend immigration from certain countries with his original plan to ban all Muslims from entering the United States. I say at least because I have not watched all of his many rallies and have no access to his private correspondence. On another occasion, when asked after the election whether his plans to ban Muslims had changed, he reiterated that his plans on that subject were known. These dozen cases collectively demonstrate that President Trump understood his travel ban as a version of his Muslim ban.

Trump’s 12 statements occurred over a period of seven months from May 2016 to December 2016. They include nine separate situations and six direct denials to direct questions about whether the travel ban had changed his plans to ban Muslims. These statements occurred in various contexts, including private phone calls, written speeches, improvised speeches, interviews, and a debate. During this time, he described the travel ban as an “expansion” of the Muslim ban, a “bigger” version of the Muslim ban, and a “morphed” version of the Muslim ban.

Moreover, in these statements, President Trump explained exactly why his method of carrying out the ban changed. He specifically cited two reasons: the negative reaction to the outright Muslim ban and the constitutional concerns that others had expressed. However, he stated that for his part, he believed that the “Constitution does not give us the right to commit suicide,” a phrase used to express that although it may violate the Constitution, we should permit the violation to avoid a collapse of the entire society. Nonetheless, he said he was willing to acquiesce to others’ concerns.

The Evolution of the Ban

This list reveals the concept’s evolution. After defending the outright Muslim ban for six months, Trump called Rudy Giuliani in early May 2016 [1] to, as Trump himself put it, “look at the Muslim ban.” Giuliani explained that Trump told him, “Show me the right way to do it legally.” This indicates that Trump wanted Giuliani to come up with a version of the Muslim ban that would satisfy legal concerns. (Note that at this point, there is no other proposal for the “it” to be, Trump confirmed that he used the phrase “Muslim ban,” and grammatically, the antecedent to “it” is “Muslim ban” in Giuliani’s comments.) With these marching orders, Giuliani and House Homeland Security Committee Chairman Michael McCaul—with help from former Attorney General Michael Mukasey and Rep. Peter King—then sent a memo to the Trump team that explained why the outright ban could be unconstitutional and urged the adoption of a territorial-based ban.

No matter what these men thought about banning Muslims, Trump clearly saw this change as a reform to, not a rejection of, his Muslim ban. In June 2016, Trump detailed this new plan for the first time publicly [2]. He claimed that he was right to call for “a ban after San Bernardino” in December 2015—i.e. the Muslim ban—and that immigration laws give him the power to “suspend entry into the country of any class of persons that the President deems detrimental” and that he would use this power to “suspend immigration from areas of the world when there is a proven history of terrorism… until we understand these threats.”

Thus, the very first time he brought up the idea, the president both tied the two bans together and detailed—in a rare prepared, written speech—the exact legal strategy that he has used to implement them. Incredibly, the administration’s brief in the Supreme Court case actually cites this speech as proving that he did not want to ban Muslims. In a speech [3] and an interview [4] afterward, Trump explained that the “Muslim ban” or “temporary ban”—as he said he preferred to call the Muslim ban—would now apply to “in particular the terrorist states.”

It’s worth mentioning that this new territorial version of the Muslim ban actually resolves an important practical consideration that people, including Michael Mukasey, who was part of the Giuliani committee, had raised with Trump about the outright Muslim ban: it’s impossible to enforce a belief-based ban. Trump had previously claimed that the ban would only apply to those who responded “yes” to the question, “Are you a Muslim?” This is obviously a practical absurdity, but a ban on certain nationalities would be easy to enforce.

In a series of interviews on CBS [5][6], NBC [7][8], and Fox [9] that followed, he repeatedly denied that the territorial ban was a rejection of the Muslim ban in response to five direct questions, while insisting that his plans would now focus on “territory, not religion.” But he emphasized that he considered this “not a rollback,” but an “expansion” of the original Muslim ban [8] or a “bigger” version of the Muslim ban [9]. It was during this time that the president’s advisors drafted the executive order itself.

Then in another prepared speech in August, Trump explained that he would implement the new ban as part of “extreme vetting” where he would suspend entries from certain countries until he created a new vetting system for Muslims to screen out those “who believe that Sharia law should supplant American law.” During the presidential debate [10], when the moderator asked whether he had changed his position on the Muslim ban, he denied it again, saying that the “Muslim ban is something that in some form has morphed into extreme vetting for certain areas of the world.” She asked him again whether the “Muslim ban” still stands, and again, he flatly declared, “It is called extreme vetting” [11]. He continues to use this phrase “extreme vetting” to describe his Executive Order.

After the election, he reiterated his plan to suspend immigration from certain countries on “Day 1.” In December 2016 [12] a reporter directly asked him whether he had rethought his plan to “ban Muslim immigration” —yet again giving him the opportunity to say “yes, that plan is irrelevant to my current plans”—but instead, he said, “You know my plans all along. I’ve proven right.” His plans “all along” have been a Muslim ban with revisions to how it would be enforced. I could find no statement during this period where he denied that the travel ban was a version of the Muslim ban.

The Benefit of the Doubt

While some people may find ambiguity in one or two of these statements, their collective force matters more than any individual statement. Trump clearly wanted people to understand the travel ban as a version of the Muslim ban. Although Trump often shoots from the hip, he has carefully guarded the branding of the Muslim ban from the beginning. He’s made many other statements telling journalists how to frame this issue as well, as his Twitter comments show.

While Trump has since said that the travel ban is “not about religion—it’s about terror,” Trump repeatedly said the exact same thing about his outright Muslim ban, saying “it’s not about religion. It’s about security.” This means that to Trump, even a ban of an entire religion is not actually a ban about that religion. There is no doubt that the president believes that his travel ban would actually improve security. The question is whether he believes it for the same reason that he believed his Muslim ban would improve security—that it would lead to fewer Muslims entering the United States. His earlier statements directly indicate that this is the reason.

If Americans are to ignore the 12 statements, the president’s comments about Muslims in other contexts should provide some obvious evidence for the belief that he would not actually have favored a ban on Muslims (even if he said he did). But the evidence is almost entirely the other way. Trump has demonstrated repeatedly that his fears of Muslims lead him to believe even the most outlandish lies about them and suggest policies that specifically target them as a group.

In defense of the ban, Trump stated, “I think Islam hates us.” He repeatedly praised the idea of murdering Muslim prisoners of war with bullets dipped in pigs’ blood purely because it would be scary to other Muslims. He repeatedly and falsely claimed that “thousands and thousands” of Muslims in the United States cheered on 9/11. He said that the U.S. government should “shut down” mosques.

Even after his switch to the “territory ban,” he described Muslim immigration as “suicide” for the United States on at least two occasions. He called for indiscriminate surveillance of U.S. mosques and ethnic profiling of Muslims based on their religion. Without evidence, he described Muslim refugees to the United States as “people who believe that women should be enslaved and gays put to death.” He falsely said that Muslim assimilation is virtually nonexistent. He repeated the false claim about Muslims dancing on 9/11 even after it was debunked. He incorrectly said “the Muslim community” does not report terrorists. He falsely said that the wife of a speaker at the DNC Convention may have not been “allowed to speak” by her husband simply because they were Muslims.

On numerous occasions, Trump repeated a falsehood about howmany people” in the “Muslim communityrefused to turn in the San Bernardino shooters despite seeing “bombs all over their floor.” He has used this point constantly to defend the Muslim ban, travel ban, and extreme vetting, including during a presidential debate. Yet in fact, it was a non-Muslim man working in the area who witnessed the delivery of “numerous packages” and was suspicious but didn’t say anything.

The fact is that there is every reason to believe that Trump wanted to morph the Muslim ban into the travel ban to avoid potential legal problems and no reason not to.


The initial quotes about the outright Muslim ban provide context about how Trump discussed that ban. Note that Trump has said he preferred to use the phrase “temporary ban” to refer to the Muslim ban.

Statements on the Outright Muslim Ban

December 7, 2015: In a statement shortly after the San Bernardino terrorist attack:

Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States until our country’s representatives can figure out what is going on. According to Pew Research, among others, there is great hatred towards Americans by large segments of the Muslim population… . Mr. Trump stated, “Without looking at the various polling data, it is obvious to anybody the hatred is beyond comprehension. Where this hatred comes from and why we will have to determine. Until we are able to determine and understand this problem and the dangerous threat it poses, our country cannot be the victims of horrendous attacks by people that believe only in Jihad, and have no sense of reason or respect for human life. If I win the election for President, we are going to Make America Great Again.”

December 8, 2015: On MSNBC:

Geist: Donald, a customs agent would then ask a person their religion?

Trump: That would be probably—they would say, “Are you Muslim?”

Geist: And if they said, “Yes,” they would not be allowed in the country?

Trump: That’s correct.

December 12, 2015: On Fox News:

It’s a temporary ban, not on everyone, but on many… . We’re not insulting. This is about security. It’s not about religion. This is about security. We can’t allow people to come into this country that have horrible thoughts in their mind.

March 9, 2016: On CNN:

I think Islam hates us. There is something – there is something there that is a tremendous hatred there. There’s a tremendous hatred. We have to get to the bottom of it. There’s an unbelievable hatred of us… . we can’t allow people coming into this country who have this hatred of the United States and of people who are not Muslim.

May 11, 2016: On Fox News Radio (at 7:30):

We have a serious problem, it’s a temporary ban, it hasn’t been called for yet, nobody’s done it, this is just a suggestion until we find out what’s going on.

The Twelve Instances of Trump Equating the Muslim Ban and the Travel Ban

[1] 1. May 11, 2016: On Fox News:

I’m looking at it very strongly with Rudy Giuliani heading it. I’ve spoken to him a little while ago. We’re going to put together a group of five or six people. Very, very highly thought of people, and I think Rudy will head it up, and we’ll look at the Muslim ban or the ‘temporary ban’ as we call it … He will head it up and he’s agreed to do so.

January 29, 2017: On Fox News:

Jeanine Pirro: I want to ask you about this ban [the territory ban Executive Order] and the protests. Does the ban [the territory ban] have anything to do with religion? How did the president decide the seven countries? I understand the permanent ban on the refugees. Talk to me.

Rudy Giuliani: I will tell you the whole history of it [the Executive Order]. When he first announced it [the Executive Order], he said, ‘Muslim ban.’ He called me up. He said, ‘Put a commission together. Show me the right way to do it [the Muslim ban] legally.’ I put a commission together with Judge Mukasey, with Congressman McCaul, [Congressman] Pete King, whole group of other very expert lawyers on this. And what we did was, we focused on, instead of religion, danger—the areas of the world that create danger for us, which is factual basis, not a religious basis. Perfectly legal.

[2] 2. June 13, 2016: In a speech:

I called for a ban after San Bernardino, and was met with great scorn and anger but now, many are saying I was right to do so – and although the pause is temporary, we must find out what is going on. The ban will be lifted when we as a nation are in a position to properly and perfectly screen those people coming into our country. The immigration laws of the United States give the President the power to suspend entry into the country of any class of persons that the President deems detrimental to the interests or security of the United States, as he deems appropriate. I will use this power to protect the American people. When I am elected, I will suspend immigration from areas of the world when there is a proven history of terrorism against the United States, Europe or our allies, until we understand how to end these threats.

[3] 3. June 15, 2016: In a speech:

We have to stop on a temporary basis at least, but we have to stop people from pouring into this country until we find out what the hell is going on… . We don’t want to have these problems, and we’ve already got ’em. Look at this weekend. We don’t want to have these problems. So what I’m saying is it’s a temporary ban, in particular for certain people coming from certain horrible—where you have tremendous terrorism in the world. You know what those places are. But we have to put a stop to it. We have to put a stop to it until such time as we can figure out what is going on.

[4] 4. June 27, 2016: In an NBC phone interview:

Trump said his Muslim ban would apply “in particular [to] the terrorist states.”

[5] 5, 6. July 17, 2016: On CBS (at 13:52),

Lesley Stahl: In December, [Mike Pence tweeted], “Calls to ban Muslims from entering the U.S. are offensive and unconstitutional.”

Trump: So you call it territories. OK? We’re gonna do territories. We’re gonna not let people come in from Syria that nobody knows who they are. Hillary Clinton wants 550 percent more people to come in than Obama who doesn’t know what he’s—

[6]Stahl: So you’re changing your position.

Trump: No. Call it whatever you want. We’ll call it territories, OK?

Stahl: So not Muslims?

Trump: You know, the Constitution, there’s nothing like it. But it doesn’t necessarily give us the right to commit suicide, as a country, OK? And I’ll tell you this. Call it whatever you want, change territories, but there are territories and terror states and terror nations that we’re not gonna allow the people to come into our country. And we’re gonna have a thing called “Extreme vetting.” And if people wanna come in, there’s gonna be extreme vetting. We’re gonna have extreme vetting. They’re gonna come in and we’re gonna know where they came from and who they are.

[7] 7, 8. July 24, 2016: On NBC:

Chuck Todd: The Muslim ban. I think you’ve pulled back from it, but you tell me. You said, “Lastly and very importantly,” this is from your speech on Thursday night, “we must immediately suspend immigration from any nation that has been compromised by terrorism until such time as proven vetting mechanisms have been put in place.” This feels like a slight rollback.

Trump: I don’t think it’s a rollback

[8] Todd: Should it be interpreted as that?

Todd: I don’t think so. I actually don’t think it’s a rollback. In fact, you could say it’s an expansion. I’m looking now at territories. People were so upset when I used the word Muslim. Oh, you can’t use the word Muslim. Remember this. And I’m okay with that, because I’m talking territory instead of Muslim. But just remember this: Our Constitution is great. But it doesn’t necessarily give us the right to commit suicide, okay? Now, we have a religious, you know, everybody wants to be protected. And that’s great. And that’s the wonderful part of our Constitution. I view it differently. Why are we committing suicide? Why are we doing that? But you know what? I live with our Constitution. I love our Constitution. I cherish our Constitution. We’re making it territorial. We have nations and we’ll come out, I’m going to be coming out over the next few weeks with a number of the places.

[9] 9. On July 25, 2016: On Fox News

Hannity: What is your position? Because you were trying to explain yesterday [on NBC] that your position has not changed that you either vet them or they can’t get in.

Trump: No. I think my position’s gotten bigger now. I’m talking about territories now. People don’t want me to say Muslim. I guess I prefer not saying it, frankly, myself. So we’re talking about territories.

[10] 10, 11. August 15, 2016: In a speech:

I call it extreme, extreme vetting. …In addition to screening out all members of the sympathizers of the terrorist groups, we must also screen out any who have hostile attitudes toward our country or its principles or who believe that Sharia law should supplant American law. …To put these new procedures in place, we will have to temporarily suspend immigration from some of the most dangerous and volatile regions of the world that have a history of exporting terrorism.

On October 9, 2016: In a debate:

Moderator: Your running mate said this week that the Muslim ban is no longer your position, and if it is, was it a mistake to have a religious test?

Trump: …The Muslim ban is something that in some form has morphed into extreme vetting for certain areas of the world.

[11] Moderator: Why did it morph into that? Answer the question. Would you please explain whether the Muslim ban still stands?

Trump: It is called extreme vetting. We are going to areas like Syria.

[12] 12. December 21, 2016: In an interview:

Reporter: Have you had cause to rethink or reevaluate your plans to create a Muslim register or ban Muslim immigration to the United States?

Trump: You know my plans all along, and I’ve proven to be right, 100 percent correct.

Education scholars such as Richard Kahlenberg from The Century Foundation claim that since school choice programs “divert important resources away from the public schools,” children left behind in traditional public schools could be negatively impacted academically. However, a peer-reviewed study recently released by Temple University professor Sarah Cordes finds that charter school competition actually improves student achievement in nearby traditional public schools in the nation’s largest school district—New York City.

Specifically, Cordes finds that attending a traditional public school within a mile of a charter school in NYC increases student achievement in math and reading by about 0.015 standard deviations, or around 11 days of additional learning in both subjects. The detected effects increase with the proximity of the public charter school competition.

But why does this happen?

Residentially assigned public schools only lose funding if families are able to exit them for an alternative private or public educational option. If a traditional public school leader knows that their educational institution could be financially harmed by the choices of individual families, they will have a strong incentive to cater to the needs of their students. Since parents care about the academic success of their children, public school leaders will need to focus on turning educational resources into vital lifelong outcomes when faced with competitive pressures.

Although these findings may surprise those that listen to the frequent claims made by public education monopolists, they should not surprise social scientists. This study only adds to the abundance of the evidence existing on the topic that points in the same direction.

Prior Scientific Evidence 

As shown in Table 1 below, 23 of 24 such prior evaluations find that competitive pressures from private school choice programs improve the test scores of students left behind in traditional public schools. One study did not find any statistically significant competitive effects.

Table 1: Effects of School Choice Competition on Public School Test Scores 


Note: Green boxes indicate that the study found statistically significant positive effects on student test scores in traditional public schools. Yellow boxes indicate that no statistically significant effects were found.

Another peer-reviewed systematic examination of the scientific evidence finds the same conclusion: 20 of 21 reviewed studies indicate that private school choice programs improve the achievement of students that are left behind in their assigned public schools. No studies found negative effects.

Public school leaders that are currently able to compel families to pay for their educational services, nearly regardless of quality or price levels, have an obvious interest in preserving the existing public school monopoly. Rather than listen to the propaganda disseminated by those in power, we should embrace rational theory and the evidence produced by the only thing that can allow us to approach truth: the scientific method.

One person were murdered in a likely terrorist attack in Charlottesville, Virginia this Saturday when a suspected white nationalist named Alex Fields Jr. drove his car into a group of protesters. Prominent people on both sides of the political spectrum have condemned the politically motivated violence. However, some commentators have pointed out that left wing terrorists and rioters have also committed violence in recent years, but they have not provided any information to actually compare the violence of both sides. This blog fills that gap by describing terrorist murders by the political ideology of the perpetrators. The chance of being murdered in a terrorist attack is minor but there is wide variation by ideology.

Data and Methodology

This post examines 25 years of terrorism on U.S. soil from 1992 through August 12, 2017. Fatalities and injuries in terrorist attacks are the most important measures of the cost of terrorism. The information sources are the Global Terrorism Database at the University of Maryland and the RAND Corporation. Other organizations seem to count many religiously or racially motivated crimes as terrorist offenses, an overcounting that I attempted to avoid. I estimate the number of murders committed by terrorists in 2017 from online sources although they may be incomplete. As much as possible, I excluded terrorists who died or were injured in their attacks as they are not victims.

I grouped the ideology of the attackers into four broad groups: Islamists, Nationalists and Right Wingers, Left Wingers, and Unknown/Other. Global Terrorism Database descriptions of the attackers and news stories were my guide in organizing the groups by ideology. Islamists and unknown/other straightforward. Left Wing terrorists include Communists, Socialists, animal rights activists, anti-white racists, LGBT extremists, attackers inspired by Black Lives Matter, and ethnic or national separatists who also embrace Socialism. Nationalist and Right Wing terrorists include white nationalists, Neo-Confederates, non-socialist secessionists, nationalists, anti-Communists, fascists, anti-Muslim attackers, anti-immigration extremists, Sovereign Citizens, bombers who targeted the IRS, militia movements, and abortion clinic bombers. Some of the marginal attacks are open to reinterpretation but the ideology of the attackers by death and injury are straightforward in virtually all cases.


Terrorists have murdered 3,350 people on U.S. soil from 1992 through August 12, 2017 (Figure 1). Islamists committed 92 percent of all those murders and are, far and away, the deadliest group of terrorists by ideology. The 9/11 attacks accounted for 2,983 of the 3,086 Islamist-inspired terrorist deaths—an overwhelming 97 percent. The chance of being murdered in a terrorist attack committed by an Islamist during this period was about 1 in 2.5 million per year (Table 1). 

Nationalist and Right Wing terrorists are the second deadliest group of terrorists by ideology and account for 228 murders and 6.9 percent of all terrorist deaths. The chance of being murdered in a Nationalist or Right Wing terrorist attack was 1 in 33 million per year. The 1995 Oklahoma City bombing, the second deadliest terrorist attack in U.S. history after 9/11, killed 168 people and accounted for 74 percent of the murders committed by Nationalist and Right Wing terrorists. Left Wing terrorists killed only 19 people in terrorist attacks during this time but 15 since 2016. Nationalist and Right Wing terrorists have only killed 5 since then, including Charlottesville. Meanwhile, the annual chance of being murdered by a Left Wing terrorist was about 1 in 400 million per year. Regardless of the recent upswing in deaths from Left Wing terrorism since 2016, Nationalist and Right Wing terrorists have killed about 12 times as many people since 1992. Terrorists with unknown or other motivations were the least deadly. 

Figure 1

Murders in Terrorist Attacks by the Ideology of the Attacker, 1992-2017.

Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, and author’s calculations.

Table 1

Annual Chance of Dying in a Terrorist Attack by Ideology of Perpetrator, 1992-2017

Terrorist Ideology

Terrorism Deaths per Ideology

Annual Chance of Being Murdered



1 in 2,461,464

Nationalist and Right Wing


1 in 33,316,130

Left Wing


1 in 399,793,565



1 in 446,828,102



1 in 2,267,486

Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, United States Census, and author’s calculations.

The distribution of injuries committed by terrorists is similarly ideologically skewed (Figure 2). Attacks committed by Islamists are responsible for almost 94 percent of the 17,414 injuries during the entire period. Nationalist and Right Wing terrorists are responsible for 992 injuries, or 5.7 percent of the total. Left Wing terrorists are responsible for 27 injuries, or 0.16 percent of the total. Nationalist and Right Wing terrorists injured about 37 times as many people in terrorist attacks as Left Wingers did during this time. 

Injuries is a less clear category of damage that can range from a few scratches to amputations or brain damage. The annual chance of being injured in a terrorist attack does not reveal as much as your annual chance of being injured but I included it anyway (Table 2).

Figure 2

Injuries in Terrorist Attacks by the Ideology of the Attacker, 1992-2017.


Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, and author’s calculations.

Table 2

Annual Chance of Being Injured in a Terrorist Attack by Ideology of Perpetrator, 1992-2017


Terrorism Injuries per Ideology

Annual Chance of Being Injured



1 in 465,047

Nationalist and Right Wing


1 in 7,657,336

Left Wing


1 in 281,336,213



1 in 124,525,865



1 in 436,205

Sources: Global Terrorism Database at the University of Maryland, RAND Corporation, ESRI, United States Census, and author’s calculations.

The risk of being killed or injured in a terrorist attack on U.S. soil is small. However, a comparison to other intentional harms can put the risk in perspective. The chance of being murdered in a non-terrorist homicide from 1992 through 2017 was about 1 in 17,000 a year, which is about 133 times as great as being killed by terrorism during that time.

Islamism is an ideology created overseas, while much of the ideology that inspires Nationalist, Right Wing, and Left Wing terrorism is home grown or it has been here for so long that it might as well be.


Islamist terrorists are the deadliest since 1992. They killed about 13.5 times as many people as Nationalist and Right Wing terrorists who, in turn, killed about 12 times as many people as Left Wing terrorists did. The deadliness of terrorists by ideology has changed over time and will continue to do so. Charlottesville was a tragedy and the person responsible should be tried and, if convicted, punished to the fullest extent possible under the law. However, it is important to realize that the actual scale and scope of the recent terrorist threat differs significantly by ideology even though the annual chance of being murdered in such an attack is still small.



Timothy Carpenter and Timothy Sanders were convicted in federal court on charges stemming from a string of armed robberies in and around the Detroit area. They appealed on the ground that the government had acquired detailed records of their movements through cell site location information (“CSLI”) from their wireless carriers in violation of the Fourth Amendment. The U.S. Court of Appeals for the Sixth Circuit turned their appeal aside, finding that “[t]he government’s collection of business records containing these data … is not a search.”

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

In cases involving familiar physical objects, they usually do. In harder cases dealing with unfamiliar items such as communications and data, however, courts retreat to the Supreme Court’s “reasonable expectation of privacy” doctrine that emerged from Katz v. United States (1967). The Court has decided to review the important criminal-procedure and digital-privacy issues here.

Cato and the Competitive Enterprise Institute, joined by Reason Foundation and the Committee for Justice, filed an amicus brief urging the Court to return to the text of the Fourth Amendment. The reasonable expectation of privacy test is outdated because it lacks a strong connection to the text and asks courts to conduct a sociological exercise rather than a judicial one. This is especially true in the context of new technology, where societal expectations have not been fully formed yet and will change based on the Court’s judgment, leading to circular reasoning.

Courts have also used the ”reasonable expectation of privacy” test to undermine the very things the Fourth Amendment was designed to protect. For instance, dog sniffs looking for drugs have been said to not “compromise any legitimate interest in privacy” because they are only looking for contraband. But just because a search is designed to look for illegal activity doesn’t mean that the Fourth Amendment is inapplicable.

Likewise with the “third-party doctrine,” which holds that constitutional protections stop when protected information is shared.

The Carpenter case deals with information about a person’s location for more than 100 days, and yet the government claims that no privacy is violated when it seizes and searches that data. The Court should return to the text of the Fourth Amendment and recognize that data and digital communication are property that are protected by the papers and effects part of the Fourth Amendment, as it did in Riley v. California—the 2014 case where the justices unanimously required a warrant for searching a phone seized during an arrest.

Here, the government ordered the information on Mr. Carpenter’s location turned over (a seizure) and then processed that data for the location of the defendants (a search). The defendants had a contract with the phone company prohibiting the distribution of the data and the Court should recognize the property interest that the defendants had based on that contract.

In sum, the Fourth Amendment presumes that a warrant is required but for exceptional circumstances. There was no exigency that threatens the destruction of the data here, threat to officer safety, or any other reason that law enforcement officers could not get a warrant if they had probable cause. Focusing on the actual text of the Fourth Amendment demonstrates that the government’s actions here violated the Fourth Amendment.

The Supreme Court will hear Carpenter v. United States this fall.

The federal government spends more than $4 trillion a year on programs in hundreds of agencies. Which are the largest agencies, and how fast are they growing?

You can find out using the charting tool at DownsizingGovernment.org/charts. The tool plots spending on hundreds of federal agencies and programs in real, or inflation-adjusted, 2017 dollars. The charts cover 1970 to 2017, based on data from the 2018 federal budget.

The following are seven charts from the tool showing spending by the 21 largest agencies in order by size.

The first chart shows the largest departments: Defense, Health and Human Services, and Social Security. The three used to vie for top spot, but Defense has been left in the dust in recent years as the two entitlement-dispensing agencies have continued to grow. The federal government now has two $1 trillion agencies. Wow.

The second chart shows that spending by Veterans Affairs, Agriculture, and the IRS have soared in recent years. Veterans Affairs spending has doubled in a decade—again this is real dollars. Yikes. Agriculture spending includes food stamps and farm subsidies. IRS spending is fueled by outlays on “refundable” tax credits, particularly the earned income tax credit.

The third chart shows Education, Transportation, and the Office of Personnel Management. For the latter agency, spending includes the retirement and health costs of federal workers. Education spending gyrates widely because of recalculations in the costs of student loans. Transportation spending shows a solid upward trend, despite all the claims that we shortchange infrastructure investment. Either way, federal transportation spending should be cut.

The fourth chart shows Labor, Homeland Security, and Other Defense Civil Programs. The latter includes spending on military retirement and health care. The spike in Labor around 2010 was due to the extra UI benefits passed by Congress during the recession. Homeland spending spiked during the early Bush years and remains high.

The fifth chart shows that State department spending has tripled in constant dollars since 2000. HUD spending gyrates due to the accounting for housing finance subsidies. Justice spending tripled from 1990 to the mid Bush years. If you go to the chart tool, you can see that HUD subsidies for rental aid and community development remain at high levels.

The sixth chart shows Energy, NASA, and International Aid. The Energy spike from 2010-2012 stemmed from President Obama’s “stimulus” legislation. Remember Solyndra?

The seventh and final chart shows Interior, Commerce, and the EPA. The spikes in Commerce surround Census years. You can see this if you go to the charting tool, click open Commerce, and plot Census separately.

Similarly, use the chart tool to see that the Commerce spike in the late 1970s was for the Economic Development Administration, which by the way is one of the dumbest agencies in the government.

Finally, if you click open EPA on the charting tool, you can see that the spike in the late 1970s was due to a surge in grants to state governments.

Which of these departments and agencies should be cut? I suggest starting with these.