A lemon rots from the inside. It may look flawless on the outside, but once you cut it open, it turns out to be bad. That’s why a bad second-hand car is called a lemon: it may be polished up for sale, but underneath it’s worn down, rusty and unreliable.
In his classic paper, economics Nobel laureate George Akerlof showed how such hidden information can break markets: if buyers can’t tell good quality second-hand cars from bad ones, they won’t pay full price – so only the worst cars get sold (Akerlof, 1970).
But how big is this problem in real life? In a study that I co-authored with Ran Gu, Søren Leth-Petersen, Hamish Low and Costas Meghir, and published recently in Quantitative Economics, we use rich Danish data to test the theory in the market for second-hand cars (Blundell et al, 2026).
Analysing Danish data on households and second-hand cars
We use comprehensive administrative data covering all cars and households in Denmark over nearly two decades. This allows us to observe when people buy and sell cars, how long they own them and how their financial situation evolves over time. Analysing a life-cycle model in which households face income shocks (perhaps because of job instability or health issues) and credit constraints (limits on their ability to borrow when they need cash), and where only sellers know the true condition of their car, we find that:
- Even good second-hand cars sell at a discount because buyers can’t tell them apart from bad ones. This discount – which we refer to as the lemons penalty – is largest in the first year of ownership: sellers lose around 12% of the purchase price, on top of normal depreciation of 19%.
- The lemons penalty fades over time. After two years of ownership, it drops to about 6%, and it shrinks to nearly zero after eight to nine years, when any hidden flaws are likely to be obvious.
- The market does not collapse. Despite asymmetric information (sellers knowing more than buyers about the true quality of the cars that they put up for sale), many people still sell good cars. Income shocks push people to sell – often because they need cash – which helps to sustain the market.
- But the lemons problem still matters. It reduces car turnover, slows down upgrades and limits the ability of households – especially those with limited savings – to use cars as financial buffers.
Implications for households and markets
For many low-wealth households, a car is one of the highest value assets that they own. But because of the lemons problem, it is not always the safety net they need. Selling a car during hard times comes at a cost, weakening its role as a tool for smoothing income shocks – especially when access to credit is limited. Our findings suggest that the lemons penalty becomes smaller during recessions, when more people are forced to sell high-quality cars, improving the average quality on offer.
Our findings highlight the real-world impact of asymmetric information in markets for durable goods. Private information – not just transaction costs such as registration fees – can create inefficiencies and shift who gains and who loses. Owners of good cars are penalised, while those selling lemons benefit from pooled pricing.
Beyond cars, these insights apply to other markets for second-hand goods, and they matter for how we think about consumer resilience, credit design and inequality.
