In Montana, legislators are considering an unusual elections proposal. In an effort seen as targeting incumbent Sen. Jon Tester (D), Senate Bill 566 would adopt a “Top Two” election system, but only for one specific election: the 2024 U.S. Senate contest. Under Top Two, all candidate regardless of party run in a so-called jungle primary. Only the top two candidates then advance to the general election in November, even if they’re both members of the same party.


Writing at FiveThirtyEight, Nathaniel Rakich pokes holes in the political theory underlying this move. It’s true Tester has won twice in the past on a plurality of less than 50%, with Libertarian candidates taking a few points in each of those elections. But it is mathematically unlikely he would have lost head-to-head. As Rakich explains, it can not simply be assumed that all of those third-party votes would have otherwise gone to the Republican candidate. Even if the larger share of them had, an implausibly lopsided supermajority would have been needed to change the results. In Tester’s most recent reelection in 2018, it’s a moot point because he won with over 50% of the vote. The partisan advantage provided by knocking spoilers off the ballot is often overstated. In reality, in most cases it wouldn’t matter even if the third party candidate “beats the spread,” that is, exceeded the margin of victory between the Republican and Democratic candidates. 


Political motives aside, Top Two is a terrible system, one of the few electoral reforms widely panned on both sides of the aisle. Top Two has been in place for several years in California and Washington state, and it’s recently under well-deserved fire in both. No other state has adopted it in well over a decade. There are good reasons it hasn’t caught on, even in an era of rising support for electoral reform.


Top Two was intended to to promote more competitive elections and encourage more moderate candidates. It has done neither. Instead, the system has produced only dysfunction and dissatisfaction. It has created manifestly undemocratic results, such as advancing two Republican candidates in a Democratic-majority district or vice versa. It also creates the unseemly spectacle whereby some voters go to the polls in November and see candidates from only a single party on their ballot, more reminiscent of elections in the Soviet Union than the United States. The number of spoiled ballots also increases exponentially under Top Two, with millions of voters refusing to mark either candidate in a one-party election. 


I wrote more about Top Two’s failures in California here. The Montana proposal is subject to the same critiques and others as well. Making such a change only for one election, targeting one particular incumbent, is an improper way to design election rules and procedures. As noted above, it probably would not accomplish its intended goal of defeating Tester by keeping third-party spoiler candidates out of the race. And it could backfire in dramatic fashion. The larger party in a state and the non-incumbent party (here, Republicans on both counts) will tend to attract more candidates to run in the primary. This results in a greater risk of the party fracturing and thus failing to advance any candidate to the general election. It puts parties in the untenable and often impossible position of trying to clear the field and discourage too many of their own candidates from running.


While Tester is not seen as likely to face a strong Democratic primary challenger, it’s worth considering the possibility. In a race where two Democrats are splitting, for example, 40% of the primary vote, and as many as six or seven Republican candidates are splitting their party’s 60%, it is entirely possible that two Democrats would take the top two spots. In November, voters would have no Republican option even though Montana voted for Trump over Biden by sixteen points, and even though more primary voters wanted a Republican candidate than a Democrat. Making sure there’s a strong second Democratic candidate in the race might even be a reasonable strategic move for Montana Democrats if Top Two passes. 


There are other options to address the problem of spoiler candidates. Runoff elections such as those used in Georgia are one possibility, as is ranked choice voting. But ultimately, electoral rules should not be bent to try to achieve a desired outcome in a specific race. Exploring alternatives to first past the post elections is a worthy goal, but Top Two is a step in the wrong direction.

This Cato study examined the role of startup businesses and the angel investors who fund them. It discussed how startups create jobs, generate innovations, and inject competition into markets.

Covering some of the same ground, a new piece in the Wall Street Journal by Christopher Mims discusses differences between large and small high-tech firms. Compared to large firms, small firms tend to have less bureaucracy, assume more risk, act more quickly, and may have better worker incentive structures.

The moment Noam Bardin, former chief executive of navigation app Waze, knew that life at a big company would be profoundly different from running a startup came soon after he sold his company to Google. ‘The first few weeks after the acquisition, we began dealing with the bewildering corporate bureaucracy,’ says Mr. Bardin. ‘What seems natural at a corporation—multiple approvers and meetings for each decision—is completely alien in the startup environment: make quick decisions, change them quickly if you are wrong.’

… big companies of every sort tend to give their employees incentives to be cautious rather than bold, to pursue overly complicated solutions rather than simple ones, and to seek promotions over serving the customer.

[The findings of a new study] show that when inventors join large firms, they get a pay bump, but they also produce fewer new innovations, relative to inventors hired by young firms.

… At big companies, people generate new ideas and get them in front of customers more slowly because of misaligned incentives, bureaucracy and institutional risk aversion, says Mr. Bardin. ‘The people who stay at a big company have to play the same games as everyone else, which means their innovative side doesn’t help them,’ he adds. ‘Their political side is what gets them promoted.’ People at big companies tend to have plenty of good ideas, he adds. The difficulty is with bringing them to fruition.

Both large and small businesses are crucial for a growing economy. We need an ecosystem, or spontaneous order, where both can thrive. What we don’t need is antitrust laws, which give policymakers the impossible task of central planning. What we don’t need is Federal Trade Commission chair Lina Khan deciding what sort of competition is “fair,” as she told Mims.

Rather, we need policymakers to avoid imposing barriers to startups and their financing. They should avoid regulatory burdens that hit small firms harder than large firms. They should repeal entry barriers such as occupational licensing and certificate of need laws. And they should cut America’s high capital gains taxes to support the flow of risk capital to startups.

More policy lessons for startups are discussed here and here.

Today Arizona Governor Katie Hobbs vetoed SB 1027, which would have placed a mandatory minimum of 10 years in prison on the first offense, and 15 years on the second offense, for anyone convicted of possessing, distributing, transferring, selling, or manufacturing heroin, fentanyl, or fentanyl analogs. In 2006 Arizona voters passed Proposition 301, imposing mandatory minimum prison sentences for possessing, transferring, selling, distributing, or manufacturing methamphetamine. This did nothing to decrease meth-related deaths. Meth-related drug deaths per 100,000 increased nationally by 1,500 percent between 2006 and 2021. From 2020 to 2021, meth-related deaths in Arizona increased by 33 percent. Why would lawmakers expect it to work differently with fentanyl?

No evidence exists that mandatory minimum laws deter the illegal drug trade or affect overdose deaths. And multiple studies show “no relationship between prison terms and drug misuse.” In 2020 Arizona ranked seventh among all states in prison population, with 526 prisoners per 100,000 population, and spent $30,000 per prisoner that year. Adding more people to the prison population will cost taxpayers money, with nothing to show for the millions spent.

Mandatory minimums will also make it easier for prosecutors to engage in coercive plea bargaining, which is responsible for more than 94 percent of state-level criminal convictions and disproportionately impacts people in lower socioeconomic groups who can’t afford to post bail or hire good lawyers.

As I testified to the U.S. House Judiciary Subcommittee on Crime and Federal Government Surveillance last month, harsher penalties—including death—will not deter the drug trade. Most drug dealers already factor these risks into their decision to get into the business and, correctly, have a greater fear of being killed by rival cartels and dealers than they fear federal or Arizona law enforcement.

I also told the Subcommittee about the “Iron Law of Prohibition” (a variant of what economists call the Alchian-Allen Effect): “The harder the law enforcement, the harder the drug.” Enforcing prohibition incentivizes those who market prohibited substances to develop more potent forms that are easier to smuggle in smaller sizes and subdivide into more units to sell.

During alcohol prohibition, bootleggers and dealers were not smuggling beer and wine but whiskey and other hard liquors. At football games, tailgaters drink beer and wine but smuggle flasks of hard liquor into stadiums that prohibit fans from bringing alcoholic beverages.

The Iron Law of Prohibition is why cannabis THC concentration has grown over the years. It is what brought crack cocaine into the cocaine market. And it made fentanyl replace heroin as the primary cause of overdose deaths in the United States.

Therefore, doubling down on penalties will do nothing to deter drug dealing and drug possession but will increase incentives to create new, more potent drugs. And it’s already happening: the veterinary tranquilizer xylazine is being added to fentanyl to boost its potency and ravage Philadelphia and other East Coast cities.

Doubling down on penalties will also undermine Arizona’s “Good Samaritan Law,” intended to make drug users unafraid of calling first responders if a person with whom they are doing drugs has an overdose. If mandatory minimums have any deterrent potential at all, it is the potential to deter people from calling 911.

The rising overdose rate understandably exasperates Arizona lawmakers. But they should think about unintended consequences before acting out of frustration. Repeating the same mistakes that drug prohibition enforcers have made for the last 50 years only worsens matters. If Arizona lawmakers want to take meaningful steps to reduce Arizona drug overdose deaths, they should remove government obstacles preventing harm reduction organizations from helping their neighbors, starting with the state’s drug paraphernalia laws. In vetoing this bill, Governor Hobbs didn’t allow anger and frustration to trump reason.

The Organisation for Economic Co-operation and Development (OECD)—an intergovernmental organization with 38 member countries—has published new data showing that the United States is losing the race for scientific talent to China and other countries. China’s strategy to recruit scientific researchers to work at China-affiliated universities is working.

In 2021, the United States lost published research scientists to other countries, while China gained more than 2,408 scientific authors. This was a remarkable turnaround from as recently as 2017 when the United States picked up 4,292 scientists and China picked up just 116. As Figure 1 shows, the rest of the OECD and China have both surpassed the United States for net inflow of scientific authors.

The OECD data are not measuring the movement of non-Chinese into China or non-Americans into the United States. The OECD tracks inflows and outflows of published scientific researchers based on changes in institutional affiliation. If an author who was previously affiliated with a different country publishes another article in a new country, the new country will be credited as receiving a new research scientist. The OECD credits more Chinese scientists returning to China for the sudden reversal in Chinese and American inflows.

This is a disturbing trend that started before the pandemic. In fact, it appears to coincide with the Trump administration’s “China Initiative”—more accurately titled the anti-Chinese initiative. Launched in November 2018, the Department of Justice’s campaign was supposed to combat the overblown threat of intellectual property theft and espionage. In reality, it involved repeatedly intimidating institutions that employed scientists of Chinese heritage and attempting malicious failed prosecutions of scientists who worked with institutions in China. U.S. Attorney Andrew E. Lelling has even admitted that the initiative that he helped lead “created a climate of fear among researchers” and now says, “You don’t want people to be scared of collaboration.”

If Chinese scientists are afraid to work in the United States, that means that the United States will not benefit from their discoveries as much or as quickly as China will. Although the Justice Department claims to have shut down its “China Initiative,” my colleagues doubt that Chinese scientists will be free from unjust scrutiny going forward. The U.S. National Institutes of Health is still bragging about having caused the firings of more than 100 scientists and shutting down research by over 150 scientists—over 80 percent of whom identify as Asian.

The administration continues to maintain contrary to evidence that Chinese industrial espionage—by scientists working in the United States—is a significant threat to the country. Universities and U.S. companies think the far greater threat is losing out on talented Chinese researchers. If the United States wants to deal a blow to the Chinese Communist Party, it should start by trying to fix the damage that it has done in the last few years and liberalize immigration from China.

There’s been quite an uproar over the past week regarding the latest Pentagon document leak. While much of the attention has been focused on material dealing with Ukraine, another U.S. ally is not happy about alleged American snooping on its internal affairs: South Korea.

Responding to the allegations, South Korean Democratic Party leader Lee Jae-myung said, “If it is true that they have spied on us, it is a very disappointing act that undermines the South Korea‑U.S. alliance, which is based on mutual trust.” 

What the South Korean minority party leader may not be aware of is that his own government has a history of spying on his U.S. counterparts in the American Congress. The evidence comes in the form of a partial declassified FBI memo dated December 5, 1975, obtained by Cato via a Freedom of Information Act request. 

The exact confidential source of the information remains classified nearly 50 years later, but the source was not only aware of FBI interest in the South Korean espionage and influence operations, but was seeking information from the Bureau as well–specifically the names and positions of the South Korean intelligence officers operating against Congress, suggesting the source may have been an intelligence or law enforcement officer from another agency or department.

The late Senators Strom Thurmond (R‑SC) and James Eastland (D‑MS) and Representative Otis Pike (D‑NY) were all involved in the drama, though to what degree is difficult to say given the still ridiculously high level of redaction in the five-page memo. The memo reveals the FBI’s Washington Field Office (WFO) was investigating at least some individuals mentioned by the confidential source. Whether any further action was taken by WFO in connection with this case is not known to the author. 

These episodes–old and new–are just reminders that allies spy on or try to influence each other nearly as much as they do on their adversaries.

Spending on the Supplemental Nutrition Assistance Program (SNAP) has doubled since 2019 to $127 billion in 2023. Such spending increases are raising federal budget deficits to unsustainable levels. Policymakers need to cut, and one place to start is programs that do not work as promised. With SNAP, the N is for nutrition, but SNAP households have less nutritious diets than other households.

The SNAP page at the U.S. Department of Agriculture (USDA) says, “SNAP provides nutrition benefits to supplement the food budget of needy families so they can purchase healthy food.” This USDA document promotes, “Increasing SNAP Benefits to Support Healthy Eating Patterns at All Life Stages,” and it mentions “nutrition” 199 times. A study by the Center on Budget and Policy Priorities says, “SNAP helped about 42 million low-income Americans afford a nutritious diet.”

However, SNAP households tend not to “purchase healthy food,” nor do they have “healthy eating patterns” or “nutritious diets.” A 2016 USDA study found that 23 percent of SNAP spending is on what can be called junk food.

More evidence comes from a 2018 USDA study. This chart from the study shows that nutritious foods make up a substantially smaller share of food acquired by SNAP households than other households.

The chart is based on a detailed USDA survey and analysis of Healthy Eating Index components for three groups: SNAP households, non-SNAP lower-income households, and non-SNAP higher-income households. The income level separating the two non-SNAP groups is 185 percent of FPG (federal poverty guidelines).

The analysis found that SNAP households receive about the same overall calories as other households, but:

Compared to lower income nonparticipants, SNAP-participating households acquired 31 percent fewer total vegetables, 40 percent fewer dark green vegetables and beans, 24 percent fewer whole fruits, 20 percent fewer whole grains, and 27 percent fewer seafood and plant proteins for every 1,000 calories acquired. They also acquired almost 6 percent (33.05/31.09) more empty calories as a share of total calories.

SNAP households bought less healthy foods than other lower-income households. The USDA study concludes, “Compared to the SNAP-nonparticipating subgroups, SNAP-participating households purchased foods of lower quality overall.” The study only examined correlations, not causations, but the results are similar to other studies: National Health and Nutrition Examination Survey data “also show that diet quality among SNAP participants is lower than among nonparticipants.”

Advocates often say that SNAP households have poor diets because they do not have nearby access to healthy foods. But the 2018 USDA results do not support that theory: “Further, on all of our measures of nutritional quality, SNAP-participating households with low household-level access to food stores did not differ from SNAP-participating households with better access.”

Policy Implications

The USDA says that “Poor nutrition is a leading cause of illness in the United States, associated with more than half a million deaths per year.” Yet the USDA’s largest food program, SNAP or food stamps, delivers poor nutrition. This is a major policy failure, especially since SNAP costs taxpayers an enormous $127 billion a year.

SNAP is part of the Farm Bill scheduled to be reauthorized this year, and so policymakers should explore ways to cut the program. One option is to replace the $127 billion program with a “Fruits and Vegetables Program,” which would cost $20 to $30 billion a year (see data note below). Nutritionists disagree about some foods, but they all agree that fruits and vegetables are essential. Taxpayers would save about $100 billion a year replacing SNAP with an F&V program.

Another option is to devolve SNAP to the states and let them decide how to reform the program. State policymakers are better placed to decide on the level of benefits and foods to support. The states alongside the private sector have more flexibility to find efficient solutions.

Unfortunately, SNAP’s nutrition failure has so far only prompted the usual response to federal failure—add more programs! Congress has added SNAP-Ed and the Gus Schumacher Nutrition Incentive Program to promote healthy food choices. These add-on programs cost taxpayers about $550 million a year. Last year, the Biden administration issued a “National Strategy on Hunger, Nutrition, and Health,” which proposed new subsidies and regulations to induce SNAP recipients and others to eat healthy.

But rather than building the federal food bureaucracy larger, policymakers should replace SNAP with a smaller Fruits and Vegetables Program or devolve it to the states for more innovative policy solutions.

For more on SNAP, see here and here.

Data Note: the 2016 USDA study found that fruits and vegetables were 14 percent of SNAP benefits, which would be about $18 billion in 2023. Replacing SNAP with a Fruits and Vegetables Program would likely increase consumption of those healthy items, perhaps to $20-$30 billion a year.

Concerns have been spreading widely that the Federal Reserve’s new payments system, FedNow, was a covert plan to launch a central bank digital currency, or CBDC. While it’s important to remain vigilant, it’s equally important to get the facts right: FedNow is not a CBDC.

Before defending the Federal Reserve (a rare position for myself) and explaining what’s happening, let me take this opportunity to note that much of the concern here is a direct result of how the Federal Reserve has become infamous for clouding its messaging to the public. To make matters worse, the Federal Reserve has consistently preserved a legal gray area with its authority to issue a CBDC. With these considerations in mind, it should be no surprise that the public’s trust in government is at historic lows and many have worried that something more sinister is at work.

With that said, let’s dive into what exactly FedNow is (and isn’t).

What is FedNow?

FedNow is an instant payments system—a sort of update to Fedwire and the Automated Clearinghouse (ACH). Individuals will not have direct access to FedNow, but they will have access to faster payments so long as their bank or credit union opts into the FedNow network. Although creating FedNow was not necessary to achieve faster payments, one big difference with FedNow will be that payments will no longer be held up on weekends, holidays, or after traditional business hours.

FedNow is Not a CBDC

Astute eyes will likely recognize that FedNow does vaguely resemble a wholesale CBDC. Where a wholesale CBDC would be restricted to financial institutions for use during interbank settlement, FedNow would also be restricted to financial institutions. The difference, however, lies in their design. Where a CBDC is a currency, FedNow is a payment rail. If we think of dollars and cents as water, then FedNow is the plumbing that gets those dollars and cents where they need to go. In contrast, a CBDC would involve replacing the water itself in this analogy.

Under the current system, interbank settlement is performed on the Federal Reserve’s payment rails, thus ultimately affecting retail banking customers’ settlement times. It’s for this reason that Federal Reserve Governor Michelle Bowman said, “My expectation is that FedNow addresses the issues that some have raised about the need for a CBDC.” This statement should not be misunderstood to say that FedNow will take CBDCs off the table, but it does show that the Federal Reserve itself sees FedNow and CBDCs as distinctly different.

Not Without Objection

Although FedNow is not a CBDC, that’s not to say there is no reason to object to it. Faster payments are needed in the United States, but FedNow is not the only option. FedNow was announced in 2019, but that was two years after the Clearing House (TCH) introduced the Real-Time Payments (RTP) Network. In fact, while not offering instant payments, even just expanding the operating hours of Fedwire and the National Settlement Service (NSS) to run 24x7x365 would have improved the U.S. financial system. Yet, the Federal Reserve seems to have chosen to ignore the simpler option and walk over the private sector. Both choices are grounds for fair and longstanding objections to FedNow.

Staying Vigilant

While FedNow is not a CBDC, that doesn’t mean those concerned about CBDCs should let their guard down. Rather, staying vigilant means working to identify issues as they come up and working to get the facts right. From ending financial privacy to destabilizing the banking system, the risks of CBDCs are all too real.



Are you interested in learning more about CBDCs? Check out this new report that explains why the benefits of CBDCs are a myth, but the risks are serious. Or, if you want something lighter to get up to speed, check out this introductory webpage.