Myth and Measurement Reprint edition by Card, David, Krueger, Alan B. (1997) Paperback

by David Card

The authors present a powerful new challenge to the conventional view that higher minimum wages reduce jobs for low-wage workers. A distinctive feature of their research is the use of empirical methods borrowed from the natural sciences, including comparisons between the "treatment" and "control" groups formed when minimum wage rises for some workers but not for others. Line drawings.

Genres: Economics, Nonfiction

Paperback

First published January 1, 1995

Reviews

[Jesse Maurais · Rating 4 out of 5]

According to the standard supply-and-demand model of economics, imposing a minimum wage on labour would effectively price-out anyone who would otherwise be earning less. In 1992 the state of New Jersey increased it's minimum wage while it's neighbour, Pennsylvania, did not. But both had previously equal rates set for the minimum.

The authors of this book set out to study what, if any, effect that this rise in the minimum wage would have on the level of employment in New Jersey, by comparing the changes in employment status among affected workers there as compared with those in Pennsylvania. This cross-state comparison allows for a "natural" experiment that is something more like a controlled experiment, since it supplies the researcher with a counterfactual case for purposes of comparison. This 'differences-in-differences' method accounts for seasonal and other trends automatically, assuming the appropriate counterfactual, and is a property than other methods lack.

Contrary to the predictions of the standard model, the authors find no evidence that there is a negative effect on employment due to the minimum wage and some evidence that there is a slightly positive effect. The book goes into further research, including a meta-analysis of previously published work, finding evidence of publication bias in favour of studies that claimed significant disemployment effects. They also examine evidence internal to other US states, international evidence, and go into detail about the possible disemployment of teenagers in particular.

Furthermore, they have attempted to explain the anomalous findings by proposing a monopsony model for the market in unskilled labour, which predicts an employment effect more like what they found in their econometric study, though the model does not fit the measured effect on prices.

[Jonathan F · Rating 4 out of 5]

I used to love theory, and while I still do, my career has refined by palate for empirical studies that produce counterintuitive or unexpected results. Working in marketing experimentation, I have learned through trial-and-error that first theories rarely survive the test and almost always require further refinement. Card and Krueger's research is the test and the perfectly competitive model of labor markets is the theory that doesn't entirely pass. And much like with marketing experimentation, what the authors propose is not doing away with the initial theory, but making it more sophisticated and better able to explain the data. I have a much deeper appreciation for Myth and Measurement today than I did a decade ago, so I am glad that I decided to revisit Card and Krueger's research.

As an economics book, Myth and Measurement is only a good introduction to the minimum wage research. Published in 1994, the literature has continued to grow since then, including additions by Card and Krueger, Dube, Neumark, Hoffman, and many others — although Card and Krueger's research continues to feature in these studies bibliographies, making it foundational. What I like about Myth and Measurement is that it puts the authors' early work together, as a collection of various articles edited into a cohesive book. If you're interested in learning more about the economics of the minimum wage and don't know where to start, this is a good place.

The right angle to approach the book is to avoid imposing a prior opinion on the minimum wage because the lessons on methods are more important than the results. Card and Krueger have been slightly successful in shifting opinion on the minimum wage among economists, but the supermajority of economists still believe that minimum wage hikes are harmful to low-wage employment despite Card and Krueger's evidence suggesting otherwise. Where they have been most successful in changing the way economists explore the problem. There is more of an onus now on proving that the perfectly competitive model offers the most accurate predictions, and where there is doubt it has caused economists to broaden their theoretical modeling to account for unexpected empirical results.

In other words, Myth and Measurement is a battery of tests that add credence to the motto: never take your assumptions and your models for granted, they need to be tested. On a deeper level, the book is also a good journey through how two extremely smart economists think about how to test their models.

Testing methodology is crucial, as anyone who does empirical testing in any field knows. There are questions on whether the control group is a good control group, whether there are any biases that will affect the results and/or their interpretation, whether the measurement of the data is accurate, and whether the model you are testing is even a good one for interpreting the results. In economics, these questions and issues are oftentimes more difficult to grapple with than in other fields, because it's extremely difficult to build random experiments. It was incredibly interesting to follow Card and Krueger as they think about the data, the empirical methods, the interpretations, and the results.

For example, they admit that empirical literature utilizing time-series analyses typically report statistically significant negative delta on employment levels with positive changes in the minimum wage. Time series data necessarily compare one period of time with another, assuming another period of time acts as a good control. Their own research looks at differences in differences within groups in the same time period, believing that by controlling for time and differentiating between affected and unaffected workers they can get better estimates of minimum wage's impact. They also show that if you extend the time-series data, the regressed negative delta disappears. Whether or not one agrees with their criticism of time series data and their preferred alternative, it was enlightening to even think about these questions because they make your analytical capabilities so much more alert to the nuances.

Thus, apart from being a good introduction to the minimum wage literature, it's also a great resource for analysts and experimenters. If you want to learn more about the economics of the minimum wage, you'll definitely need to follow up with some of the more recent research, but this is a good start and it will prepare you for many of the themes that the new literature explores. Even so, Myth and Measurement is a book that I will periodically re-read not because of its conclusions, but because of how it gets to those conclusions.

[D · Rating 4 out of 5]

some of the math was over my head. basic findings are compelling and research methodologies seem reasonable. probably no need to read the book unless you really want to get into the details to alternative theoretical models for labor market dynamics.

[Avi Moorthy · Rating 4 out of 5]

Technical at times but a very compelling argument for minimum wages. Rip Alan Krueger

[Usue Sagüés · Rating 4 out of 5]

Es un libro muy interesante, sobre todo a nivel metodológico.
Aborda varios temas relacionados con el salario mínimo: pobreza, distribución de la renta, beneficios empresariales... aunque hace especial hincapié en el efecto sobre el empleo.
Explica de forma técnica pero clara las decisiones metodológicas adoptadas, exponiendo en todo momento los resultados econométricos obtenidos y la procedencia de los datos, lo que no solo lo dota de transparencia, sino que facilita la replicabilidad del estudio.
Los resultados son muy interesantes y ponen de tela de juicio las presunciones básicas del modelo estándar del mercado laboral, proponiendo el funcionamiento de este en base a un monopsonio.
El único aspecto negativo es que a veces da la sensación de que los autores generalizan el efecto positivo del salario mínimo hallado en el experimento realizado en el sector de comida rápida de algunos estados tras la subida del salario mínimo, algo dificilmente extrapolable a otros países con distinta coyuntura económica e incluso a otros sectores.
A pesar de ello, los autores intentan mantener un tono cauto a lo largo de toda la exposición, abogando por la profundización en torno a los efectos del salario mínimo a través de técnicas experimentales.
Por último, indicar que es un libro que requiere conocimientos sobre teoría económica (preferiblemente economía laboral) y sobre econometría.