“Can America Win the AI War with China?”

A long video chat, with Geoffrey Cain, who is more hawkish than I am.  Bari Weiss moderates.  One argument I make is that America may prefer if China does well with AI, because the non-status quo effects of AI may disrupt their system more than ours.  I also argue that for all the AI rival with China (which to be sure is real), much of the future may consist of status quo powers America and China working together to put down smaller-scale AI troublemakers around the rest of the world.  Interesting throughout.

Three Simple Principles of Trade Policy

Are we in a trade war today? Who knows? Doesn’t really matter. It’s always a good time to review important principles. A good source is Doug Irwin’s Three Simple Principles of Trade Policy published in 1996. Below I have updated occasionally with more recent data.

Principle 1: A Tax on Imports is a Tax on Exports

Exports are necessary to generate the earnings to pay for imports, or exports are the goods a country must give up in order to acquire imports….if foreign countries are blocked in their ability to sell their goods in the United States, for example, they will be unable to earn the dollars they need to purchase U.S. goods.

…The equivalence of export and import taxes is not an obvious proposition, and it is often counterintuitive to most people. Imagine taking a poll of average Americans and asking the following question: “Should the United States impose import tariffs on foreign textiles to prevent low-wage countries
from harming thousands of American textile workers?” Some fraction, perhaps even a sizeable one, of the respondents would surely answer affirmatively. If asked to explain their position, they would probably reply that import tariffs would create jobs for Americans at the expense of foreign workers and thereby reduce domestic unemployment.

Suppose you then asked those same people the following question: “Should the United States tax the exportation of Boeing aircraft, wheat and corn, computers and computer software, and other domestically produced goods?” I suspect the answer would be a resounding and unanimous “No!” After all, it would be explained, export taxes would destroy jobs and harm important industries. And yet the Lerner symmetry theorem says that the two policies are equivalent in their economic effects.

Exports and imports rise and fall together. It is surely obvious that if you want more imports you must export more (barring a bit of borrowing see below). The same thing is true in other countries. As a result, it is also true that when you import more you export more.

Principle 2: Businesses are Consumers Too

Business firms are, in fact, bigger consumers of imported products than are U.S. households.

As of 2024, more than 64% of imports are intermediate products. See here for the data.

By viewing imports not as final consumer goods but as inputs to U.S. production, policy makers can more clearly recognize that the issue is not so much one of “saving” jobs but of “trading off’ jobs between sectors. This brings home forcefully the most important lesson in all of economics-there is no such thing as a free lunch. Every action involves a trade-off of some sort. Higher domestic steel prices help employment in the steel industry but harm employment in steel-using industries. Higher domestic semiconductor prices help employment in the semiconductor industry but harm employment in semiconductor using industries. As john Stuart Mill wrote in 1848 in the context of import protection, “The alternative is not between employing our own country-people and foreigners, but between employing one class or another of our own country-people.”

Principle 3: Trade Imbalances Reflect Capital Flows

There is a fundamental equation of international finance that relates this net borrowing and lending activity to the current account. The equation is:

Exports – Imports = Savings – Investment

The powerful implication of this equation is that if a country wishes to reduce its trade deficit, the gap between its domestic investment and its domestic savings must be reduced.

A country’s trade balance is related to international capital flows–not with open or closed markets, unfair trade practices, or national competitiveness. If a country wants to solve the “problem” of its trade deficit, it must reverse the international flow of capital into its country. In many cases net foreign borrowing can be reversed by reducing the government fiscal deficit. [emphasis added, AT]

Doug concludes:

These three simple principles of trade policy…[have] stood the test of time, they come as close to truths as anything economists have to offer in any area of policy controversy. Yet they are routinely denied, explicitly or implicitly, in trade policy debates in the United States and elsewhere. I do not imagine that a greater appreciation of these principles would invariably bring about more liberal trade policies; I offer them, rather, in the more modest hope that they might lead to sounder debates in which the real consequences of government policies are confronted more seriously than at present.

Hat tip: Erica York.

What should I ask Sheilagh Ogilvie?

She is a Canadian economic historian at Oxford, here is from her home page:

I am an economic historian. I explore the lives of ordinary people in the past and try to explain how poor economies get richer and improve human well-being. I’m interested in how social institutions – the formal and informal constraints on economic activity – shaped economic development between the Middle Ages and the present day.

And:

My current research focusses on serfdom, human capital, state capacity, and epidemic disease. Past projects analysed guilds, merchants, communities, the family, gender, consumption, finance, proto-industry, historical demography, childhood, and social capital. I have a particular interest in the economic and social history of Central and Eastern Europe.

Here is her Wikipedia page.  Her book on guilds is well known, and her latest is Controlling Contagion: Epidemics and Institutions from the Black Death to Covid.  Here are her main research papers.

So what should I ask her?

Deep Research considers the costs and benefits of US AID

You can read it here, summary sentence:

Based on the analysis above, the net assessment leans toward the conclusion that USAID’s benefits outweigh its costs on the whole, though with important qualifiers by sector and context.

Here is a useful Michael Kremer (with co-authors) paper.  Here are some CRS links.  Here is a Samo analysis.  AID is a major contributor to the Gavi vaccine program, which is of high value.  The gains from AID-supported PEPFAR are very high also.

To be clear, I consider this kind of thing to be scandalous.  And I strongly suspect that some of the other outrage anecdotes are true, though they are hard to confirm, or not.  How about funds to the BBC?  While the “Elonsphere” on Twitter is very much exaggerating the horror anecdotes and the bad news, I do see classic signs of “intermediaries capture” for the agency, a common problem amongst not-for-profit institutions.

The Samo piece is excellent.  For one thing he notes: “The agency primarily uses a funding model which pays by hours worked, thus incentivizing long-duration projects.”  And the very smart Samantha Power, appointed by Biden to run AID, “…is in favor of disrupting the contractor ecosystem.”  Samo also discusses all the restrictions that require American contractors to be involved.

Here is a study on how to reform AID, I have not yet read it.

Ken Opalo, in a very useful and excellent post, writes:

For example, in 2017 about 60% of USAID’s funds went to just 25 American organizations.  Only 11% of U.S. aid goes directly to foreign organizations. The rest gets management via U.S. entities or multilateral organizations. This doesn’t mean that the 89% of aid gets skimmed off, just that an inefficiently significant share of the 89% gets gobbled up by overhead costs. In addition, this arrangement denies beneficiaries a chance at policy autonomy.

According to the very smart, non-lunatic Charlie Robertson:

My data suggests US AID flows in 2024 were equivalent to: 93% of Somalia’s government revenues, 61% in Sudan, just over 50% in South Sudan and Yemen

While I do not take cutting off those flows lightly, that seems unsustainable and also wrong to me as a matter of USG policy.  Those do not seem like viable enterprises to me.

There are various reports of AID spending billions to help overthrow Assad.  I cannot easily assess this matter, either whether the outcomes was good or whether AID mattered, but perhaps (assuming it was effective) such actions should be taken by a different agency or institution?

While US AID appears to pass a cost-benefit test, it does seem ripe for reform.  Based on what I have read and heard, I would focus all the more on public health programs, and forget about “trade promotion,” “democracy promotion,” and more.  I would get rid of virtually all of the consultants, and make direct transfers to worthy African and Ukraine programs, thus lowering overhead.  If such worthy programs exist, why not give them money directly?  Are they so hard to find?  And if so, how trustworthy are these intermediaries really?  What are they intermediating to?

So a housecleaning is needed here, but the important sources of value still should be supported.

I wonder if he enjoyed our Markets in Everything series?

Now he is in it:

A former senior adviser to the Federal Reserve Board of Governors was arrested Friday and accused of leaking inside information from the Fed to the Chinese government over a period of several years, at one point receiving a $450,000 payment, and then lying about it to Fed investigators.

Economist John Harold Rogers, 63, of Vienna, Virginia, worked in the Division of International Finance of the Fed from 2010 until 2021, according to an indictment unsealed Friday in federal court in the District. Last year, he told a podcaster that he had retired from the Fed in May 2021, approximately a year after he had been questioned by investigators for the Fed’s inspector general and allegedly lied about how he accessed and transmitted sensitive information to two unnamed Chinese co-conspirators.

After leaving the Fed, Rogers moved his family to Shanghai and began working as a professor at Fudan University, according to comments he made to the EconVue podcast last year and posted in online biographies.

Here is the full story.

Wednesday assorted links

1. Douglas Irwin on dismantling the Indian license raj.  And Indians praying for visas under Trump.  And podcast on Indian biotech potential.

2. New data on LLMs refereeing economics papers.

3. Do LLMs use trigonometry to do addition?

4. Is Occam’s Razor obsolete?

5. Oliver Kim on exchange rates.

6. New asteroid risk.

7. Heat and economic decision-making.

8. Czech beaver DOGE.

9. Zvi on Deep Research.

10. Is the Peter Principle actually true?

Trumpian policy as cultural policy

The Trump administration has issued a blizzard of Executive Orders, and set many other potential changes in the works.  They might rename Dulles Airport (can you guess to what?).  A bill has been introduced to add you-know-who to Mount Rushmore.  There is DOGE, and the ongoing attempt to reshape federal employment.

At the same time, many people have been asking me why Trump chose Canada and Mexico to threaten with tariffs — are they not our neighbors, major trading partners, and closest allies?

I have a theory that tries to explain all these and other facts, though many other factors matter too.  I think of Trumpian policy, first and foremost, as elevating cultural policy above all else.

Imagine you hold a vision where the (partial) decline of America largely is about culture.  After all, we have more people and more natural resources than ever before.  Our top achievements remain impressive.  But is the overall culture of the people in such great shape?  The culture of government and public service?  Interest in our religious organizations?  The quality of local government in many states?  You don’t have to be a diehard Trumper to have some serious reservations on such questions.

We also see countries, such as China, that have screwed-up policies but have grown a lot, in large part because of a pro-business, pro-learning, pro-work culture.  Latin America, in contrast, did lots of policy reforms but still is somewhat stagnant.

OK, so how might you fix the culture of America?  You want to tell everyone that America comes first.  That America should be more masculine and less soft.  That we need to build.  That we should “own the libs.”  I could go on with more examples and details, but this part of it you already get.

So imagine you started a political revolution and asked the simple question “does this policy change reinforce or overturn our basic cultural messages?”  Every time the policy or policy debate pushes culture in what you think is the right direction, just do it.  Do it in the view that the cultural factors will, over some time horizon, surpass everything else in import.

Simply pass or announce or promise such policies.  Do not worry about any other constraints.

You don’t even have to do them!

They don’t even all have to be legal!  (Illegal might provoke more discussion.)

They don’t all have to persist!

You create a debate over the issues knowing that, because of polarization, at least one-third of the American public is going to take your side, sometimes much more than that.  These are your investments in changing the culture.  And do it with as many issues as possible, as quickly as possible (reread Ezra on this).  Think of it as akin to the early Jordan Peterson cranking out all those videos.  Flood the zone.  That is how you have an impact in an internet-intensive, attention-at-a-premium world.

You will not win all of these cultural debates, but you will control the ideological agenda (I hesitate to call it an “intellectual” agenda, but it is).  Your opponents will be dispirited and disorganized, and yes that does describe the Democrats today.  Then just keep on going.  In the long run, you may end up “owning” far more of the culture than you suspected was possible.

Yes policy will be a mess, but as they say “man kann nicht alles haben.”  The culture is worth a lot, both for its own sake and as a predictor of the future course of policy.

Now let’s turn to some details.

In the first week, Trump makes a huge point of striking down DEI and affirmative action (in some of its forms) as the very beginnings of his administration.  The WSJ described it as the centerpiece of his program.  Take origins seriously!

Early on, we also see so many efforts to make statements about the culture wars.  Trans issues, for instance trans out of the military.  No more “Black History Month” for the Department of Defense.  There are more of these than I can keep track of, use Perplexity if you must.

It is no accident that these are priorities.  And keep in mind the main point is not to eliminate Black History Month, though I do not doubt that is a favored policy.  The main point is to get people talking about how you are eliminating Black History Month.  Just as I am covering the topic right now.

How is that war against US AID going?  Will it be abolished?  Cut off from the Treasury payments system?  Simply rolled up into the State Department?  Presidential “impoundment” invoked?  I do not know.  Perhaps nobody knows, not yet.  The point however is to delegitimatize what US AID stands for, which the Trumpers perceive as “other countries first” and a certain kind of altruism, and a certain kind of NGO left-leaning mindset and lifestyle.

The core message is simply “we do not consider this legitimate.”  Have that be the topic of discussion for months, and do not worry about converting each and every debate into an immediate tangible victory.

What about those ridiculous nominations, starting with RFK, Jr.?  As a result of the nomination, people start questioning whether the medical and public health establishments are legitimate after all.  And once such a question starts being debated, the answer simply cannot come out fully positive, whatever the details of your worldview may be.  People end up in a more negative mental position, and of course then some negative contagion reinforces this further.

JFK and UAP dislcosure?  The point is to get people questioning the previous regime, why they kept secrets from us, what really was going on with many other issues, and so on.  It will work.  The good news, if you can call it that, is that we can expect some of the juicier secrets to be made public.

I think by now you can see how the various attempts to restructure federal employment fit into this picture.  And Trump’s “war against universities” has barely begun, but stay tuned.  Don’t even get me going on “Gaza real estate,” the very latest.

Finally, let’s return to those tariffs (non-tariffs?) on Canada and Mexico.  We already know Trump believes in tariffs, and yes that is a big factor, but why choose those countries in particular?  Well, first it is a symbol of strength and Trump’s apparent ability to ignore and contradict mainstream opinion.  But also those are two countries most Americans have heard of.  If Trump announced high tariffs on say Burundi, most people would have no idea what it means.  They would not know how to debate it, and they would not know if America was debasing itself or thumbing its nose at somebody, or whatever.

Canada and Mexico gets the cultural point across.  Canada, all the more so, and thus the Canadian tariffs might be harder to truly reverse.  At least to many Yankee outsiders, Canada comes across as exactly the kind of “wuss” country we need to distance ourselves from.

To be clear, this hypothesis does not not not require any kind of cohesive elite planning the whole strategy (though there are elites planning significant parts of what Trump is doing).  It suffices to have a) conflicting interest groups, b) competition for Trump’s attention, and c) Trump believing cultural issues are super-important, as he seems to.  There then results a spontaneous order, in which the visible strategy looks just like someone intended exactly this as a concrete plan.

In a future post I may consider the pluses and minuses of this kind of political/cultural strategy.

Deep Research

I have had it write a number of ten-page papers for me, each of them outstanding.  I think of the quality as comparable to having a good PhD-level research assistant, and sending that person away with a task for a week or two, or maybe more.

Except Deep Research does the work in five or six minutes.  And it does not seem to make errors, due to the quality of the embedded o3 model.

It seems it can cover just about any topic?

I asked for a ten-page paper explaining Ricardo’s theory of rent, and how it fits into his broader theory of distribution.  It is a little long, but that was my fault, here is the result.  I compared it to a number of other sources on line, and thought it was better, and so I am using it for my history of economic thought class.

I do not currently see signs of originality, but the level of accuracy and clarity is stunning, and it can write and analyze at any level you request.  The work also shows the model can engage in a kind of long-term planning, and that will generalize to some very different contexts and problems as well — that is some of the biggest news associated with this release.

Sometimes the model stops in the middle of its calculations and you need to kick it in the shins a bit to get it going again, but I assume that problem will be cleared up soon enough.

If you pay for o1 pro, you get I think 100 queries per month with Deep Research.

Solve for the equilibrium, people, solve for the equilibrium.

Tuesday assorted links

1. Congestion Pricing Tracker update, sigh…Economists like to think they are useful, but if the initial policy is not right you can expect the results to be disappointing, as these are so far.  It is true congestion prices that work, not “cordon prices.”

2. How is GPT-4o as a judge?

3. North Koreans at McDonald’s.

4. Redux of my sovereign wealth fund column for Bloomberg.  I am against the idea for the United States, while recognizing that any arbitrarily small version of it will appear to make sense.

5. Trump’s CEA nominee wrote a long memo on how to restructure the global trading system.  I turned to o1 pro for comment.

6. Peter Coy is ending his NYT newsletter.

7. Scott Alexander update on model cities.

8. My Bloomberg column on the economics of the Luka trade.

9. Economist Richard Nelson has passed away.

The New Consensus on the Minimum Wage

My take is that there is an evolving new consensus on the minimum wage. Namely, the effects of the minimum wage are heterogeneous and take place on more margins than employment. Read Jeffrey Clemens’s brilliant and accessible paper in the JEP for the theory. A good example of the heterogeneous impact is this new paper by Clemens, Gentry and Meer on how the minimum wage makes it more difficult for the disabled to get jobs:

…We find that large minimum wage increases significantly reduce employment and labor force participation for individuals of all working ages with severe disabilities. These declines are accompanied by a downward shift in the wage distribution and an increase in public assistance receipt. By contrast, we find no employment effects for all but young individuals with either non-severe disabilities or no disabilities. Our findings highlight important heterogeneities in minimum wage impacts, raising concerns about labor market policies’ unintended consequences for populations on the margins of the labor force.

Or Neumark and Kayla on the minimum wage and blacks:

We provide a comprehensive analysis of the effects of minimum wages on blacks, and on the relative impacts on blacks vs. whites. We study not only teenagers – the focus of much of the minimum wage-employment literature – but also other low-skill groups. We focus primarily on employment, which has been the prime concern with the minimum wage research literature. We find evidence that job loss effects from higher minimum wages are much more evident for blacks, and in contrast not very detectable for whites, and are often large enough to generate adverse effects on earnings.

Remember also that a “job” is not a simple contract of hours of work for dollars but contains many explicit and implicit margins on work conditions, fringe benefits, possibilities for promotion, training and so forth. For example, in Unintended workplace safety consequences of minimum wages, Liu, Lu, Sun and Zhang finds that the minimum wage increases accidents, probably because at a higher minimum wage the pace of work increases: 

we find that large increases in minimum wages have significant adverse effects on workplace safety. Our findings indicate that, on average, a large minimum wage increase results in a 4.6 percent increase in the total case rate.

Note that these effects don’t always happen, in large part because, depending on the scope of the minimum wage increase and the industry, large effects of the minimum wage may be passed on to prices. For example here is Renkin and Siegenthaler finding that higher minimum wage increase grocery prices:

We use high-frequency scanner data and leverage a large number of state-level increases in minimum wages between 2001 and 2012. We find that a 10% minimum wage hike translates into a 0.36% increase in the prices of grocery products. This magnitude is consistent with a full pass-through of cost increases into consumer prices.

Similarly, Ashenfelter and Jurajda find there is no free lunch from minimum wage increases, indeed there is approximately full pass through at McDonalds:

Higher labor costs induced by minimum wage hikes are likely to increase product prices.4 If both labor and product markets are competitive, firms can pass through up to the full increase in costs (Fullerton and Metcalf 2002). With constant returns to scale, firms adjust prices in response to minimum wage hikes in proportion to the cost share of minimum wage labor. Under full price pass-through, the real income increases of low-wage workers brought about by minimum wage hikes may be lower than expected (MaCurdy 2015). There is growing evidence of near full price pass-through of minimum wages in the United States….Based on data spanning 2016–20, we find a 0.2 price elasticity with respect to wage increases driven (instrumented) by minimum wage hikes. Together with the 0.7 (first-stage) elasticity of wage rates with respect to minimum wages, this implies a (reduced-form) price elasticity with respect to minimum wages of about 0.14. This corresponds to near-full price pass-through of minimum-wage-induced higher costs of labor.

You can draw your own conclusions about the desirability of the minimum wage, but the fleeting hope that it raises wages without trade-offs is gone. The effects of the minimum wage are nuanced, heterogeneous, and by no means entirely positive.

Gradual Empowerment?

The subtitle is “Systemic Existential Risks from Incremental AI Development,” and the authors are Jan Kulveit, et.al.  Several of you have asked me for comments on this paper.  Here is the abstract:

This paper examines the systemic risks posed by incremental advancements in artificial intelligence, developing the concept of `gradual disempowerment’, in contrast to the abrupt takeover scenarios commonly discussed in AI safety. We analyze how even incremental improvements in AI capabilities can undermine human influence over large-scale systems that society depends on, including the economy, culture, and nation-states. As AI increasingly replaces human labor and cognition in these domains, it can weaken both explicit human control mechanisms (like voting and consumer choice) and the implicit alignments with human interests that often arise from societal systems’ reliance on human participation to function. Furthermore, to the extent that these systems incentivise outcomes that do not line up with human preferences, AIs may optimize for those outcomes more aggressively. These effects may be mutually reinforcing across different domains: economic power shapes cultural narratives and political decisions, while cultural shifts alter economic and political behavior. We argue that this dynamic could lead to an effectively irreversible loss of human influence over crucial societal systems, precipitating an existential catastrophe through the permanent disempowerment of humanity. This suggests the need for both technical research and governance approaches that specifically address the risk of incremental erosion of human influence across interconnected societal systems.

This is one of the smarter arguments I have seen, but I am very far from convinced.  When were humans ever in control to begin with?  (Robin Hanson realized this a few years ago and is still worried about it, as I suppose he should be.  There is not exactly a reliable competitive process for cultural evolution — boo hoo!)

Note the argument here is not that a few rich people will own all the AI.  Rather, humans seem to lose power altogether.  But aren’t people cloning DeepSeek for ridiculously small sums of money?  Why won’t our AI future be fairly decentralized, with lots of checks and balances, and plenty of human ownership to boot?

Rather than focusing on “humans in general,” I say look at the marginal individual human being.  That individual — forever as far as I can tell — has near-zero bargaining power against a coordinating, cartelized society aligned against him.  With or without AI.  Yet that hardly ever happens, extreme criminals being one exception.  There simply isn’t enough collusion to extract much from the (non-criminal) potentially vulnerable lone individuals.

I do not in this paper see a real argument that a critical mass of the AIs are going to collude against humans.  It seems already that “AIs in China” and “AIs in America” are unlikely to collude much with each other.  Similarly, “the evil rich people” do not collude with each other all that much either, much less across borders.

I feel if the paper made a serious attempt to model the likelihood of worldwide AI collusion, the results would come out in the opposite direction.  So, to my eye, “checks and balances forever” is by far the more likely equilibrium.

Emergent Ventures Africa and Caribbean winners, sixth cohort

Maya Chouikrat, Algeria, to support training for an international olympiad of informatics team.

Mercy Muwanguzi and Kwesiga Pather, Uganda, for sanitation robotics to be used in medical centers.

Johan Fourie, South Africa,  Professor of Economics at Stellenbosch University, to write a graphics novel on classical liberalism in a South African context.

Ken Opalo, Associate Professor, Georgetown University, for blogging on African economic development.

Katharine Patterson, Botswana, to support graduate internship in robotics research at NASA Jet Propulsion Laboratory.

Cyril Narh, Ghana, for general career development.

Jon Ortega, travel grant to Silicon Valley.

Alex Kyabarongo, Uganda, Doctor of Veterinary Medicine from Makerere University, to pursue graduate school in the USA for biosecurity.

Joshua Regrello, Trinidad and Tobago, first Steelpannist to perform on the Great Wall of China, Guinness Record Holder for longest steelpan performance, for general career development.

Liam O’Dea, London/Argentina, data science research into parliamentary records of the Caribbean for the last 200 years.

Joshua Payne, undergrad at University of Chicago, for research into mRNA vaccine optimisation, and career development.

Abdoulaye Faye, Senegal, developing Catyu, a firm that designs remotely operated robots.

Deveron Bruce, Barbados, PhD candidate at UWI, to support research in political reform in the Caribbean.

Tony Odhiambo, Kenya, undergrad at MIT, for enhanced training of top performers in mathematics olympiads in Kenya.

Sebastian Naranjo, Panama, PhD candidate at Renmin University of China, to support research on the diplomatic relations of China in Central America.

Ivoine Strachan, Bahamas, for research into designing and developing a VR bodysuit

Phumiani Majozi, South Africa, to establish a think tank promoting classical liberalism in South Africa

Pearl Karungi, Rwanda, for research into redesigning menstrual products.

Emmanuel Nnadi, Nigeria, Microbiologist, to support visiting research at the University of Waterloo in phage therapies.

Youhana Nassif, Egypt, to support an animation and arts showcase in Cairo.

Frida Andalu, Tanzania, to support visiting research in petroleum engineering at the University of Aberdeen.

Rupert Tawiah-Quashie, Ghana, to support his research internship at Harvard University concerning symbolic reasoning in AI models.

I thank Rasheed Griffith for his excellent work on this, and again Nabeel has created excellent software to help organize the list of winners, using AI.

Those unfamiliar with Emergent Ventures can learn more here and here. The EV African and the Caribbean announcement is here and you can see previous cohorts here. If you are interested in supporting this tranche of Emergent Ventures, please write to me or to Rasheed.

Does the Gender Wage Gap Actually Reflect Taste Discrimination Against Women?

One explanation of the gender wage gap is taste discrimination, as in Becker (1957). We test for taste discrimination by constructing a novel measure of misogyny using Google Trends data on searches that include derogatory terms for women. We find—surprisingly, in our view—that misogyny is an economically meaningful and statistically significant predictor of the wage gap. We also test more explicit implications of taste discrimination. The data are inconsistent with the Becker taste discrimination model, based on the tests used in Charles and Guryan (2008). But the data are consistent with the effects of taste discrimination against women in search models (Black, 1995), in which discrimination on the part of even a small group of misogynists can result in a wage gap.

That is a new NBER working paper by Molly Maloney and David Neumark.