This post updates an October post with full-year and state data.

President Trump promised an “impenetrable” border wall. Trump built (with some help from Biden) about 450 miles of his wall, accounting for two-thirds of the border fence, and the remaining sections are the more recent designs that the Trump administration left because it largely approved of them.

Despite the construction of these exceedingly expensive new barriers, crossers are breaching the border wall more frequently than before, according to data obtained by the Cato Institute through a recent Freedom of Information Act request. In fiscal year 2022, the border wall was breached 4,101 times—more than 11 times per day. This was far more than the number of breaches in any of the prior six years and double the number of breaches for fiscal year 2016 before any of the Trump wall was built. In June, it was nearly 16 times per day, 113 times per week. Figure 1 shows the number of breaches by fiscal year. 

Every region of the border has seen a dramatic increase in border wall breaches. Figure 2 shows the number of border wall breaches by state. From FY 2016 to 2022, the number of breaches in California doubled from 713 to 1,403. Arizona saw a nearly sevenfold increase from just 94 to 674, and breaches also doubled from 997 to 2,025 in Texas and New Mexico (which are not listed separately).

These billion dollar walls are easily cut through with conventional power saws that cost as little as $100. Of course, these are just the breaches that have been discovered. Earlier this year, the Washington Post reported:

After smuggling crews cut through, they often disguise the breaches with tinted putty, making it difficult for agents to recognize which bollards have been compromised. The smugglers can return again and again to the site until the damage is detected, using the breach like a secret entrance. “They cut it with a fair amount of precision,” said one person with detailed knowledge of the sawing tactics who, like others, spoke on the condition of anonymity because he was not authorized to speak to reporters. “You have to look really closely to see it.”

As I have explained previously, the Trump border wall failed for all the predictable reasons. Immigrants used cheap ladders to climb over it, or they free climb it. They cut through small pieces and squeezed through, and they cut through big sections and drove through. In one small section in 2020, they sawed through at least 18 times that Border Patrol knew about in a month. They also made tunnels. Some tunnels were long, including the longest one ever discovered, but some were short enough just to get past the barrier.

Figure 3 shows the total miles of Trump border fencing compared to the number of known successful illegal entries. The number of successful entries—that Border Patrol saw—more than doubled after the Trump border wall was constructed. The fence has failed to accomplish its stated purpose of stopping illegal entries. Citing some easily surveilled minor gaps here and there in the Trump design as the reason for this failure is just sticking your head in the sand to avoid seeing all the smugglers sawing through it.

The border wall has been a $15 billion boondoggle. If a $15 billion project can be so easily defeated with $100 power tools and $5 dollar ladders, maybe the Border Patrol union was right the first time when it initially said more fencing wouldn’t work. No fair-minded advocate can say now that the project has provided any benefit whatsoever.

The United States is home to immigrants from every country in the world, but people are much more likely to leave some countries than others. The probability of a person from a given country immigrating to the United States can be broadly measured by dividing the number of immigrants in the United States by the country’s population plus the number of U.S. immigrants.

Figure 1 shows the origin countries for people with the highest likelihood of immigrating to the United States. The immigration data come from the U.S. Census Bureau’s American Community Survey as of July 1, 2021 and the country population data come from the United Nations. Immigrants are defined as anyone born abroad who was not a U.S. citizen by birth.

• The Marshall Islands tops the list with nearly 39 percent of its population in the United States. It is a former territory of the United States to which the United States still permits nearly free migration.
• Puerto Rico, which is still a U.S. territory with fully free migration between it and the mainland, is shown for comparison purposes. About 36 percent of people born in Puerto Rico (1.8 million) live in the U.S. states.
• Third is Dominica, the small Caribbean island, with 29 percent of its population in the United States.
• Guyana—the South American former-British colony where the official language is English—is fourth with 26.5 percent of its population living in the United States.
• Jamaica—another Caribbean island which also has English as its official language—is fifth with 22 percent of its population in the United States: 810,629 people.
• Then there is a quartet of smaller other Caribbean Island nations—Saint Kitts and Nevis, Grenada, Saint Vincent and the Grenadines, and Antigua and Barbuda with between 19 and 21 percent of their populations in the United States.
• El Salvador—the Central American country—is the most populous country in the top 10, but 18.4 percent of its population lives in the United States: 1.4 million people.

Other notable countries include the Dominican Republic and Cuba—both of which have seen more than 10 percent of their populations move to the United States. Strikingly, Mexico is by far the largest country in the top 20 with nearly 8 percent of its population living in the United States: 10.7 million people.

Table 1 shows the data for all 147 countries on which the American Community Survey collects information. The most unlikely immigrant in the world is from Indonesia. With a population of nearly 273 million 2021, the country was the country of origin for just 99,509 U.S. immigrants. Despite their 2.1 million immigrants, Chinese are actually quite unlikely to immigrate to the United States with just 0.15 percent of its population living as immigrants in the United States.

The number of immigrants (legal and illegal) from many of these countries has certainly risen since July 1, 2021. However, since the number of immigrants to countries other than the United States are not included, Figure 1 somewhat overstates the likelihood of immigrating to the United States in particular.

Regardless, the data indicate that many countries have withstood dramatic emigration to the United States. The experience of the Marshall Islands and Puerto Rico may indicate the upper bound for what would happen in countries with close ties to the United States over many decades if legal immigration were not restricted (more than a third immigrating). Surprisingly, however, several countries have sustained an emigration rate exceeding 10 percent of their populations without unrestricted immigration. Countries like Jamaica, El Salvador, the Dominican Republic, and Cuba give an indication of the scale that migration can take in those circumstances.

Over the summer, the Biden administration successfully slashed illegal immigration from Haiti by letting Haitians apply for asylum at legal crossing points along the U.S.-Mexico border. In November, following its decision to expel Venezuelans who cross illegally, it has also begun to let many Venezuelans enter legally as well—both at airports and at the land borders. It shows how the administration can easily end the border crisis by increasing legal migration.

The number of Venezuelans at legal crossing points—officially labeled “inadmissibles”—jumped to 6,817 from just 781 in October and 184 in September. Customs and Border Protection (CBP) processed more Venezuelans at ports of entry in November than it has since it introduced Title 42 in March 2020.

About 80 percent were processed away from the border at U.S. airports under the Process for Venezuelans parole (P4V) program that the administration created at the end of October. Under this process, Venezuelans with U.S. sponsors can fly directly from South America to the United States and request “parole” (which is a temporary, discretionary status—effectively a waiver of the normal restrictions on legal entry).

But the P4V program is not the only avenue opening up either. CBP also processed as many Venezuelans requesting asylum at southwest legal border crossings in November as it has during the entire pandemic. This process is not automatic. It requires connecting with a select group of nonprofit entities along the border, requesting that they ask CBP for an exception to Title 42 based on specific vulnerabilities and CBP must be willing to grant the request.

The new legal procedures have also helped increase the share of Venezuelans crossing legally from 1 percent In November to 50.3 percent in October. In other words, most are now entering legally. Although these new processes coincided with Venezuelans being subject to Title 42 expulsions to Mexico for the first time, the legal avenues are providing those who would otherwise be expelled a reason to wait for a chance to cross legally rather than trying to sneak in (as Title 42 has encouraged other nationalities to do).

The administration should go much further to reduce illegal immigration by Venezuelans. Nearly all Venezuelans in Mexico cannot access the P4V program because they entered Panama or Mexico illegally, and many others in South America have no one to sponsor them in the United States. It should open up processing at ports of entry—not require a nonprofit referral and not limit it only to certain immigrants deemed “vulnerable” in the subjective opinion of the nonprofit and CBP.

CBP almost completely eliminated illegal immigration by Haitians over the summer by vastly increasing their access to ports of entry to apply for asylum, and it virtually ended the flow of Ukrainians to the U.S.-Mexico border by granting them authorization to travel directly to the United States from abroad. CBP should expand these efforts to other nationalities, starting by processing the many Nicaraguans and Cubans who are seeking asylum right now.

Contrary to the confident-sounding claims of experts in the media, economists cannot accurately predict the macroeconomy. Economists have an awful record at forecasting inflation, interest rates, gross domestic product, and other macro variables. Below I excerpt from a Wall Street Journal column summarizing the failed record of forecasting 2022’s inflation, interest rates, and stock market.

This is important for public policy because if economists cannot predict the macroeconomy, then governments cannot successfully intervene and manipulate. The administration thought that its $1.9 trillion American Rescue Plan passed in March 2021 would aid the economy, but it proved to be hugely damaging by helping to spur high inflation.

Businesses and stock market investors often make mistakes, but they can change direction quickly as conditions change. The government, by contrast, is a rigid institution led by people who rarely admit mistakes. So when politicians move economic resources around, the resources usually get stuck in low‐​value uses for years on end, as discussed in this study on government failure.

The Wall Street Journal reports that while economists projected corporate profits accurately this year, they were far off in projecting the stock market, inflation, and interest rates:

The average forecast last December for this month’s interest rate was just 0.5%, according to Consensus Economics. The Fed this month raised rates to a range of 4.25% to 4.5%. Miss a change of this importance and there is little hope of getting anything else right.

Few even called the direction of stocks correctly. JPMorgan, Goldman Sachs and Citigroup were all bullish, expecting the S&P 500 to hit 5100, 5050 and 4900 respectively. Bank of America‘s strategists were rightly bearish, for the right reason, predicting a rates shock. But their 4600 target was a drop of just 3% from when they published their prediction. The S&P 500 closed on Friday at 3844.82, down 19% for the year so far.

Underlying all these errors, and those from pretty much everyone else, was the mistaken belief that inflation would quickly go away. Covid-related supply-chain problems would fade away, they thought, and falling inflation would allow the Fed to raise rates gently, sparing asset prices. Instead, inflation spread to virtually all categories of goods and services, worsened by the energy- and food-price spikes that followed Russia’s invasion of Ukraine.

If they cannot successfully manipulate the economy, then what should governments do? Economist Adam Smith advised them to adopt the “simple system of natural liberty.” By refraining from intervening:

the sovereign is completely discharged from a duty, in the attempting to perform which he must always be exposed to innumerable delusions, and for the proper performance of which no human wisdom or knowledge could ever be sufficient; the duty of superintending the industry of private people, and of directing it towards the employments most suitable to the interest of the society.

More on failed macro forecasting here, here, here, and here.

Late on the afternoon of December 23, the U.S. House of Representatives passed the approximately 1.7 trillion-dollar omnibus spending package, which now awaits President Biden’s signature. As is usually the case with such legislative monstrosities, the bill contains a few positive features and many negative ones. One positive part is that it repeals the so-called “X‑waiver” (an X is added to the DEA narcotics prescribing license), which the Drug Enforcement Administration has long required health care practitioners to apply for to prescribe the Schedule III opioid buprenorphine as medication-assisted treatment for substance use disorder. Inexplicably, health care practitioners have not needed an X‑waiver to prescribe the drug to treat pain. The X‑waiver requirement, which also limits the number of patients providers can treat, discouraged many practitioners from treating addiction.

Buprenorphine and methadone are both effective for treating opioid addiction, with some patients responding better to buprenorphine and others responding better to methadone. Unfortunately, the Drug Enforcement Administration regulates methadone outpatient treatment programs (OTPs) and requires most patients to take the methadone in the presence of clinic staff. Buprenorphine, on the other hand, is prescribed as a take-home medication by health care providers in their medical clinics.

In response to the COVID emergency, the DEA, in conjunction with the Substance Abuse and Mental Health Services Administration, permitted OTPs to disburse up to 28 days of take-home methadone to “stable patients” and 14 days of take-home methadone “less stable” patients. The success of the take-home program led the SAMHSA to extend it to one year past the date when the President declares an end to the COVID Public Health Emergency.

I and many others have called on Congress to end the DEA’s requirement that health care providers apply for an X‑waiver on their narcotics license to treat substance use disorder. By including the Mainstream Addiction Treatment Act (MAT Act) within the giant omnibus bill that Congress just passed, Congress finally removed the X‑waiver requirement and, along with it, any limits on the number of patients practitioners are allowed to treat. The act also makes it easier to prescribe buprenorphine using telemedicine.

Unfortunately, a fly in the ointment is that the omnibus bill also contained the Medication Access and Training Expansion (MATE) Act which, among other things, will require all health care practitioners who apply for or seek to renew their DEA narcotics prescribing license to take an 8‑hour government-approved course in understanding and treating addiction. This unnecessary burden might provide money-making opportunities to organizations that offer courses. Still, it may make many health care practitioners not renew or seek a narcotics prescribing license, making it even more difficult for patients in pain to get relief.

In a July letter to the Chair and Ranking Members of the Senate Health, Education, Labor, and Pensions Committee, the President of the American Medical Association rightly opposed the 8‑hour course requirement, stating:

There could be serious unintended consequences for patient care were many physicians to forgo registration with the DEA to prescribe controlled substances instead of taking the eight-hour training. This could have the opposite of the MATE Act’s intended effect by leading to fewer instead of more physicians and other health professionals having the ability to treat opioid and other SUDs with evidence-based medications. It could lead to less access for patients with psychiatric and other mental health conditions that are effectively managed with controlled substances having access to medically necessary care. There could be severe consequences for patients with painful conditions or postoperative pain obtaining effective pain control as far fewer physicians may be able to prescribe any scheduled medications for them. In addition, physicians who continue to be registered with the DEA could be inundated by patients who could no longer be treated by their usual source of care.

Many states, including my state of Arizona, impose their own course requirements on licensed health care practitioners who wish to prescribe controlled substances. In Arizona, practitioners must take a 3‑hour course every two years. The letter from the AMA mentioned this:

There is no evidence showing that a one-time mandated education requirement, as the MATE Act requires, will have the result that we all want, which is improving patient outcomes and stopping patients from dying from a drug-related overdose. In response to increasing numbers of people dying from a drug related overdose, beginning in 2016–2017, states began enacting mandates to restrict opioid prescribing, require use of state prescription drug monitoring programs and require prescribers to take specific continuing medical education (CME) classes. These policy interventions have not had a positive effect on reducing drug-related overdose or increasing access to evidence-based treatment for SUDs. This includes the 40 states with mandated CME requirements. A one-time training mandate for SUDs, no matter how well-intentioned, will not have meaningful impact on reducing drug-related overdose for the same reasons state-level CME mandates have not had the desired effect.

Sadly, but not uncharacteristically, lawmakers give with one hand and take away with the other.

On this day in 1783, George Washington appeared before Congress at the statehouse in Annapolis to resign his military commission at the end of the Revolutionary War. In so doing, he created the first of a series of precedents establishing the peaceful, voluntary transfer of power in the United States, along with his decision to not seek a third term as president in 1796 and, four years later, when John Adams became the first president to concede defeat for re-election.

The significance of the moment was not lost at the time. Washington’s former aide-de-camp, John Trumbull, wrote that his resignation “excites the astonishment and admiration of this part of the world. ‘Tis a Conduct so novel, so inconceivable to People, who, far from giving up powers they possess, are willing to convulse the Empire to acquire more.” Trumbull’s monumental painting of the occasion stands today in the rotunda of the Capitol.


On the other side of the Atlantic, even King George III later remarked that Washington’s action would “place him in a light the most distinguished of any man living” and make him “…the greatest character of the age.”


The idea that the highest power in the land would be voluntarily surrendered, subject to the rule of law, was unheard of at the time. Though the Founders had some contemporary examples of quasi-republican government, elective heads of state, and victorious generals after a revolution, these rulers typically held power for life. Figures such as the doge of Venice, the elective kings of the Polish-Lithuanian Commonwealth, and Oliver Cromwell after his victory in the English Civil War reigned until death or until forcibly overthrown. The only analogous history Washington’s contemporaries could find was in Cincinnatus, the Roman general and statesman who had likewise surrendered his military command and returned to civilian life more than two thousand years prior. It was an example he embraced and sought to emulate.


It is fitting that on this anniversary, Congress has adopted a crucial measure to ensure the peaceful transfer of power to a new commander-in-chief every four years. The Electoral Count Reform Act reflects a bipartisan, bicameral consensus to repudiate the lawless, unconstitutional attempts to overturn the results of the 2020 election. It is an exercise in lowercase‑c constitutional lawmaking, remedying the longstanding defects in the procedures for formally confirming who will become the next president.


Along with a broad cross-ideological coalition, my Cato colleagues and I have written extensively about the need to reform the Electoral Count Act of 1887, a muddled and unclear law. This arcane statute, little-noticed for over a century, opened the door to arguments that Congress could usurp the constitutional role of the states and the Electoral College in deciding who should hold the nation’s highest office. The updated law largely reflects our recommendations and scholarly consensus across the political spectrum. It confirms that the Constitution must be followed at each stage of the process, from the proper certification of presidential electors by the states through to the ceremonial role of the vice president during the electoral count on January 6.


Beyond the technical details, reforming the Electoral Count Act represents a commitment to abide by Washington’s example. No man is above the law, no man is indispensable, and no man rules for life.

Entertaining crackdowns on Americans’ voluntary use of software has been a theme of policy leaders’ response to fraud charges at centralized crypto exchange FTX. Following last week’s FTX hearing, Senate Banking Committee Chairman Sherrod Brown (D‑OH) floated the idea of “maybe banning” crypto. Earlier this month, Commodity Futures Trading Commission Chairman Rostin Behnam considered how the federal government might help offshore crypto exchanges that block U.S. users “protect those firewalls.”

Banning open-source software and barring the doors to the global Internet is no way to lead the free world. It shortsightedly undermines new ways of organizing civil society and commerce by prohibiting a technology that may prove key to developing robust cultural institutions over digital networks. Furthermore, it is inconsistent with American values, rejecting tools used to counter authoritarian overreach the world over. U.S. crypto policy must embrace, not eliminate, Americans’ freedom to write code, access information, and digitally organize.

A state-sanctioned informational firewall is a tool of authoritarian regimes, whereas cryptographically secure software is a tool of democratic civil society. To overcome state censorship and access dissident content, Chinese anti-lockdown protesters had to bypass “China’s Great Firewall” to reach American websites. Blockchain technology, by contrast, allowed protesters in Hong Kong to preserve the archives of a pro-democracy newspaper that authorities pressured to shut down.

While crypto has been used by repressive actors themselves, such as North Korean state-sponsored hackers, a blanket U.S. crypto ban would not distinguish between those using technology to subsidize human rights violations and those using it to combat them. One also suspects such a ban would do more to harm law-abiding software developers and civic innovators in the U.S. than to thwart the ambitions of terrorists and their state sponsors who respect no laws.

When it comes to the role that open-source and cryptographically secure networks will play in the future of civic life and geopolitics, one of the most formidable analyses comes from tech entrepreneur and investor Balaji Srinivasan’s The Network State.

In a book that combines technological forecasting, cultural criticism, and political theory, Srinivasan describes how social movements can leverage digital networks to organize online communities devoted to a shared belief, build new institutions, and, most provocatively, found new polities. Srinivasan defines these polities, or “network states,” as:

[S]ocial network[s] with a moral innovation, a sense of national consciousness, a recognized founder, a capacity for collective action, an in-person level of civility, an integrated cryptocurrency, a consensual government limited by a social smart contract, an archipelago of crowdfunded physical territories, a virtual capital, and an on-chain census that proves a large enough population, income, and real-estate footprint to attain a measure of diplomatic recognition.

Srinivasan sees crypto as key to this potential phenomenon because “only decentralized networks can give rise to network states.” Unlike centralized networks, decentralized ones provide exit rights that allow users to opt out of services while still holding onto their digital assets. In addition, integrated cryptocurrency blockchains provide a recordkeeping and payment system that resists tampering and censorship. Trustworthy records are key because, as Srinivasan argues, studying history is essential to making sense of and building in the present. Hence Big Brother’s eagerness to destroy or falsify the historical record in Orwell’s 1984. Moreover, by using smart contracts (self-executing software that can document individuals’ consent and limit administrators’ discretion), network states can incorporate the idea of the voluntary social contract.

The Network State makes bold claims and predictions, and Srinivasan is frank about the limits of projections and potential failures of both theories and technologies. Nonetheless, dramatic policy proposals—like banning crypto—require exploring broad implications.

While some might view the prospect of new network states competing with America as another reason to ban crypto, attempting such a ban could provide the ultimate justification for bringing such network states to fruition. (Srinivasan explores the related idea of a Bitcoin seizure triggering political conflict that contributes to the appeal of network states.)

But even if one fully accepts the possibility of network states, such states may bring about more benign (or salutary) impacts than existential risks: think chains of special economic zones in the developing world, as opposed to peer competitors seceding from or displacing the Union.

For that reason, one of the greatest lessons of The Network State for U.S. crypto policy is the potential benefits and vitality that a crypto ban would sap from America itself. Along the path to network states, Srinivasan traces intermediate outcomes short of sovereignty, including “network unions” that leverage crypto networks to help people organize around common goals.

There’s a version of this idea that should sound familiar to students of American history. Socially, network unions can be tools of Tocquevillian civil society of every stripe—from socially-conservative Benedict Option communities to libertarian Galt’s Gulches to progressive communes. Culturally, regardless of whether one plans to spend any time in the metaverse, an open-source version of the idea has far greater creative potential than a walled garden. Economically, network unions can form the bases for incorporating new real-world townships with zoning and other development reforms. (When combined with physical property, network unions become “network archipelagos” in Srinivasan’s phrasing.) Politically, network unions can facilitate organizing on any part of the ideological spectrum, allowing users to communicate with one another and track campaign goals using trustworthy channels and ledgers without relying on third-party applications.

Regardless of whether one buys into the potential of crypto, let alone network unions or network states, an outright crypto ban would destroy much of that possible upside without any guarantee of mitigating risks. By contrast, crypto policy tailored to relevant risks can help capture potential benefits from novel institutional designs while protecting against known downsides. Moreover, risk-based crypto policy, unlike a crypto ban, would be consistent with an American tradition of technological and cultural dynamism that we’ve struggled more to reclaim than lose in recent years. And, ultimately, a crypto-permissive policy stands against the censoriousness of Great Firewalls and for Americans’ ability to write, use, and organize with software as free people.