Supply and Demand: Who Sets the Price?

The Price Tag in Front of the Fish-Shaped Bun Stand

On a winter evening, a street stall at the corner of the alley sells fish-shaped buns. The price tag reads "three for 2,000 won."

Who set this price? It is easy to answer that the owner set it. Yet the owner cannot just set it however they like. If they raise the price too much, customers walk past. If they lower it too much, they cannot even cover the cost of flour. They also have to align with the price at the stall in the next alley and the number of people walking by today.

A price is not decided by one person's declaration. It is decided where the person who wants to buy and the person who wants to sell meet. This article asks three things: how prices are determined, why prices move when conditions change, and what happens when prices are set by law. On the final question, there are places where the literature conflicts, so the conflicting accounts are presented as they are.

This article explains general concepts and does not advise what to do. Instances from a country identify the country in place.

The Buying Side: Demand

The relationship pairing how much of an item people want to buy at each price is called demand.

Let us suppose that if fish-shaped pastry costs 1,000 won each, people in this neighborhood want to buy 400 of them a day. If it is 500 won, they want to buy 600. The numbers are made up for explanation. The entire bundle where the quantity people want to buy is paired one-to-one with each price is demand. If you draw this as a picture, it becomes a line sloping down to the right, and this line is called the demand curve.

The tendency for the quantity people want to buy to decrease when the price goes up and increase when the price goes down, assuming other conditions remain the same, is called the law of demand.

When the price goes up, the amount you can buy with the same money decreases, and other things worth buying catch your eye instead. The incentives covered in "Incentives: How Do People Respond to Prices and Rewards?" are prices and rewards that change choices. Price is one of them.

The law only states a direction, not a proportion. How much the quantity people want to buy decreases when the price doubles varies from item to item. Explanations writing that it is inversely proportional to price are pointed out to be careless formulations. This is from a commentary reviewing educational and academic literature in the Russian-speaking world (DOI: 10.52957/2221-3260-2024-10-21-42).

There are narrow cases discussed as exceptions to this rule. This is when very poor households depend for their meals on a single staple food. The theory is that even if the price of that staple food goes up, they may buy even more of it to maintain minimum consumption. A field experiment randomly subsidizing the price of rice for poor households in Hunan Province, China, reported strong evidence of such behavior in rice. Evidence was less solid for flour in Gansu Province, but it was stronger when looking only at households meeting the theoretical conditions (DOI: 10.2139/ssrn.310863 · a working paper not yet peer-reviewed by an academic journal). It is an exception with conditions attached, both theoretically and experimentally. The same commentary suggests through its own model that this phenomenon can also appear in consumers with higher incomes (DOI: 10.52957/2221-3260-2024-10-21-42).

The Selling Side: Supply

The relationship pairing how much of an item people are willing to sell at each price is called supply.

Suppose that when carp bread is 500 won per piece, street vendors want to sell 200 pieces a day. At 1,000 won they want to sell 400 pieces, and at 1,500 won they want to sell 600 pieces. If you draw this as a picture, it becomes a line going up to the right, which is called the supply curve.

The tendency for the quantity people are willing to sell to increase when the price goes up and decrease when the price goes down, assuming other conditions remain the same, is called the law of supply.

When the price goes up, the margin left over from selling a single piece grows larger, making it worth baking longer and putting in more ingredients. When the price is at rock bottom, it is better to do something else at that time. The value of that other thing is the opportunity cost, namely the value of the best among the alternatives given up. This was covered in "Scarcity and Opportunity Cost: What Do You Give Up When You Choose?".

The Price Where the Two Sides Meet: Equilibrium Price

Let us place the numbers from above side by side. At 500 won, the quantity people want to buy is 600, and the quantity people want to sell is 200. At 1,000 won, both quantities are 400. At 1,500 won, the quantity people want to buy is 200, and the quantity people want to sell is 600.

The price where the quantity people want to buy equals the quantity people want to sell is called the equilibrium price. The equilibrium price in this neighborhood is 1,000 won. At this price, everyone who wants to buy purchases what they want, and everyone who wants to sell sells what they want. The point where the demand curve and the supply curve meet is precisely this price.

What happens when the price moves away from equilibrium?

A state at a certain price where the quantity people want to buy is greater than the quantity people want to sell is called excess demand, and is commonly referred to as a shortage. At 500 won, the quantity people want to buy is 400 higher. Lines grow longer, and buyers emerge who are willing to pay more. Then the price goes up.

A state at a certain price where the quantity people want to sell is greater than the quantity people want to buy is called excess supply, and is commonly referred to as a surplus. At 1,500 won, the quantity people want to sell is 400 higher. Unsold fish-shaped pastries are left over, and the sellers lower their prices. A shortage acts as a force that pushes prices up, and a surplus acts as a force that pushes prices down, causing the price to tilt toward the equilibrium price.

A book review in a primer introduces a classroom experiment. When all participants can see the trades taking place, the price quickly converges to the level predicted by theory. This happened even though no single participant had the information to calculate that price on their own (DOI: 10.1111/1468-0297.13915). The same review also writes the following. In textbooks, people know the curves and read prices off a chart, but real markets do not work that way. The curves and the equilibrium price are frameworks for understanding the market. Real-world prices often deviate from equilibrium. We learn this framework typically by looking at markets with many buyers and sellers where no single person can move the price on their own (DOI: 10.71097/ijsat.v16.i2.5670 · An article explaining price determination through supply and demand).

When the Price Changes Versus When the Conditions Change

There is a spot that people learning the demand curve for the first time confuse most often. We can tell them apart in one go using a single scene of bungeoppang (carp-shaped red bean pastries).

When the amount people want to buy changes because the price of that item itself has changed, this is called a movement along the curve. When bungeoppang rise from 1,000 won to 1,500 won, the quantity people want to buy drops from 400 to 200. The point has slid down along the same demand curve. The curve itself stays put.

When the amount people want to buy changes at the exact same price because a condition other than price has changed, this is called a shift of the curve itself. Let us say a cold wave sweeps in and everyone looks for something warm. bungeoppang remain 1,000 won, but the quantity people want to buy increases from 400 to 600. Because the quantity people want to buy has increased at every price, the demand curve itself has moved to the right.

We can tell them apart with a single question. Is what changed the price of that item? If so, we move along the curve. Is it something other than the price? If so, the curve itself moves.

The conditions that move the curve are several. One chapter of an introductory textbook lists consumer income, the prices of competing and complementary goods, and tastes as factors that change the quantity people want to buy. Factors that change the quantity people want to sell include the prices of other goods that can be made, the prices of production factors like ingredients, technology, and corporate goals (DOI: 10.1079/9781780644752.0026).

Two goods that can be used in place of each other, so that if the price of one falls, people buy less of the other, are called substitutes. When the price of hotteok (sweet pancakes) drops significantly, some people who meant to buy bungeoppang switch over to hotteok, and the demand curve for bungeoppang shifts to the left.

Two goods whose usefulness grows greater when used together, so that if the price of one rises, people buy less of the other, are called complements. Sliced bread and jam are like that. If the price of jam rises significantly, the demand for sliced bread can also drop. The story differs depending on whether the reason one side's price rose came from the demand side or the supply side. However, a critique points out that the textbooks examined lack a discussion on what drove the price up (DOI: 10.1177/056943450905400211 · A paper reviewing textbooks).

The supply side is the same. When the flour price rises, the share left over after baking bungeoppang shrinks, reducing the quantity people want to sell at the same price, and the supply curve shifts to the left.

Now let us see how the equilibrium changes when the two curves move. When a cold wave shifts the demand curve to the right, the equilibrium price rises and the bungeoppang traded increase as well. When a rise in the flour price shifts the supply curve to the left, the equilibrium price rises just the same, but the bungeoppang traded decrease. Even though the outward appearance of the price rising is the same, the trade volume moves in opposite directions depending on which side of the curve moved.

How Sensitive Are People to Prices: Elasticity

A measure of how sensitively the quantity people want to buy changes when the price changes is called elasticity. It is measured by the percentage change in the quantity people want to buy when the price rises by 1%. Because the quantity people want to buy decreases when the price rises, this number is usually written with a minus sign.

Even for the same price increase, people react differently depending on the item. For example, even if the price of salt goes up by 10%, the amount of salt people want to buy stays almost the same. If the price of bungeoppang goes up by 10%, some people will switch to hotteok, so the amount people want to buy will drop noticeably. Whether there is a substitute item you can switch to is one reason that determines this sensitivity.

Introductory textbooks also explain different reactions under the same name. They are concepts that separately measure the reaction when income changes and the reaction when the price of another good changes (DOI: 10.1093/hebz/9780198806523.003.0006).

Prices Are Signals Before They Are Numbers

The way a price tells everyone at once how scarce something is—this is called a price signal.

Let us suppose the wheat harvest suffers a major crop failure. The customers at a carp-shaped pastry stall do not know that news. Even so, when the price of flour rises and the price of the pastry goes up, a customer buys one fewer. The price has delivered the news in its place. A higher price is a notice that this item has grown scarce, and a lower price is a notice that it has grown plentiful.

The economist Hayek viewed the market as a system that processes scattered information. Countless people move separately, each with only the knowledge around them. His concept is that prices, arising within dispersed competition, coordinate those actions. This concept is introduced in a retrospective paper on Hayek (DOI: 10.1257/jep.31.3.215). A review of the primer introduces the author's thesis, which is that the market's economic function largely lies in gathering scattered information into prices as signals of value and scarcity (DOI: 10.1111/1468-0297.13915).

Signals are not always accurate. One paper evaluates that Hayek's thesis has strength in explaining the failure of central planning. However, it notes that the financial crisis near that time, discussed in a 2013 paper, showed market prices can give very wrong signals for a long time (DOI: 10.23941/ejpe.v6i1.120). The authors of the Hayek retrospective paper acknowledge his market framework. However, they do not view it as logically supporting his policy arguments for a free market (DOI: 10.1257/jep.31.3.215). Prices are signals conveying the situations of the buying side and the selling side, and signals are sometimes wrong.

When the Law Sets the Price: Price Ceilings and Price Floors

A legally mandated price above which you cannot charge is called a price ceiling. A legally mandated price below which you cannot exchange goods is called a price floor.

Let us suppose the equilibrium price of a bungeoppang is 1,000 won, but the law dictates it must be sold for 500 won or less. If the ceiling is lower than the equilibrium price, theory predicts a shortage. At 500 won, the quantity demanded is 600 while the quantity supplied is 200, leaving a shortfall of 400. Even when the price is held down, the shortage of bungeoppang does not simply vanish. It changes form into waiting in line, extra premiums, and under-the-counter transactions. Conversely, if the floor is higher than the equilibrium price, theory predicts a surplus. If they cannot sell below 1,500 won, the quantity supplied exceeds the quantity demanded by 400.

There are reports that these predictions manifest in altered forms in reality. The working paper below has not yet undergone peer review for an academic journal.

On-street parking in American cities is mostly free or priced low. A paper treating this as an example of a price ceiling evaluates it as follows. It finds that in a significant number of cities, the typical consequences of a ceiling—shortages, misallocation, the cost of searching for a vacant spot, and a black market—manifest (DOI: 10.1108/s2044-994120140000005019 · a book chapter containing an author's evaluation).

The official retail prices in the former Soviet Union from 1981 to 1988 rose by about 0.6% each year. The author interprets household savings balances, inventories, and subsidy indicators during the same period as pointing to suppressed pressure. They then conclude that price controls shifted and postponed visible inflation into forced savings, shortages, and fiscal subsidies. A small regression analysis states that due to the short sample, it is merely diagnostic (DOI: 10.2139/ssrn.6664838 · working paper).

In the Netherlands, bundles of medicines where the price ceiling fell by more than 10% differed from comparison bundles under the same conditions. The number of suppliers decreased, and the proportion of the market in a stockout state increased (DOI: 10.2139/ssrn.7092858 · working paper).

When ceilings are enforced loosely or are binding only for some, predictions can also be inverted. In rental advertisement data from Vienna, Austria, the authors report that the stricter the ceiling—which is enforced only through tenant reports—the higher the advertised rent (DOI: 10.2139/ssrn.6436778 · working paper, including a theoretical model). There is also a study on price ceilings in isolated markets with only a few gas stations. Gas stations bound by the ceiling lowered their prices, but gas stations not bound by it instead raised their prices (with large variance), which the authors interpret as the ceiling serving as a focal point for collusion (DOI: 10.2139/ssrn.6438081 · working paper). This case is an instance where the ceiling was binding for only some. The outcome of a ceiling varies depending on how close it is to the equilibrium price and how strictly it is observed.

When Rents Are Capped: Rent Control

Rent control is a system that places limits on the level of rent or how much it can rise. The literature does not speak with one voice on what this system does. The literature gathered in this article places both sides side by side.

First is the side reporting side effects. A review spanning empirical literature summarizes it this way. Rent control appears quite effective at slowing the rise of rents in controlled housing, but it also produces various side effects across society as a whole (DOI: 10.1016/j.jhe.2024.101983). A working paper using household data from Stockholm, Sweden, estimates that regulated rents are lower than market rents. It interprets that higher-income households, who benefit more from the regulation, occupy larger homes and crowd out other households (DOI: 10.2139/ssrn.6629277). In New York, there is a report that a substantial proportion of apartments appear misallocated among demographic groups (DOI: 10.1257/000282803769206188). A paper looking at the 2024 rent control in Catalonia, Spain, reports a decrease in the number of rental contracts. It notes that the decrease in the rent growth rate was less robust and calls for caution due to data limitations (arXiv:2602.08631v1 · preprint, not peer-reviewed).

The other side consists of papers reporting effects or countering conventional wisdom about side effects. There is a Korean study examining the 2020 two lease laws for apartments in Seoul. When tenants used the right to extend their contract for one more term (the right to request contract renewal), the rent increase rate was 12.29 percentage points lower than when they did not use it. This comparison is not the effect of the 5% cap alone, but the difference between exercising versus not exercising the renewal right, and the two systems were introduced together. The authors interpret that this system contributed to stabilizing rents. At the same time, they note concerns that the 5% cap may distort the market in the long run. They also state that the effect is strong in jeonse and weak in monthly rent, and varies by region (DOI: 10.24957/hsr.2025.33.2.133). A chapter of a book advocating for rent control as a policy refutes conventional wisdom. It argues that the common belief that caps harm housing supply or quality lacks foundation (DOI: 10.7591/cornell/9781501782091.003.0008 · a piece advocating the policy).

This is as far as the literature confirms. Reports that rent increases are slowed appear in the review and the Seoul study. Regarding supply, allocation, and transactions, the empirical studies gathered in this article report side effects, while arguments opposing this come from a chapter in a book advocating the policy.

When You Raise the Price of Work: The Minimum Wage

There are buyers and sellers in the job market as well. The store looking for workers is the buyer, the person wanting to work is the seller, and the wage is the price of the work. Setting a legal lower limit on the hourly wage that can be paid in exchange for work is called the minimum wage. This is an example of a price floor.

If the floor is higher than the equilibrium wage, the basic prediction of the theory is that there will be more people wanting to work than there are jobs. Soon, employment decreases. However, there is not just one theory. In markets where there are few buyers of labor so wages can be held down, there is an argument that a floor may not reduce employment or could even increase it. A preprint using German data reports that industry-specific minimum wages reduce employment in competitive labor markets. In markets where the buyers of labor are highly concentrated, that effect is said to flip toward increasing employment (arXiv:2111.13692v7 · preprint; the authors read this as empirical support for this argument).

What about empirical research? One paper summarizes that it is difficult to find a negative effect that reduces employment, while it is easy to find an effect on wages. The author believes this situation is unlikely to be resolved with better data or methods. There are two reasons: the path by which the minimum wage reaches labor costs is weaker than thought, and imperfect competition is widespread in the labor market (DOI: 10.1257/jep.35.1.3).

Even in meta-analyses that re-analyze multiple studies together, the results do not point in a single direction. The analyses below may have overlapping samples, so they are not summed or ranked.

A meta-analysis of 64 US studies found that this literature is contaminated by publication selection bias. Publication selection bias is the tendency for significant results to be published more easily. Correcting for the bias leaves little to no negative association between the minimum wage and employment (DOI: 10.1111/j.1467-8543.2009.00723.x).

An analysis looking at studies from 2010 to 2014 is a follow-up to the preceding analysis. It is posted in a public repository, and whether it has been published in an academic journal is unverified. This analysis reports that there is publication selection in the elasticity samples, and correcting for it leaves a small negative effect. In the coefficient samples, there is no bias and the true effect is negative but small (DOI: 10.7910/dvn/x2azej).

A meta-analysis gathering US studies concludes that the estimated range of the effect has become closer to zero than before. The effect is negative but noticeably smaller and is mostly concentrated among teenagers (DOI: 10.1111/labr.12162).

A meta-analysis gathering UK studies finds that overall there is no practically meaningful reduction in employment, and publication bias is lower than in the US. However, it notes that there may be an actual reduction in employment in the nursing home sector (residential care) (DOI: 10.1111/bjir.12031). A meta-regression manuscript gathering 187 estimates from 18 EU studies reports no substantive effect overall. It also finds no evidence of publication selection bias, but notes negative effects in nursing homes, the retail sector, female employment, and Eastern Europe (DOI: 10.13140/rg.2.2.24184.90882).

A preprint re-bundling existing meta-analyses addresses three topics together: informal labor, the minimum wage, and monopsony power. This preprint concludes that the employment effects of the minimum wage are consistently small. It also states that across all three topics, the publication bias-corrected values are generally close to zero (DOI: 10.48550/arxiv.2509.20465 · preprint).

The pointing out of publication bias, the failure to find evidence of bias, and the finding that a small negative effect remains even after correction all exist within the same literature. Directly supporting the result that a small negative effect remains even after correction is a single follow-up analysis gathered in this piece, whose peer-review status is unverified.

One reason discussed for why the effects vary is whether the minimum wage actually has the power to pull up wages. In US county-level data, when all counties were pooled, the effect was small and insignificant. However, in counties where the minimum wage was likely higher than the market wage, the effect of reducing teen employment was much larger (DOI: 10.1177/001979390906200305).

There is also the case of South Korea. In 2018, South Korea's minimum wage rose by 16.4%. One working paper compared two groups using individual panel data. These were workers whose 2017 wage was lower than the new minimum wage (affected workers) and comparison workers whose wages were slightly higher. The hourly wages of affected workers who remained employed rose, and the paper reports no significant effect on employment in the overall sample. When broken down by industry, affected workers in manufacturing saw their probability of employment drop by about 10 percentage points. In non-manufacturing, working hours decreased without a similar employment loss. The author interprets that adjustments occurred through different channels across industries (DOI: 10.2139/ssrn.6889587 · working paper). In Puerto Rico, the minimum wage rose by 44.8% across three stages. A working paper examining this case reports finding no evidence that total employment was reduced (DOI: 10.2139/ssrn.6854082).

This is as far as the literature goes. While there are multiple reports that wages rise, the impact on employment varies in direction and magnitude across studies and conditions.

Putting It to the Test in the Real World

Let us see what numbers come out when we measure sensitivity to prices—that is, elasticity—in the real world. The numbers vary depending on the item, region, and time period.

There is a preprint compiling the price elasticity of heating and cooling energy. It gathered estimates from over 400 original studies. It found strong signs of publication bias, where insignificant or positive estimates are underreported, and of the practice of selecting analyses so that significant values remain (p-hacking). The bias-corrected elasticity is -0.05 to -0.2 in the short run and -0.1 to -0.3 in the long run (DOI: 10.31219/osf.io/4sjy5_v2 · preprint). Unpacking these numbers, when the price rises by 10%, usage drops by about 0.5 to 2% in the short run and 1 to 3% in the long run.

The global elasticity of sugar-sweetened beverages is -0.423. This is an estimate based on sales volume and real price data from 2010 to 2024. When divided into the six World Health Organization regions, only Europe and the Americas were estimated significantly, while the remaining regions were not significant (DOI: 10.3389/fpubh.2026.1782007).

The method of measurement also changes the numbers. In Spain, fuel taxes varied by region. Looking at this data, a working paper reports that the elasticity of gasoline purchases in response to price differences with neighboring areas near regional boundaries was very large. If you ignore the movement of going to live in another region, fuel demand is estimated to be very elastic (sensitive to price). When you account for it, it becomes inelastic (insensitive to price) (DOI: 10.2139/ssrn.6694958). Even for the same item, elasticity can vary depending on how it is measured.

Where to Look Next

Prices are determined where buyers and sellers meet, and when those prices are suppressed by law, the literature is mixed. Rent controls and minimum wages are open questions without settled verdicts, and new research keeps coming out.

Prices do not always tell you everything. There are moments when something is not captured in the price. Costs that spill over outside the two people making the transaction are like that. "When Markets Fail — The Moments a Price Cannot Tell the Whole Story" covers this next.

Source Papers

• "Reinventing the Bazaar: A Natural History of Markets." (2003) — DOI: 10.1111/1468-0297.13915 — Book review: The author's thesis that the function of markets is to gather dispersed information into prices, which are signals of value and scarcity; in experiments, when all participants can see trades, prices converge to theoretical levels.
• "The Price of Lower Price Caps: How Pharmaceutical Price Caps Affect the Supplier Landscape and Shortages" (2026) — DOI: 10.2139/ssrn.7092858 — Working paper: In drug bundles in the Netherlands where caps fell by more than 10%, the number of suppliers decreased and the proportion of out-of-stock markets increased.
• "Perverse Effects of Imperfectly Enforced Price Controls" (2026) — DOI: 10.2139/ssrn.6436778 — Working paper: Among vacant rental listings, the stricter the cap, the higher the advertised rent was by an average of 76 euros.
• "The Impact of Rent Control on Housing Consumption" (2026) — DOI: 10.2139/ssrn.6629277 — Working paper: Stockholm's regulated rents are an average of 36.1% lower than market rents, and higher-income households with larger benefits use larger homes.
• "The Effects of Minimum Wage Increases on Employment and Firm Dynamics in Puerto Rico" (2026) — DOI: 10.2139/ssrn.6854082 — Working paper: Found no evidence of a decrease in total employment from a 44.8% increase.
• "Labor Market Effects of a Large Minimum Wage Increase: Evidence from South Korea's 2018 Reform" (2026) — DOI: 10.2139/ssrn.6889587 — Working paper: After a 16.4% increase, wages rose 6.8 to 7.8%, with no overall employment effect and a roughly 10 percentage point drop in the probability of manufacturing employment.
• "Integrated analysis of informality, minimum wage, and monopsony power: A synthesis of meta-analyses with unified theoretical underpinnings" (2025) — DOI: 10.48550/arxiv.2509.20465 — Preprint: Employment effects are consistently small and close to zero when publication bias is corrected.
• "Do the Right to Lease Renewal and the Rent Cap Curb Apartment Rent Increases?: Dissecting Rent Dynamics in a Regulated Housing Market" (2025) — DOI: 10.24957/hsr.2025.33.2.133 — Seoul's 2020 Lease Act: The increase rate for tenants exercising the right to renewal was 12.29 percentage points lower, and the effect varies by type and region.
• "Role of Demand and Supply in Price Determination under Perfect Competition Market" (2025) — DOI: 10.71097/ijsat.v16.i2.5670 — Explains price determination under a perfectly competitive market using the laws of demand and supply.
• "Rent control effects through the lens of empirical research: An almost complete review of the literature" (2024) — DOI: 10.1016/j.jhe.2024.101983 — Review: Slows rent increases for controlled housing but creates various side effects for society as a whole.
• "Demand, the law of demand, Giffen and Veblen effects. Part 1. Functional demand" (2024) — DOI: 10.52957/2221-3260-2024-10-21-42 — Commentary: Using the law of demand as an inverse relationship is a careless formulation.
• "The Elusive Employment Effect of the Minimum Wage" (2021) — DOI: 10.1257/jep.35.1.3 — Employment effects are hard to find, while wage effects are easy to find.
• "The elasticity of demand" (2019) — DOI: 10.1093/hebz/9780198806523.003.0006 — Textbook chapter: Concepts of price, income, and cross-price elasticity.
• "Retrospectives: Friedrich Hayek and the Market Algorithm" (2017) — DOI: 10.1257/jep.31.3.215 — Hayek's vision of the market as an information-processing system and questions regarding his policy arguments.
• "Demand and supply: the price mechanism in a market economy." (2014) — DOI: 10.1079/9781780644752.0026 — Textbook chapter: Factors that shift demand and supply.
• "Publication Selection Bias in Minimum-Wage Research? A Meta-Regression Analysis" (2009) — DOI: 10.1111/j.1467-8543.2009.00723.x — Meta-regression of 64 US studies: Almost no negative association after correcting for publication selection.
• "Cross Price Elasticity and Income Elasticity of Demand: Are Your Students Confused?" (2009) — DOI: 10.1177/056943450905400211 — Textbook explanations of cross-price and income elasticity are ambiguous.
• "Using Local Labor Market Data to Re-Examine the Employment Effects of the Minimum Wage" (2009) — DOI: 10.1177/001979390906200305 — Negative effects on teen employment are much larger in counties with high binding constraints.
• "The Misallocation of Housing Under Rent Control" (2003) — DOI: 10.1257/000282803769206188 — A substantial proportion of New York apartments appear to be misallocated.
• "Giffen Behavior: Theory and Evidence" (2002) — DOI: 10.2139/ssrn.310863 — Strong evidence in Hunan rice, less robust evidence in Gansu flour.
• "Price Controls, Repressed Inflation, and Monetary Overhang in the Late Soviet Economy" (2026) — DOI: 10.2139/ssrn.6664838 — Working paper: Price controls shifted inflation into reduced choices, forced savings, shortages, and subsidies.
• "Trends in sugar-sweetened beverage prices, sales, and elasticities: policy evidence from WHO regions, 2010–2024" (2026) — DOI: 10.3389/fpubh.2026.1782007 — Global elasticity -0.423, Europe -0.828, Americas -0.509.
• "Parking Pricing" (2014) — DOI: 10.1108/s2044-994120140000005019 — The author's evaluation of curbside parking as an example of a price ceiling.
• "Hayek on the wisdom of prices" (2013) — DOI: 10.23941/ejpe.v6i1.120 — Prices can give the wrong signals for a long time.
• "Effectiveness of Rent Controls: Evidence from Spain" (2026) — arXiv:2602.08631v1 — Preprint: Rental contracts decrease in Catalonia, while the decrease in rent growth rates is less robust.
• "The Impact of Price Ceiling Regulation: Evidence from the Retail Gasoline Market" (2026) — DOI: 10.2139/ssrn.6438081 — Working paper: Gas stations unaffected by the cap raised their prices, interpreted as a coordination benchmark.
• "The Price Elasticity of Heating and Cooling Energy Demand" (2025) — DOI: 10.31219/osf.io/4sjy5_v2 — Preprint: Short-run -0.05 to -0.2, long-run -0.1 to -0.3 after correction.
• "How We Fix This—Rent Control" (2025) — DOI: 10.7591/cornell/9781501782091.003.0008 — Chapter from a pro-advocacy book: Refutes the conventional wisdom that caps harm supply or quality.
• "Minimum Wages in Concentrated Labor Markets" (2021) — arXiv:2111.13692v7 — Preprint: The effect of industry-specific minimum wages in Germany shifts from negative to positive depending on labor market concentration.
• "15 Years of Research on US Employment and the Minimum Wage" (2019) — DOI: 10.1111/labr.12162 — Meta-analysis of 37 studies: Consensus range shifts toward zero.
• "Are employment effects of minimum wage the same across the EU? A Meta-regression analysis" (2018) — DOI: 10.13140/rg.2.2.24184.90882 — 18 studies, 187 estimates: Virtually no overall effect, with negative effects in some sectors and Eastern Europe.
• "Is There Publication Selection Bias in Minimum Wage Research during the Five-year Period from 2010 to 2014?" (2015) — DOI: 10.7910/dvn/x2azej — 45 studies: Small negative effects remain even after correction.
• "Does the UK Minimum Wage Reduce Employment? A Meta-Regression Analysis" (2013) — DOI: 10.1111/bjir.12031 — 16 studies: No overall substantial reduction in employment, though possible in care homes (residential care).
• "Localized Fuel Taxes and Consumer Behavior: Spatial Relocation and the Price Elasticity of Fuel Demand" (2026) — DOI: 10.2139/ssrn.6694958 — Working paper: Ignoring regional mobility leads to fuel demand being estimated as elastic.