The importance and challenges of measuring work hours

Measuring hours worked is important, but different surveys can tell different stories

Bureau of Labor Statistics, USA, and IZA, Germany


Elevator pitch

Work hours are key components in estimating productivity growth and hourly wages as well as being a useful cyclical indicator in their own right, so measuring them correctly is important. The US Bureau of Labor Statistics (BLS) collects data on work hours in several surveys and publishes three widely-used series that measure average weekly hours. The series tell different stories about average weekly hours and trends in those hours but qualitatively similar stories about the cyclical behavior of work hours. The research summarized here explains the differences in levels, but only some of the differences in trends.

The key work hours series exhibit similar
                        cyclical behavior

Key findings


On average, respondents to the BLS’s household survey correctly report weekly hours worked per person and weekly hours worked on their main jobs.

All three series exhibit similar cyclical behavior.

Differences in levels between the three series can be explained by differences in concepts and coverage.

All three series exhibit similar trends since the beginning of the 1990s.


There are significant differences between the three series in levels and trends, and the differences in long-term trends cannot be completely reconciled.

In the household survey, respondents under-report multiple jobholding and over-report hours worked on second jobs, but aggregate hours are approximately correct because these errors in reporting mostly offset each other.

Estimating annual work hours from average weekly hours overstates the annual number of hours worked, because the surveys’ reference periods exclude most holidays.

These results may not generalize to other countries because of differences in survey methods.

Author’s main message

The BLS publishes three widely-used weekly-hours series: one from its household survey, one from its establishment survey, and one that combines data from the two. The difference in level between these series can be explained by survey features, but the difference in long-term trends can be only partially explained. However, all three series tell qualitatively similar stories about the cyclical behavior of weekly work hours. It is important for decision makers to understand the advantages and limitations of the different hours series.


Measuring weekly work hours correctly is important for estimating average hourly earnings and labor productivity. If average weekly hours are overestimated, then average hourly wages are underestimated. To illustrate, some research using US data has shown that college graduates tend to overestimate their hours, while high school graduates tend to slightly underestimate theirs [1]. This imparts a downward bias on the average hourly wage of college graduates and an upward bias on the average wage of high school graduates, thus underestimating the college–high school wage difference. For measuring productivity, the actual number of hours worked is not as important as the growth rate. Most countries compare the growth in output to the growth in total labor hours (average weekly hours × employment × number of weeks per period). If hours growth is underestimated, then productivity growth will be overestimated.

Average weekly hours is also an important economic indicator in its own right. Firms can adjust work hours more easily than employment levels. Thus, at the beginning of a recession, firms historically have reduced hours per worker before laying off workers. Similarly, at the end of a recession, firms are often uncertain about the strength of the recovery and tend to increase hours per worker before incurring the cost of hiring additional workers.

The average weekly hours series from the BLS’s main sources of data on hours have both similarities and differences. Average weekly hours for all workers from the household survey (the Current Population Survey) have the highest level and exhibit a relatively flat trend—hovering around 39 hours a week but with a fair bit of cyclical variation. In contrast, weekly hours for production and nonsupervisory workers drawn from the establishment survey (the Current Employment Statistics survey) are lower in level, have less cyclical variation, and exhibit a strong downward trend until the early 1990s, with most of the decline occurring before 1982. The BLS’s Office of Productivity and Technology private nonfarm business hours series, which is derived from the establishment survey production worker series combined with data from the household survey and the National Compensation Survey, behaves much like the establishment survey production worker series. This is not too surprising since the production worker series is its primary data source.

Another source of data on work hours is the American Time Use Survey. The BLS publishes estimates of time spent working, but does not publish an official hours series. It is discussed here mainly because researchers have used microdata from time-use surveys to assess the quality of hours data from household surveys. The behavior of this series is similar to that of the hours series from the household survey.

Discussion of pros and cons

This section discusses the results of research that has examined the differences in these series with the goal of helping policymakers evaluate these important economic data. The series have different strengths and weaknesses, which need to be taken into account in deciding which to use.

A brief description of the hours series

The Current Population Survey is a monthly survey that collects information about employment status and hours worked from a sample of 60,000 households each month. Respondents are asked about usual and actual hours worked on their main and secondary jobs. For actual hours worked, respondents are usually asked to report for the week that includes the 12th of the month, which was chosen to minimize the effect of holidays. However, the November and December survey reference weeks are sometimes moved to the week of the 5th due to the impact of the November and December holidays on data collection or survey operations. These individual monthly reports are averaged to arrive at an estimate of average weekly hours worked. Although the survey collects hours worked on each job separately, the published estimates are on a per-employed-person basis.

The Current Employment Statistics survey collects information, also monthly, from business establishments on employment and hours paidfor the pay period that includes the 12th of the month. Until recently, the establishment survey collected hours data only for production workers (in goods-producing industries) and nonsupervisory workers (in services-providing industries); the series derived from these data is referred to as the production worker series. In 2006, the BLS started publishing an all-employee average weekly hours series in addition to the production worker series. The establishment survey production worker series has a much longer time series and it is the official hours series of the BLS. The two establishment survey series move together, but the all-employee hours series is higher in level by about three-quarters of an hour per week compared to the production worker series, which implies that average weekly hours are about 3.5 hours greater for nonproduction and supervisory workers than for production and nonsupervisory workers (nonproduction and supervisory workers are about 20% of total employment in the establishment survey).

As in the household survey series, average weekly hours from the establishment survey are calculated as total hours divided by total employment. However, because employment in the survey is establishment based, the denominator is the total number of jobs rather than the total number of employed people. Thus multiple jobholders are counted once in the household series but multiple times in the establishment survey. There are also important differences in coverage between the hours series from the household and establishment surveys. The establishment survey provides hours estimates for the total private sector, while the household survey covers all employed individuals. The household survey series includes government workers, private household workers, unincorporated self-employed workers, and unpaid family members—all of which are excluded from the establishment survey hours series.

The productivity series produced by the BLS Office of Productivity and Technology does not come from a separate survey but is constructed by combining data from three surveys—the Current Employment Statistics survey, the Current Population Survey, and the National Compensation Survey. Productivity measurement requires data on hours worked that cover all private nonfarm workers and have industry definitions that are consistent with those in the output data. Neither the establishment survey data nor the household survey data are ideal. Industry classifications in the establishment survey are more consistent with those used in the surveys that collect output data, but the hours data are hours paid and cover only production and nonsupervisory workers. The household survey collects data on hours worked by all workers but does not use industry classifications that are consistent with those in the output data sources. It is important that industry definitions be consistent for outputs and the inputs used to create them. Otherwise, industry productivity measures will be biased, with the extent of the bias depending on the amount of misclassification and the difference in growth rates.

The primary data source for the Office of Productivity and Technology productivity series is the establishment survey’s production worker hours series (the all-employee hours series is not used because the time series is too short). The data are adjusted from an hours-paid to an hours-worked basis using industry-level ratios of hours worked to hours paid calculated from National Compensation Survey data. These hours-worked-to-hours-paid ratios capture changes in the amount of annual leave granted and the amount of sick leave taken. This adjustment produces an estimate of production and nonsupervisory worker hours worked. Because of this adjustment, the productivity hours series shows fewer hours worked than the establishment production worker series, even though the productivity series includes nonproduction and supervisory workers while the establishment series does not.

Average weekly hours worked for nonproduction and supervisory workers are estimated using data from the household survey. The household survey data are adjusted to convert them from a per-person basis to a per-job basis, and jobs that are not covered by the establishment survey are dropped from the sample (for example, government workers, the self-employed, private household workers, and unpaid family members). The ratio of the average weekly hours of nonproduction and supervisory workers to those of production and nonsupervisory workers is calculated and then multiplied by production worker average weekly hours worked to arrive at an estimate of nonproduction and supervisory worker hours. Average weekly hours for all private wage and salary workers are equal to the weighted average of production worker and nonproduction worker hours worked [2]. To fill in remaining gaps, the data on wage and salary workers are supplemented with household survey data on unincorporated self-employed workers, employees of government enterprises (for example, the Postal Service), and unpaid family members.

Coverage of the productivity hours series differs from the household and establishment survey series because it is driven by the data needed for productivity estimates. The productivity hours series excludes government and nonprofit organizations because outputs for these sectors are derived from inputs (specifically, wages and salaries), which makes productivity estimates meaningless. Private household workers are excluded because there are no output data for this sector.

Another source of data on work hours is the American Time Use Survey, which is a time-diary survey that collects information about how people spend their time and can be used to generate estimates of average weekly hours worked. The BLS does not publish an hours series based on these data, but they are useful for evaluating the accuracy of hours-worked data from the household survey.

Figure 1 summarizes the key differences among the three BLS surveys that collect information on work hours: the household survey, the establishment survey, and the time-use survey.

Comparison of US Bureau of Labor Statistics
                        data sources on work hours

Reconciling differences

A natural first step in reconciling differences in work hour estimates between the household and establishment surveys is to examine the accuracy of hours data from the household survey using data from the time-use survey. Research using two methodologies has found that household survey respondents report their total work hours correctly on average, although some demographic groups tend to overestimate hours (for example, college graduates, full-time workers, and women) while others underestimate hours (for example, high school graduates and part-time workers) [1][3][4][5].

The research also found that the household survey reference period is not representative of the entire month—workers work longer hours during household survey reference weeks [1]. About one-third of the difference between reference and nonreference weeks was due to the exclusion of holidays from reference weeks. Because the household survey reference weeks are chosen to avoid holidays (to make it easier to interpret month-to-month changes), extrapolating published household series estimates to the entire month overestimates hours worked during the month, even though respondents correctly report their hours for the reference weeks.

Looking at main and second jobs separately, the research found that weekly hours from the household survey are correctly reported on average for the main job but not for second jobs [4][5]. Specifically, household survey respondents under-report the incidence of second jobs (about 5–6% of workers in the household survey report holding more than one job, compared with about 10–11% in the time-use survey) and over-report hours worked on second jobs (about 14 hours a week compared with about 9 hours). It is likely that the higher incidence of multiple jobholding in the time-use survey is due to self-reporting (compared with mostly proxy reporting in the household survey), which can result in greater reporting of low-hour second jobs. The net effect of this misreporting is that average weekly hours worked on all jobs is overestimated by 0.2–0.4 hours per job per week [4].

Studies making similar comparisons for the UK and Denmark come to similar conclusions about the accuracy of hours reports from household surveys [6][7]. Hours reports from the UK’s household surveys are found to be accurate at an aggregate level, and only minor differences are found between Danish household survey estimates and those from time diaries.

Some of the differences between the household and establishment survey hours series are due to the differences noted in the previous section. For example, research simulating the establishment survey hours data using household survey data (adjusted to establishment survey concepts) found that the difference in levels between the two series is due almost entirely to differences in coverage (all workers in the household survey, production and nonsupervisory workers only in the establishment survey), in the treatment of multiple jobholders (counted once in the household survey, multiple times in the establishment survey), and in the hours concept (hours worked in the household survey, hours paid in the establishment survey) [5].

Other differences were not as easy to reconcile. Although the simulated establishment production and nonsupervisory worker series replicated the actual establishment survey series in levels, it did not replicate the downward trend in the establishment survey hours series [5]. In fact, the simulated establishment survey series was nearly parallel to the original household survey series, but at a lower level of hours worked. There were two periods when the trends in the two series diverged: between 1984 and about 1991 and, to a lesser extent, after 2003 (when measured on a per-job basis). The establishment survey’s sample expanded considerably during the 1980s, but there are no data to determine whether the expansion played a part in the decline in the establishment survey series average weekly hours [5].

The effects of other, more subtle, differences between the series also were examined [5]. For example, the greater number of hours worked during the week of the 12th combined with the lengthening of pay periods in the establishment survey between the early 1980s and the mid-2000s could have imparted a slight downward trend because longer pay periods include more low-hour weeks. For the post-1998 divergence, most of the differences in overall hours trends were found to be due to differences in the trends in three industries (retail trade, leisure and hospitality, and professional and business services) rather than to differences in the distribution of workers across industries (composition effects).

Cyclical behavior

As predicted by theory, all of the BLS hours series exhibit procyclical behavior, although the cyclical patterns are more pronounced in some series than in others (Figure 2). It is easier to see the cyclical pattern in the household series because there is no upward or downward long-term trend. Average weekly hours decline during recessions and rise during expansions. The downward trends of the establishment survey series and the productivity series before 1982 make it more difficult to see how these series vary over the business cycle in those early years. During this period, the rate of decline accelerated during recessions, and the increases after recessions were moderated somewhat by the long-term downward trend. The increases in hours following the last three recessions were much slower than the increase after the 1980–1982 recessions.

Alternative US Bureau of Labor Statistics
                        measures of average weekly hours: quarterly, seasonally adjusted

A study comparing hours data from the time-use survey with data from the household survey and the establishment survey finds that the time-use survey hours are more cyclical than establishment survey hours but less cyclical than household survey hours [8]. While the study does not directly compare household and establishment survey hours series, it can be inferred that hours from the household survey are more cyclical than hours from the establishment survey.

Other researchers have examined the effect of standard-workweek reporting for salaried workers in the establishment survey on the cyclicality of the establishment survey hours series. The study simulated the establishment survey hours-paid concept for salaried workers using household survey data and compared that series to the household survey hours-worked series [9]. It found that the hours-paid series varies less with the business cycle than the hours-worked series but that the two series tell the same story about long-term changes in average weekly hours.

Limitations and gaps

The BLS collects hours data from household and establishment surveys, and each source of data has its advantages and disadvantages.

The BLS’s household survey (the Current Population Survey) allows for comparisons across demographic groups. But some groups appear to over-report hours in the household survey, while others under-report. Further, the small sample size limits the amount of industry detail, and the industry definitions in the household surveys do not exactly match the definitions in establishment surveys that collect output data, both of which are important for measuring productivity.

The BLS’s establishment survey (the Current Employment Statistics survey) collects data on hours paid, which differ from hours worked because of paid leave and off-the-clock work by salaried workers. And, until recently, the establishment survey collected hours data only for production and nonsupervisory workers. However, a major advantage of the establishment survey is its large sample, which allows for more industry detail in published estimates. Moreover, industry definitions in the establishment survey are more consistent with those used in surveys that collect output data.

The American Time Use Survey is a newer source of hours data. Because hours reports from time-diary surveys are generally considered to be accurate, it is a useful data source for validating data from the household surveys. But its small sample and infrequent publication (annually) limits its usefulness as a primary source of hours data.

One advantage of the household survey over the other sources of hours data is the high response rate—about 90%, with very low item nonresponse for the hours questions. In contrast, the response rate for the hours questions in the establishment survey is only 32%, due mainly to high item nonresponse. The response rate for the all-employee count is about 70%, but only 45% of those respondents provide information about payroll and hours. It is not possible to directly assess the effect of nonresponse on the quality of the establishment survey hours data because no benchmark data are available. The response rate for the time-use survey is higher than that of the establishment survey, but still well below that of the household survey. However, research suggests that nonresponse imparts minimal bias to estimates of hours worked in the time-use survey [10].

This article has focused mainly on the US because, apart from the studies mentioned earlier, there does not appear to be any research examining the accuracy of hours data in countries other than the US. Because survey methods and estimation procedures likely vary from country to country, the results summarized here may not generalize to other countries’ hours data. For example, whether there is a reference period effect (as there is in the US) depends on the reference period used and how work hours are distributed. However, these studies highlight some of the issues that might arise when collecting data on work hours and illustrate the types of research projects other national statistical agencies might undertake to assess the quality of their data on hours worked.

Summary and policy advice

Given the difficulty of measuring work hours, it is important for decision makers to understand the advantages and limitations of the different BLS hours series.

For comparing work hours across demographic groups, the Current Population Survey data are the most appropriate. The household survey can also provide information about hours worked by workers not covered by the establishment survey. The American Time Use Survey data also have demographic information, but the small sample size is a limitation, especially for comparing groups over time. And since the BLS does not publish hours estimates from the time-use survey, it is generally necessary to calculate estimates from the microdata. The time-use survey can, however, shed light on the timing of work, both by time of day and across days of the week.

In contrast, the establishment survey and productivity hours series are preferred for comparing hours across industries. Which series is most appropriate depends on the question being asked. The productivity series provides a more comprehensive measure of hours worked by all private nonfarm business sector workers, whereas the establishment survey series measures hours paid to production and nonsupervisory workers. In March 2006, the BLS began publishing hours data for all employees in addition to the production and nonsupervisory hours series.

All the BLS hours series tell qualitatively similar stories about the cyclical behavior of weekly hours, although they differ quantitatively. But the hours series tell different stories about the long-term trend in weekly hours. It has not been possible to explain all of these differences, so one series cannot be recommended over another for comparing long-term trends. Again, which series is most advantageous depends on the question being asked.


The author thanks US Bureau of Labor Statistics staff, an anonymous referee, and the IZA World of Labor editors for many helpful suggestions on earlier drafts. The paper was produced as part of the author’s work at the US Bureau of Labor Statistics. The article may be copied freely as long as it is properly cited.

Competing interests

The IZA World of Labor project is committed to the IZA Guiding Principles of Research Integrity. The author declares to have observed these Principles.


El cruzado del tarifazo

El presidente Mauricio Macri encabeza la campaña para buscar legitimidad política y social a una estructura tarifaria desproporcionada, injusta y regresiva. Datos duros desmienten sus principales argumentos para justificar el tarifazo.

El argumento más rústico, para justificar el aumento desproporcionado, injusto y regresivo de las tarifas de gas y luz, dice que existe una pauta de consumo de derroche en los hogares. El presidente Mauricio Macri encabeza esa cruzada repitiéndolo una y otra vez cuando no existe evidencia empírica ni respaldo en investigaciones en el área energética para sostener semejante afirmación. Que Macri y su familia vivan con el aire acondicionado prendido todo el año, como informó en su momento su esposa, o que algunos hogares privilegiados calefaccionen sus piletas en invierno no deberían orientar a sentenciar que toda la sociedad dilapida energía como esos representantes de la clase alta. En la defensa enfática de garantizar ganancias extraordinarias a conglomerados energético vía tarifazos, el gobierno carga la responsabilidad de los aumentos en sectores vulnerables, desde clases medias y bajas hasta pequeños y medianos comercios e industrias. El tarifazo sería culpa de ellos por años de despilfarro de los recursos energéticos. Macri lo afirma con la misma firmeza como cuando asegura que la inflación está bajando, sin recibir de sus interlocutores ocasionales la interpelación que no es cierto lo que dice.


Macri también critica la política de subsidios energéticos para buscar legitimidad social a los tarifazos, sin mencionar que ahora toda la sociedad está subsidiando a las petroleras pagando un precio del gas muy por encima del costo de producción. Hace años que la ortodoxia y la heterodoxia conservadora han demonizado a los subsidios a las tarifas de servicios públicos, logrando que gran parte de la sociedad aceptara que era necesario eliminarlos para vivir mejor. Ahora lo están padeciendo en los bolsillos.

La distorsión acerca de la comprensión acerca del alcance de la política de subsidios ha sido inmensa. Gracias a la cuenta de Twitter @elbosnio se descubre la existencia de un documento del FMI que se ocupa de ese tema. El objetivo de ese trabajo fondomonetarista es la de promover la disminución de los subsidios energéticos, pero resulta revelador porque informa que el nivel de subsidios que había hasta el 2015 en Argentina, en comparación internacional, era muy bajo.

En “How Large Are Global Energy Subsidies?”, el Fondo postula la necesidad de disminuirlos, exponiendo argumentos ambientales, fiscales y sociales. Por la cuestión ambiental, advierte que los subsidios, al abaratar el consumo de energías no renovables, incrementan la contaminación y gases de efecto invernadero.  Por el frente fiscal, indica que su financiamiento implica aumento de la deuda pública y más impuestos, además de desplazamiento de gasto público de otras áreas, como salud, educación e infraestructura, que afecta el crecimiento económico. Por el impacto macroeconómico y social, sentencia que “son una forma altamente ineficiente de brindar apoyo a los hogares de bajos ingresos, ya que la mayoría de los beneficios de los subsidios a la energía son, generalmente, capturados por hogares ricos”.

El informe del FMI calcula que los subsidios energéticos sumaban 5,3 billones de dólares en 2015, 6,5 por ciento del PIB mundial, aumentando de la estimación de 4,2 billones de dólares de 2011. El aspecto más interesante de ese documento es el cálculo particular de los subsidios energéticos en los países, mostrando que no eran significativos en Argentina en la comparación. Aquí va el detalle del monto de los subsidios energéticos en algunos países, en 2015, según el FMI:

  • Estados Unidos, 2177 dólares per cápita.
  • China, 1652.
  • Canadá, 1283.
  • Australia, 1259.
  • Japón, 1240.
  • Israel, 1113.
  • Bélgica 909.
  • Alemania, 684.
  • Inglaterra, 635.
  • España, 521.
  • Chile, 515.

El promedio mundial de subsidios a la energía eléctrica era de 640 dólares por habitante en 2015. Argentina subsidiaba a cada habitante con 413 dólares, hasta 2015, el 35 por ciento menos que el promedio mundial.


Quienes estudian con seriedad el tema energético explican que malgastar es muy diferente que el uso eficiente. En el mercado argentino no existe derroche pero sí una utilización deficiente de la energía, por ejemplo por el tipo de construcción de las viviendas y por la calidad de los electrodomésticos. Aunque para muchos es difícil de absorber porque ha habido años de desinformación, y más aún ahora con un Presidente de la Nación que insiste hasta el hartazgo con el tema del dispendio, el consumo energético de los hogares no se define por el nivel de las tarifas. El consumo no es mayor o menor según el precio del gas o de la luz, sino que está definido por la cuestión climática y por la calidad de la infraestructura de la vivienda.

Por el tarifazo no habrá ahorro; y sí, transferencias inmensas de ingresos a las compañías energéticas.

La demanda de gas natural está definida por la temperatura, por eso es mayor en las provincias más frías, y es inelástica respecto de los cambios tarifarios y el nivel de ingreso. Esto último significa que la utilización de gas no sube o disminuye según el nivel de la tarifa y el monto de los ingresos de los hogares. Son conclusiones de la investigación “El consumo de gas natural por red. Análisis territorial y temporal”, de Pedro Chévez, Carlos Discoli e Irene Martini, publicada en “Realidad Económica” N°308.

A partir del análisis realizado de la demanda nacional de gas natural residencia en un período extenso, que abarca de 1993 a 2014, la troika de investigadores

  • explican que el consumo de gas natural demuestra una clara incidencia del factor climático y en consecuencia una relación directa con la ubicación geográfica;
  • destacan que la demanda con un mismo nivel de temperatura no resulta constante a lo largo del territorio, sino que se acrecienta a medida que la rigurosidad del clima es mayor;
  • mencionan que existen diferentes variables tecnológicas y de comportamiento, que conjugadas con situaciones climáticas extremas justifican mayores consumos y que no necesariamente significan sobreconsumos adjudicables a un mal uso del recurso energético;
  • indican que entre las dimensiones y variables consideradas hay que tener en cuenta la mayor permanencia en los hogares debido a las condiciones climáticas, ocasionando una mayor demanda de energía, como así también menores niveles de radiación solar y heliofanía (número de horas reales de insolación diaria), minimizando el efecto de la ganancia energética directa en las viviendas; y
  • afirman que a nivel nacional el consumo promedio de gas natural por usuario acompañó en cierta manera la coyuntura del país sin cambios muy pronunciados, concluyendo que la demanda de gas es inelástica.

O sea, con más o menos poder adquisitivo, el consumo se mantiene constante. Si sube la capacidad de compra por mejoras en términos reales de los salarios o jubilaciones, los hogares incrementan el consumo de otros bienes motorizando la demanda y la expansión del mercado interno. En cambio, si disminuye el salario o el haber jubilatorio real, no se reduce el consumo de gas, sino el de otros bienes porque de esa forma generan el ingreso necesario para pagar un servicio público esencial. El tarifazo afecta en forma negativa de ese modo el nivel de actividad del mercado interno.


Hasta las empresas energéticas, beneficiarias del shock del tarifazo, se alejan de la campaña oficial de confusión y muestran que la cuestión climática es una variable relevante del nivel de consumo de los hogares. El balance anual 2017 de Edenor informa que hubo “un consumo inferior en aproximadamente 3 por ciento respecto al año anterior”, atribuyendo esa disminución “a un invierno poco severo, con temperaturas medias que superaron en 2°C las del año 2016”. El “ahorro” en los hogares no fue por el tarifazo; fue por razones climáticas.

El efecto monetario directo de los aumentos desproporcionados de tarifas estuvo reflejado en los balances de las empresas energéticos. Hubo menos ingresos disponibles en los hogares para otros consumos y estrangulamiento en la estructura de costos de comercios e industrias (ver nota aparte) con la contrapartida del incremento de las ganancias de las empresas energéticas

El recorrido por los ejercicios anuales presentados en la Bolsa de Comercio, eligiendo firmas representativas de diferentes eslabones de la cadena energética, permite observar utilidades generalizadas originadas por el tarifazo.

  • Distribución eléctrica.

Edenor: La reseña del ejercicio 2017 indica que tuvo “mejoras significativas” del cuadro de resultados respecto al año anterior. Los ingresos pasaron de 13.080 a 24.340 millones de pesos; los resultados operativos, de una pérdida de 656 a una ganancia de 2495 millones de pesos; y el resultado neto, de un quebranto de 1189 a una utilidad de 682 millones de pesos. Señala que esos saldos “nos han permitido recomponer el patrimonio, aunque seguimos registrando resultados acumulados negativos”. Informa que en 2017 invirtieron 4137 millones de pesos, un 53 por ciento más que lo ejecutado en 2016, que “a su vez habían resultado en el máximo hecho por la sociedad en toda su historia”.

Edesur: Por razones no explicitadas, el comportamiento de sus resultados es opuesto al de Edenor. El operativo fue positivo en 1546 millones de pesos, menor en comparación con los 2024 millones del ejercicio anterior. El integral total fue una pérdida neta de 1044 millones, superior a la de 275 millones de 2016. Informa que las inversiones alcanzaron los 3559 millones de pesos en obras destinadas “a mejorar la calidad del servicio”. Tiene un capital de trabajo negativo de 10.018 millones de pesos, principalmente como consecuencia de la deuda contraída con Cammesa para financiar la operación de la compañía durante los años pasados. En 2017, los fondos aplicados a inversiones totalizaron 3994 millones de pesos comparados con los 3088 millones de 2016.

  • Transportadoras de gas.

TGN: Tiene 6806 kilómetros de gasoductos, y a través de sus dos troncales (Norte y Centro Oeste) abastece a ocho de nueve distribuidoras de gas y a numerosas generadoras eléctricas e industrias ubicadas en 15 provincias. Registró en 2017 ganancias de 842 millones de pesos en comparación con la pérdida de 259 millones de pesos en 2016, de 518 millones en 2015 y 257 millones en 2014. La variación de las ventas de 2040 millones de pesos entre los ejercicios 2017 y 2016 se explica por una mayor facturación de 1917 millones asociada “al incremento en las tarifas de transporte”, y el resto por servicios de “operación y mantenimiento de gasoductos”.

TGS. Por sus gasoductos, de 9183 kilómetros de extensión en total, transporta el 60 por ciento del gas consumido del país, atendiendo a 5,8 millones de usuarios finales. La ganancia en 2017 fue de 2793 millones de pesos, 200 por ciento más respecto de los resultados netos positivos del año anterior (930,6 millones). En el balance anual indica que el 44 por ciento de la tarifa del servicio público de gas se explica por el precio del gas en boca de pozo, y el restante 56 por ciento se divide en partes iguales entre los impuestos y los segmentos del transporte y la distribución.

  • Generadoras.

Central Puerto. Esta empresa representa el 12,1 por ciento de la generación energética total, y el 17,5 por ciento de la generación térmica. En 2017, la ganancia operativa fue de 3111 millones de pesos, mientras que en el mismo período de 2016 fue de 2100 millones. La ganancia neta, en este caso, a diferencia de la mayoría de los balances, fue superior a la operativa por utilidades obtenidas en los resultados financieros: el saldo positivo en ese renglón del balance fue de 3508 millones de pesos.

  • Distribuidora de gas.

Metrogas: Es la empresa de distribución de gas natural por red más grande del país, con el 18 por ciento del mercado con 2,4 millones de clientes. El resultado neto del año pasado fue una ganancia de 775 millones de pesos, cuando en 2016 había contabilizado una pérdida de 604 millones de pesos.

  • Transporte de electricidad.

Transener. Junto a su controlada Transba anotó una ganancia neta del ejercicio 2017 de 2282 millones de pesos, cuando el año anterior había tenido una pérdida de 57 millones. Posee, opera y mantiene la red de transporte de energía eléctrica de alta tensión (550kV). Transba presta el servicio público de transporte de energía eléctrica de la provincia de Buenos Aires (66kV a 220Kv) por líneas de distribución troncal.


Como lo hizo con la reforma previsional asaltando el ingreso de los jubilados, Macri está lesionando también el de otro sector de su base electoral. Con el tarifazo castiga el presupuesto de los hogares de clases medias urbanas, y lo hace con dedicación para garantizar ganancias extraordinarias a las empresas energéticas.

El tarifazo es desproporcionado en relación al nivel de ingresos de la mayoría de la población. Esto lo dejó en evidencia el propio gobierno con la propuesta de disminuir o eliminar impuestos o tasas de las facturas de servicios públicos. Con esa medida aspira a aliviar el monto final de la factura de luz, gas y agua.

Ese descuento poco y nada cambiará el presupuesto de los hogares, y sólo muestra la desesperación del oficialismo ante la resistencia social y política que provoca el tarifazo. En esa búsqueda de compartir los costos de una medida anti popular abrazó a los gobernadores, que están actuando como aliados políticos, varios de ellos integrantes de una oposición, hasta ahora, colaboracionista con el ajuste.

Los tarifazos de la Alianza Cambiemos están derivando en una situación política muy incómoda para el oficialismo, expresada en las dos cámaras del Congreso. Pero también en una muy crítica por su impacto socioeconómico. La cruzada del tarifazo está ahogando el presupuesto de los hogares de las clases media y baja, que tiene como saldo un aumento de colgados a la red eléctrica: Edenor registra que el 17,1 por ciento del total de la energía entregada al mercado no la puede cobrar, que le significa unos 4000 millones de pesos menos de recaudación, mientras que en Edesur representan el 12,0 por ciento de la energía provista, contabilizando además un incremento de la morosidad en 6 puntos en los dos primeros años del gobierno de Macri respecto a la existente en 2015.

Minuto tecnológico: Amazon y la respuesta sindical global

Hemos comentado en varias ocasiones en este minuto tecnológico, lo que va a suponer la robótica y en general la transformación digital en el mundo del trabajo, de la empresa y de las relaciones laborales. Paro masivo provocado por la destrucciones de puestos de trabajo de personas que pierden sus empleos al ser sustituidas por robots; y también mejoras en las condiciones del resto de los trabajadores que tendrán menos trabajo y más tiempo libre…Todas estas cuestiones las habrán escuchado seguro que más de una vez.

Pero lo paradójico es que de momento nos encontramos con titulares que al menos sorprenden. Voy a coger como ejemplo una de las grandes tecnológicas: Amazon. La empresa que ha revolucionado el comercio. Que en una hora te lleva a tu casa lo que hayas comprado, sea lo que sea. Una de las grandes locomotoras de la Bolsa, que sólo en el primer trimestre ha facturado más de 50.000 millones de dólares y ha ganado más de 1.600 millones. Una empresa que en 2017 contrató ella sola a lo largo del mundo a casi 140.000 trabajadores y que en total tiene ya 560.000. Casi seiscientos mil trabajadores.

Amazon es grande, innovadora, global, moderna, divertida. Y está generando nuevas situaciones laborales, pocas veces antes vistas. Por ejemplo sus trabajadores llevan en los últimos meses denunciando las condiciones laborales que tienen. Y ya han realizado algunas huelgas en varios países. Pero sin entrar en el fondo de esos motivos, ahora se produce un efecto de la globalización. Una movilización y protesta laboral también global. Sindicatos de España, Francia e Italia en contacto con trabajadores del Reino Unido y Alemania buscan un comité de empresa europeo para globalizar la lucha y coordinar acciones. Nuevos entornos también porque las huelgas de los trabajadores tienen que enfrentarse a nuevos retos: competencia de robots y sobre todo a la tiranía de los algoritmos.

Cuatro choques externos que transformarán América Latina

La región se enfrenta a unos años difíciles en los que tendrá que digerir las consecuencias de la revolución digital y el cambio climático

“Muchos años después, frente al pelotón de fusilamiento, el coronel Aureliano Buendía había de recordar aquella tarde remota en que su padre lo llevó a conocer el hielo”.

Este es la primera frase de Cien años de soledadla novela de Gabriel García Márquez que narra la vida de los pocos pero fascinantes habitantes de Macondo, un imaginario pueblo perdido en las selvas de Colombia. Cada cierto tiempo, a Macondo lo sacude lo que los economistas, politólogos y sociólogos de hoy llaman un “choque externo”. Es un invento foráneo que trastoca la economía, las relaciones de poder, los hábitos y costumbres, en fin, la vida del pueblo. Así, además del hielo, a Macondo lo sacudieron la llegada de los imanes, de la pianola, de ideas políticas en conflicto que dispararon una larga y sangrienta guerra civil, o la bananera, una empresa multinacional más poderosa que todos los Gobiernos y todos los ejércitos. La novela es, por supuesto, mucho más que el relato de los “choques externos” que le cambian la vida a la gente del pueblo. Pero las convulsiones de Macondo ofrecen una excelente metáfora para discutir las sacudidas que le vienen a América Latina.

Esta ha sido siempre la región con las economías más volátiles del mundo. Periódicamente, una etapa de expansión y prosperidad es súbitamente reemplazada por otra de estancamiento y miseria.

Según la ONU América Latina es una de las zonas más vulnerables al cambio climático

Estos ciclos de auge y caída suelen estar determinados por los precios internacionales de las materias primas que exporta la región, y por la disponibilidad de préstamos e inversiones que vienen de afuera. Cuando los precios del petróleo, cobre, café, soja, etcétera, suben en el mercado mundial, América Latina prospera. Cuando caen, se empobrece. Cuando los bancos y empresas extranjeras invierten y abren el crédito, las economías latinoamericanas mejoran. Pero cuando los préstamos e inversiones foráneas cesan (y con frecuencia eso sucede al mismo tiempo que bajan los precios de las exportaciones) viene la debacle: devaluación, inflación, desempleo, suspensión de programas sociales y quiebras de bancos y empresas. Naturalmente, los Gobiernos latinoamericanos también son responsables por no hacer que sus economías sean menos vulnerables a las oscilaciones internacionales. Pero es justo reconocer que no es fácil neutralizar el impacto de un masivo choque económico externo.

Los nuevos choques externos

En los próximos años, a las sacudidas económicas que periódicamente agitan América Latina se le van añadir otras. El cambio climático, la revolución digital, una nueva intolerancia a ciertos problemas sociales o la revolucionada política mundial serán las fuentes de estos potentes e inéditos choques externos.

El cambio climático

Ninguna región escapará a sus efectos. Pero según la Organización de Naciones Unidas, América Latina es una de las más vulnerables a los accidentes climáticosque seguirán aumentando en frecuencia, fuerza, fatalidades y costos. Las razones de esta alta vulnerabilidad van de la geografía al clima, de las condiciones socioeconómicas a la demografía. Es la zona más urbanizada del planeta: el 80% de sus habitantes viven en ciudades, la gran mayoría de ellos son pobres y sus viviendas, muy precarias. La corrupción también agudiza la fragilidad de la región ante el cambio climático. Es frecuente, por ejemplo, que funcionarios venales autoricen construcciones en lugares inadecuados o que hagan la vista gorda con las violaciones de normas urbanísticas a cambio de sobornos.

El cambio climático traerá los choques externos más transformadores que ha vivido América Latina. Cambiarán dónde y de qué viven los latinoamericanos, lo que producen y lo que gastan. O a qué conflictos domésticos e internacionales tendrán que enfrentarse.

La revolución digital

Inteligencia artificial, big data, robótica, blockchain, computación cuántica y redes neuronales son solo algunos de los campos en los que se dan las revoluciones tecnológicas que van a cambiar el mundo. Es fácil intuir que se nos avecinan enormes transformaciones. Pero lo que más nos sorprenderá es la rapidez con la que se harán notar sus efectos.

La posibilidades que abren estas nuevas tecnologías son maravillosas. Pero también son enormes los problemas que plantean. Un importante efecto indeseable de la revolución digital es que puede destruir muchos puestos de trabajo existentes, antes de crear otros nuevos. Eso también va a pasar en todas partes. Pero en América Latina el impacto sobre el mercado laboral será más fuerte. Según la ONU, en las próximas décadas dos de cada tres empleos formales en Latinoamérica serán automatizados. El choque externo producido por la revolución digital puede ser tan determinante como el del cambio climático

Nueva intolerancia a viejos males

La desigualdad económica y la corrupción son dos plagas perennes en Latinoamérica. A pesar de que su erradicación ha sido siempre la promesa de populistas y revolucionarios, en la práctica las sociedades las aceptaban como realidades inevitables. Recientemente, esto ha comenzado a cambiar. La coexistencia pacífica de los latinoamericanos con la corrupción y la desigualdad se está acabando. Entre otras razones, el aumento de la desigualdad en Estados Unidos y Europa ha creado un intenso debate mundial que ha reactivado esa discusión en América Latina. Lo mismo sucede con la corrupción. Las fechorías de los corruptos, que siempre han existido, ahora se han hecho más visibles e inaceptables. Las nuevas clases medias, más numerosas, educadas, informadas y conectadas, se han activado y están hartas de los desmanes y de la impunidad. Guatemala, Brasil y Perú se han unido a la lista de países como Corea del Sur, Ucrania, Arabia Saudí y China, donde las acusaciones de corrupción han llevado a la cárcel a políticos y empresarios antes intocables.

Los choques externos son una amenaza creciente para América Latina. Pero no tanto como las amenazas que constituyen la complacencia y la falta de previsión

Si bien estas acciones son locales, en muchos casos los estímulos que prenden la mecha vienen de afuera. El escándalo de la empresa Odebrecht, por ejemplo, es brasileño pero ha resultado en un choque externo que ha convulsionado la política de muchos países de América Latina. La inestabilidad producida por la lucha de la gente contra la corrupción y la desigualdad va a seguir.

La política

A finales del siglo XIX, un periodo tan caótico como el actual, el pensador italiano Antonio Gramsci escribió: “El viejo mundo se está muriendo. El nuevo tarda en llegar. En ese claroscuro se ceban los monstruos”. Esta frase capta muy bien lo que está sucediendo hoy en el mundo. Es natural que estos nuevos monstruos foráneos también hagan de las suyas en América Latina. Para eso no solo se aprovecharán de las oportunidades que les ofrecen las nuevas relaciones de poder dentro y fuera de la región, sino también de las nuevas tecnologías. Internet está siendo utilizado para polarizar, destruir reputaciones e influir en procesos electorales. El impacto en sociedades aprensivas, que ya están crispadas y confundidas, puede ser enorme. Los protagonistas de siempre pierden influencia y son reemplazados por recién llegados que alteran la agenda y actúan de manera imprevisible. Nuevos caudillos mundiales, como Xi Jinping o Vladímir Putin, mueven fronteras y cambian reglas. Donald Trump desestabiliza a su país y al mundo. ¿A quién creer? ¿Qué es verdad? ¿En quién confiar?

La política es siempre muy local, pero ahora lo local se mezcla con lo global casi instantáneamente, creando contundentes choques políticos externos. Si la principal cadena de transmisión de este tipo de choques externos son las redes sociales, entonces América Latina es, de nuevo, la región más vulnerable. Según un estudio de ComScore, los latinoamericanos pasan en las redes sociales el 29% del tiempo que están en Internet. En ninguna otra parte el tiempo en redes sociales es tan alto. En EE UU es el 14%, y en Asia el 8%, por ejemplo.

Las buenas noticias

Los choques externos que impactarán a Latinoamérica tendrán altos costos, pero también abrirán oportunidades inéditas. Y los latinoamericanos son expertos en sobrevivir a la volatilidad. Llevan décadas, sino siglos, adaptándose, esquivando, improvisando y manejando los efectos de los choques externos. Las empresas latinoamericanas de hoy son ágiles y eficaces sobrevivientes de los periódicos revolcones que súbitamente cambian las reglas del juego.

Los Gobiernos latinoamericanos también son responsables de no hacer que sus economías sean menos vulnerables

Además, estos cuatro choques externos también tienen aspectos positivos. La nueva intolerancia con la desigualdad y la corrupción es una buena noticia. El cambio climático trae catástrofes, pero también cambios en los ciclos agrícolas que pueden aumentar la productividad y el rendimiento de las cosechas. Habrá una inmensa demanda de nuevas industrias especializadas en la adaptación a los cambios del clima o la mitigación de sus riesgos.

Lo mismo vale para la revolución digital. Las nuevas tecnologías seguramente crearán nuevos mercados y reducirán las barreras que tanto inhiben la competencia en el sector privado. Y si bien la polarización y la crispación política de la sociedades crea las peligrosas convulsiones políticas de las que se aprovechan los populistas, también es cierto que puede estar abriendo espacios para nuevos líderes democráticos que traigan otras propuestas.

Los choques externos son una amenaza creciente para América Latina. Pero no tanto como las amenazas que constituyen la complacencia y la falta de previsión.

El más de lo mismo ya no va a funcionar. Quienes antes rompan con el conformismo y salgan de su zona de comodidad tendrán más oportunidades de evitar que los choques externos los arrollen.

El sindicalista corrupto no va luchar por intereses verdaderos


Alejandro Argueta, abogado de profesión especializado en Derecho Laboral, se ha enfocado en el estudio y atención de casos relacionados a la defensa de derechos de trabajadores, las organizaciones sindicales, pactos colectivos del sector público y privado. El abogado, también con conocimientos y estudios en Historia y Economía, conversó con “La Hora” respecto a las consecuencias que existen a partir de la corrupción de organizaciones de sindicatos en el sector público y de la necesidad de eliminar estas malas prácticas en favor de la gestión de las instituciones públicas.

Argueta analizó e identificó anomalías en el Pacto Colectivo firmado entre autoridades del Ministerio de Salud Pública y el sindicato de salubristas de esa institución. A criterio del experto, existe una brecha sumamente ancha entre la verdadera naturaleza de los sindicatos y la forma en que han funcionado, en la práctica. Según Argueta, en ese juego de corrupción y defensa de intereses personales, los más afectados son los trabajadores honestos y la población.

¿Cuál es o debería ser la naturaleza de un sindicato?
Los sindicatos son muy importantes. En el ámbito laboral y profesional, la única forma de lograr mejorar y superarse es organizándose, en lo individual hay muy pocas historias de éxito, creer que solo uno va lograr las cosas no se puede, entonces hay que organizarse. Un sindicato siempre debe buscar el bienestar de los trabajadores. Hay muchos casos que demuestran la necesidad de trabajar y proponer en colectivo. Una segunda función de los sindicatos es que son un interlocutor político, porque para hacer política es importante tener representantes políticos y legítimos.

¿Cómo son los sindicatos en Guatemala?
En Guatemala, los sindicatos actualmente son el producto de una política de represión antisindical. En lo personal hay muchos que no me agradan, son abusivos y son inconsecuentes, tienen todos los atributos negativos para no ser referentes de un verdadero movimiento sindical. No es culpa de los trabajadores, la destrucción de la fuerza sindical ha sido una causa de algunos patronos.

¿Quién es el principal beneficiado con la corrupción de un sindicalista?
Es el patrono porque si el sindicalista fuera honesto a la representación que ejerce entonces el patrono realmente tendría que negociar con una fuerza colectiva real y ceder parte de sus intereses económicos y políticos con tal de mantener la estabilidad de su actividad. Por el contrario, si el sindicalista es controlado por la presión del patrono, lo que hace es que deja de ser sindicalista y se convierte en un cómplice del patrono. Hay una frase que dice “no se puede vivir de defender una injusticia y pretender no ser parte de ella”. Entonces, cuando el sindicalista traiciona su representatividad y se presta a intereses opacos del patrono, se convierte en cómplice del patrono. Un patrono legal, responsable y honesto no necesita de un sindicalista corrupto. El sindicalista corrupto no va luchar por intereses verdaderos.

¿Cómo se aplica esto a las instituciones públicas?
En las instituciones públicas, lo que puede suceder es que los funcionarios corruptos corrompen al sindicalista para que en la traición de sus intereses lo primero que traicionen sea ser el primer testigo de los actos de corrupción del funcionario. Cuando en realidad ellos tendrían que ser los primeros en denunciar la corrupción de las instituciones. O sea que lo primero que gana es el silencio del sindicalista de la inapropiada gestión del funcionario.

Un sindicalista estatal conoce bien las instituciones, lo saben todo, tienen el sentido de un funcionario correcto y uno que no está actuando bien, saben sobre los procesos administrativos, sobre las compras, las presiones financieras, se enteran de todo. Cuando lo ocultan es donde traicionan la esencia de su papel como sindicalistas. La corrupción que vemos ahora es el resultado y producto de un deterioro de la administración pública que empieza con los funcionarios. La corrupción es la otra cara de la moneda a la represión. Te pega o te paga.

¿Cuáles son las consecuencias de esa dinámica de corrupción entre sindicalistas y funcionarios?
Lo que sucede es que aunque existen muchos sindicatos en las instituciones, en lo público o en lo privado, en realidad los trabajadores no están representados, lo que existen son sindicatos de cartón, son fachadas políticas que ejercen una función de validación de la corrupción y han dejado de ser representativos para el resto de trabajadores.

¿De qué forma se puede resolver esto?
Considero que lo primordial es dar oportunidades de fomentar nuevos liderazgos y sobre todo de jóvenes que estén conscientes de las deficiencias actuales y tengan la intención de hacer las cosas de manera distinta.

También considero que se deben aprovechar las coyunturas actuales de la lucha contra la corrupción y aprovechar que hay de parte del sector empresarial y estatal personas que comprenden la necesidad de apoyar estos procesos.

Actualmente resultan muy cuestionados los Pactos Colectivos. ¿De qué forman funcionan estas negociaciones?
El Pacto Colectivo es la formalización de los acuerdos entre los sindicalistas y los representantes de las instituciones, el problema es el tipo de negociación. Hasta ahora no tengo evidencia para hacer una excepción respecto a las anomalías en las negociaciones.

El problema con los Pactos Colectivos es que, con el modelo con el que se plantean, no son sostenibles. La forma en que se negocian entre instituciones públicas y sindicalistas, hacen que los beneficios no sean sostenibles. En el caso de los salubristas hay que reconocer que sí ganan muy poco, los ingresos son bajos para los que están haciendo gestión púbica. Y entonces vienen este tipo de prácticas entre sindicalistas y funcionarios públicos y llegan a acuerdos de beneficios económicos que aparentan ser favorables pero no son sostenibles y al final los trabajadores salen defraudados porque al final no reciben nada.

No les pagan porque no hay respaldo legal para hacerlo, no hay suficiente capacidad presupuestaria, otras veces, la mayoría de beneficios deben ser erradicados porque no son sostenibles.

¿Cuál es el perfil que debe tener un sindicalista?
Yo considero que antes que cualquier cosa, un sindicalista debe ser responsable y formal en el cumplimiento de sus funciones legales como empleado públicos y a partir de ahí proponer una mejora. No al revés, no proponer cosas que les quiten responsabilidades como empleados públicos porque de esa forma afectan la gestión pública y esa no es la verdadera intención de un sindicalista.

Un sindicalista debe identificarse con la gestión y a partir de esa identificación y proyección del bien común, exigir mejores condiciones de vida. Así es como me imagino un sindicato: intransigente con las prácticas opacas, exigente de condiciones de trabajo que le dignifiquen y le den carrera.

¿En algún momento, los sindicatos se vuelven herramientas políticas?
Se convierten en grupos de poder. Primero, las personas los apoyan y creen que van a lograr beneficios insostenibles con la representación de los sindicalistas, buscan un alivio a su precaria condición de vida. Con esa intención, la gente apoya la sindicalización. La gente participa de una u otra manera, pero con ese apoyo y con el uso de licencias se convierten en una herramienta de presión política para intereses muy particulares que muchas veces no tienen que ver con los intereses de los empleados.

Tomando en cuenta que existen anomalías en la actividad sindical, ¿quiénes deberían vigilar el actuar de los sindicatos?
Todos. Es una tarea que va desde el Presidente, los ministros, las organizaciones como la Contraloría General de Cuentas, la Oficina Nacional de Servicio Civil, el Ministerio Público, los tribunales, el Ministerio de trabajo. Prácticamente es una tarea de muchas instituciones. Son varias las autoridades que deben poner atención en este tema para resolver lo que está sucediendo.

La mejor solución para empezar a desatar el nudo es atacar la corrupción. Esto no se trata de criticar el movimiento sindical en general sino a la corrupción que se ha impregnado en estas actividades.

En lo personal hay muchos que no me agradan, son abusivos y son inconsecuentes, tienen todos los atributos negativos para no ser referentes de un verdadero movimiento sindical.

La corrupción es la otra cara de la moneda a la represión. Te pega o te paga.

La mejor solución para empezar a desatar el nudo es atacar la corrupción.

Un sindicalista debe identificarse con la gestión y a partir de esa identificación y proyección del bien común, exigir mejores condiciones de vida.

Alejandro Argueta, abogado especializado en derecho laboral.

Los países del mundo en los que se trabaja más horas (y los dos primeros son de América Latina)

La Asamblea Nacional de Corea del Sur aprobó en marzo una ley que le dará a una cantidad importante de su fuerza laboral un merecido descanso.

Los legisladores aprobaron de manera abrumadora un proyecto de ley que reduce el máximo de horas de trabajo semanales a 52, de las 68 horas que se permitían antes.

Corea del Sur es la nación desarrollada que tiene la jornada laboral más larga, según la Organización para la Cooperación y el Desarrollo Económico (OCDE).

La ley entrará en vigor en julio, aunque inicialmente solo se aplicará en las grandes empresas y luego llegará a compañías más pequeñas.

A pesar de la oposición de la comunidad empresarial, el gobierno cree que la ley es necesaria para mejorar la calidad de vida, crear más empleos y aumentar la productividad.

Excepción a la regla

Corea del Sur tiene una jornada laboral más extensa que cualquier otro país desarrollado: un promedio de 2.069 horas por año, por trabajador.

Pero aunque es el país desarrollado que más trabaja, hay dos naciones en vías de desarrollo en el que se trabaja aún más horas. Y ambas están en América Latina.

El análisis de la OCDE, que incluyó a 38 países, mostró que los mexicanos (2.225 horas/año) y los costarricenses (2.212) tienen las jornadas laborales más extensas del mundo.

Gráfico de los países con jornadas laborales más extensas

Estudios llevados a cabo por la Organización Internacional del Trabajo (OIT) muestran que los países de ingresos bajos y medios tienden a trabajar más horas que sus contrapartes más ricos.

Esto se debe a una serie de factores que van desde la proporción de trabajadores cuentapropistas o autónomos a los salarios más bajos, la inseguridad laboral y cuestiones culturales.

Pero Corea del Sur no es la única excepción a la regla.

Una minera en Sierra LeonaDerechos de autor de la imagenGETTY IMAGES
Image captionVarios países del mundo no tienen límites legales para la jornada laboral.

Japón tiene un problema con la “muerte por exceso de trabajo“, algo que no solo está expresado por las estadísticas sino también por el hecho de que el idioma japonés tiene una palabra para este tipo de fallecimiento: karoshi.

La palabra se refiere a los empleados que mueren a causa de dolencias relacionadas con el estrés (ataques cardíacos, derrames cerebrales) o que se quitan la vida debido a las presiones del trabajo.

El promedio de 1.713 horas trabajadas por año en Japón no está entre las más altas de la lista de la OCDE, pero más allá del número, existe la sombría realidad de que el país no cuenta con una legislación que estipule un límite máximo de horas semanales y no hay ningún límite para las horas extras.

Empresario japonés dormido en una estaciónDerechos de autor de la imagenGETTY IMAGES
Image captionTrabajar de más es algo habitual en Japón.

En el año fiscal 2015-16, el gobierno registró un récord de 1.456 casos de karoshi.

Los grupos defensores de los derechos de los trabajadores afirman que las cifras reales podrían ser mucho más altas debido al subregistro.

Países con la semana laboral legal más extensa
Tailandia 84
Islas Seychelles 74
Costa Rica 72
Nepal 68
Irán 64
Malasia 62
Singapur 61

Según las cifras más recientes de la OIT, Asia es el continente donde el mayor número de personas trabaja la mayor cantidad de horas: la mayoría de los países asiáticos (el 32%) no tienen un límite máximo para la jornada laboral.

Otro 29% tiene umbrales altos (60 horas semanales o más). Y solo el 4% de los países cumplen con las recomendaciones de la OIT y acatan las normas laborales internacionales que establecen un máximo de 48 horas o menos para la semana laboral.

En las Américas y el Caribe, el 34% de las naciones no tienen límite de horas semanales, la tasa más alta entre todas las regiones.

Surcoreanos cruzando una calle.Derechos de autor de la imagenGETTY IMAGES
Image captionLos surcoreanos trabajaron un promedio de 2.069 horas el año pasado.

Uno de los países sin límite es Estados Unidos.

Pero es en Medio Oriente donde los límites legales son más extensos: ocho de cada 10 países permiten jornadas laborales que exceden las 60 horas por semana.

En Europa, en tanto, todos los países tienen un horario semanal máximo y solo Bélgica y Turquía tienen jornadas laborales legales de más de 48 horas.

Pero es en África donde hay más países en los que más de un tercio de la fuerza laboral trabaja más de 48 horas por semana. La tasa en Tanzania, por ejemplo, es de 60.

Ciudades “adictas al trabajo”

También hay encuestas que muestran en qué ciudades se trabaja más.

En 2016, Swiss Bank UBS publicó un análisis de 71 ciudades que mostraban a Hong Kong con un promedio de 50,1 horas semanales de trabajo, por delante de Bombay(43.7), Ciudad de México (43.5), Nueva Delhi (42.6) y Bangkok (42.1).

Hong KongDerechos de autor de la imagenGETTY IMAGES
Image captionUna encuesta de 2016 halló que Hong Kong es la ciudad en la que se trabaja más horas.

Los mexicanos, además de trabajar las jornadas más largas, también están sujetos a uno de los regímenes vacacionales más mezquinos del mundo: sus vacaciones mínimas pagas legales son de menos de 10 días, lo mismo que ocurre en Nigeria, Japón y China, por ejemplo.

En cambio Brasil ofrece un mínimo de entre 20 a 23 días.

Podría ser peor, sin embargo.

En India, donde no existe un límite para el horario laboral máximo, los trabajadores no tienen una cantidad mínima garantizada de vacaciones pagas anuales.

How much is an hour worth? The war over the minimum wage

Some economists say it should be raised. Others say it’s already too high. But what if both sides are missing the point? By 

No idea in economics provokes more furious argument than the minimum wage. Every time a government debates whether to raise the lowest amount it is legal to pay for an hour of labour, a bitter and emotional battle is sure to follow – rife with charges of ignorance, cruelty and ideological bias. In order to understand this fight, it is necessary to understand that every minimum-wage law is about more than just money. To dictate how much a company must pay its workers is to tinker with the beating heart of the employer-employee relationship, a central component of life under capitalism. This is why the dispute over these laws and their effects – which has raged for decades – is so acrimonious: it is ultimately a clash between competing visions of politics and economics.

In the media, this debate almost always has two clearly defined sides. Those who support minimum-wage increases argue that when businesses are forced to pay a higher rate to workers on the lowest wages, those workers will earn more and have better lives as a result. Opponents of the minimum wage argue that increasing it will actually hurt low-wage workers: when labour becomes more expensive, they insist, businesses will purchase less of it. If minimum wages go up, some workers will lose their jobs, and others will lose hours in jobs they already have. Thanks to government intervention in the market, according to this argument, the workers struggling most will end up struggling even more.

This debate has flared up with new ferocity over the past year, as both sides have trained their firepower on the city of Seattle – where labour activists have won some of the most dramatic minimum-wage increases in decades, hiking the hourly pay for thousands of workers from $9.47 to $15, with future increases automatically pegged to inflation. Seattle’s $15 is the highest minimum wage in the US, and over double the federal minimum of $7.25. This fact alone guaranteed that partisans from both sides of the great minimum-wage debate would be watching closely to see what happened.

But what turned the Seattle minimum wage into national news – and the subject of hundreds of articles – wasn’t just the hourly rate. It was a controversial, inconclusive verdict on the impact of the new law – or, really, two verdicts, delivered in two competing academic papers that reached opposite conclusions. One study, by economists at the University of Washington (UW), suggested that the sharp increase in Seattle’s minimum wage had reduced employment opportunities and lowered the average pay of the poorest workers, just as its critics had predicted. The other study, by economists at the University of California, Berkeley, claimed that a policy designed to boost worker income had done exactly that.

The duelling academic papers launched a flotilla of opinion columns, as pundits across the US picked over the economic studies to declare that the data was on their side – or that the data on their side was the better data, untainted by ideology or prejudice. In National Review, the country’s most prominent rightwing magazine, Kevin D Williamson wrote that the UW study had proven yet again “that the laws of supply and demand apply to the labor market”. Of course, he added, “everyone already knew that”.

Over on the left, a headline in the Nation declared: “No, Seattle’s $15 Minimum Wage Is Not Hurting Workers.” Citing the Berkeley study, Michelle Chen wrote: “What happens when wages go up? Workers make more money.” The business magazine Forbes ran two opposing articles: one criticising the UW study (“Why It’s Utter BS”), and another criticising liberals for ignoring the UW study in favour of the Berkeley study (“These People are Shameless”). This kind of thing – furious announcements of vindication from both sides – was everywhere, and soon followed by yet another round of stories summarising the first round of arguments.

When historians of the future consider our 21st-century debates about the minimum wage, one of the first things they will notice is that, despite the bitterness of the disagreement, the background logic is almost identical. Some commentators think the minimum wage should obviously go up. Some think all minimum-wage laws are harmful. Others concede we may need a minimum wage, but disagree about how high it should be or whether it should be the same everywhere – or whether its goals could be better accomplished by other measures, such as tax rebates for low-income workers.

But beneath all this conflict, there is a single, widely shared assumption: that the only important measure of the success of a minimum wage is whether economic studies show that it has increased the total earnings of low-wage workers – without this increase being outweighed by a cost in jobs or hours.

It is no coincidence that this framing tracks closely with the way the minimum wage is typically discussed by academic economists. In the US’s national organs of respectable public discourse – New York Times op-eds, Vox podcasts and Atlantic explainers – the minimum-wage debate is conducted almost entirely by economists or by journalists steeped in the economics literature. At first glance, this seems perfectly natural, just as it may seem completely natural that the debate is framed exclusively in terms of employment and pay. After all, the minimum wage is obviously an economic policy: shouldn’t economists be the people best equipped to discuss its effects?

But to historians of the future, this may well appear as a telling artifact of our age. Just imagine, for a moment, combing through a pile of articles debating slavery, or child labour, in which almost every participant spoke primarily in the specialised language of market exchange and incentives, and buttressed their points by wielding competing spreadsheets, graphs and statistical formulas. This would be, I think we can all agree, a discussion that was limited to the point of irrelevance. Our contemporary minimum-wage debates are similarly blinkered. In its reflexive focus on just a few variables, it risks skipping over the fundamental question: how do we value work? And is the answer determined by us – by politics and politicians – or by the allegedly immutable laws of economics?

In the last four years, some of the most effective activists in America have been the “Fight for $15” campaigners pushing to raise the minimum wage – whose biggest victory so far had come in Seattle. Thanks to their efforts – widely viewed as a hopelessly lost cause when they began – significant minimum-wage increases have been implemented in cities and states across the US. These same activists are laying plans to secure more increases in this November’s midterm elections. The Democratic party, following the lead of Bernie Sanders, has made a $15 minimum part of its official national platform. US businesses and their lobbyists, historically hostile to all minimum-wage increases but well aware of their robust popularity, are gearing up to fight back with PR campaigns and political talking points that paint the minimum wage as harmful to low-wage workers, especially young workers in need of job experience.

In the UK, Jeremy Corbyn has pledged that a Labour government would raise the national minimum wage to £10 “within months” of taking office. (It is currently on schedule to rise slowly to £9 by 2020, which has been criticised by some on the right, citing Seattle as evidence that it will eliminate jobs.) In recent years, EU policymakers have raised the possibility of an EU-wide minimum-wage scheme. All this activity – combined with concern about rising economic inequality and stagnating wages – means the minimum wage is being studied and debated with an intensity not seen for years. But this is a debate unlikely to be resolved by economic studies, because it ultimately hinges on questions that transcend economics.

So what are we really talking about when we talk about the minimum wage?

The first minimum-wage laws of the modern industrial era were passed in New Zealand and Australia in the first decades of the 20th century, with the goal of improving the lives and working conditions of sweatshop workers. As news of these laws spread, reformers in the US sought to copy them. Like today’s minimum-wage proponents, these early reformers insisted that a minimum wage would increase the incomes of the poorest, most precarious workers. But they were also explicit about their desire to protect against capitalism’s worst tendencies. Without government regulation, they argued, there was nothing to stop companies from exploiting poor workers who needed jobs in order to eat – and had no unions to fight on their behalf.

In the field of economics, the concern that a state-administered minimum wage – also known as a wage floor – could backfire by reducing jobs or hours had been around since John Stuart Mill at least. But for many years, it was not necessarily the dominant view. Many mainstream economists supported the introduction of a minimum wage in the US, especially a group known as “institutionalists”, who felt economists should be less interested in abstract models and more focused on how businesses operated in the real world. At the time, many economists, institutionalist and otherwise, thought minimum-wage laws would likely boost worker health and efficiency, reduce turnover costs, and – by putting more cash in workers’ pockets – stimulate spending that would keep the wheels of the economy spinning.

During the Great Depression, these arguments found a prominent champion in President Franklin Roosevelt, who openly declared his desire to reshape the American economy by driving out “parasitic” firms that built worker penury into their business models. “No business which depends for existence on paying less than living wages to its workers has any right to continue in this country,” he said in 1933.

Inevitably, this vision had its dissenters, especially among business owners, for whom minimum-wage increases represented an immediate and unwelcome increase in costs, and more generally, a limit on their agency as profit-seekers. At a 1937 Congressional hearing on the proposed Fair Labor Standards Act (FSLA) – which enacted the first federal minimum wage, the 40-hour work week and the ban on child labour – a representative of one of the US’s most powerful business lobby groups, the National Association of Manufacturers, testified that a minimum wage was the first step toward totalitarianism: “Call it Bolshevism or communism, if you will. Call it socialism, Nazism, fascism or what you will. Each says to the people that they must bow to the will of the state.”

Despite these objections, the FLSA passed in 1938, setting a nationwide minimum wage of $0.25 per hour (the equivalent of $4.45 today). Many industries were exempt at first, including those central to the southern economy, and those that employed high proportions of racial minorities and women. In subsequent decades, more and more of these loopholes were closed.

But as the age of Roosevelt and his New Deal gave way to that of Reagan, the field of economics turned decisively against the minimum wage – one part of a much larger political and cultural tilt toward all things “free market”. A central factor in this shift was the increasing prominence of neoclassical price theory, a set of powerful models that illuminated how well-functioning markets respond to the forces of supply and demand, to generate prices that strike, under ideal conditions, the most efficient balance possible between the preferences of consumers and producers, buyers and sellers.

Viewed through the lens of the basic neoclassical model, to set a minimum wage is to interfere with the “natural” marriage of market forces, and therefore to legislatively eliminate jobs that free agents would otherwise have been perfectly willing to take. Low-wage workers could lose income, teenagers could lose opportunities for work experience, consumer prices could rise and the overall output of the economy could be reduced. The temptation to shackle the invisible hand might be powerful, but was to be resisted, for the good of all.

Throughout the 70s, studies of the minimum wage’s effects were few and far between – certainly just a small fraction of today’s vast literature on the subject. Hardly anyone thought it was a topic that required much study. Economists understood that there were indeed rare conditions in which employers could get away with paying workers less than the “natural” market price of their labour, due to insufficiently high competition among employers. Under these conditions (known as monopsonies), raising the minimum wage could actually increase employment, by drawing more people into the workforce. But monopsonies were widely thought to be exceptionally unusual – only found in markets for very specialised labour, such as professional athletes or college professors. Economists knew the minimum wage as one thing only: a job killer.

In 1976, the prominent economist George Stigler, a longtime critic of the minimum wage on neoclassical grounds, boasted that “one evidence of the professional integrity of the economist is the fact that it is not possible to enlist good economists to defend protectionist programs or minimum wage laws”. He was right. According to a 1979 study in the American Economic Review, the main journal of the American Economic Association, 90% of economists identified minimum-wage laws as a source of unemployment.

“The minimum wage has caused more misery and unemployment than anything since the Great Depression,” claimed Reagan during his 1980 presidential campaign. In many ways, Reagan’s governing philosophy (like Margaret Thatcher’s) was a grossly simplified, selectively applied version of neoclassical price theory, slapped with a broad brush on to any aspect of American life that Republicans wanted to set free from regulatory interference or union pressure. Since becoming law in 1938, the US federal minimum wage had been raised by Congress 15 times, generally keeping pace with inflation. Once Reagan was president, he blocked any new increases, letting the nationwide minimum be eroded by inflation. By the time he left office, the federal minimum was $3.35, and stood at its lowest value to date, relative to the median national income.

Today, invectives against Reaganomics (and support for minimum-wage increases) are a commonplace in liberal outlets such as the New York Times. But in 1987, the Times ran an editorial titled “The Right Minimum Wage: $0.00”, informing its readers – not inaccurately, at the time – that “there’s a virtual consensus among economists that the minimum wage is an idea whose time has passed”. Minimum-wage increases, the paper’s editorial board argued, “would price working poor people out of the job market”. In service of this conclusion, they cited not a single study.

But the neoclassical consensus was eventually shattered. The first crack in the facade was a series of studies published in the mid-90s by two young economists, David Card and Alan Krueger. Through the 1980s and into the 90s, many US states had responded to the stagnant federal minimum wage by passing laws that boosted their local minimum wages above what national law required. Card and Krueger conducted what they called “natural experiments” to investigate the impact of these state-level increases. In their most well-known study, they investigatedhiring and firing decisions at fast-food restaurants located along both sides of the border separating New Jersey, which had just raised its wage floor, and Pennsylvania, which had not. Their controversial conclusion was that New Jersey’s higher wage had not caused any decrease in employment.

In Myth and Measurement, the duo’s book summarising their findings, they assailed the existing body of minimum-wage research, arguing that serious flaws had been overlooked by a field eager to confirm the broad reach of neoclassical price theory, and willing to ignore the many ways in which the labour market might differ from markets in consumer goods. (For one thing, they suggested, it was likely that monopsony conditions were much more common in the low-wage labour market than had been previously assumed – allowing employers, rather than “the market”, to dictate wages). The book was dedicated to Richard Lester, an economist from the institutionalist school who argued in the 1940s that neoclassical models often failed to accurately describe how businesses behave in the real world.

Card and Krueger’s work went off like a bomb in the field of economics. The Clinton administration was happy to cite their findings in support of a push, which was eventually successful, to raise the federal minimum to $5.15. But defenders of the old consensus fought back.

In the Wall Street Journal, the Nobel prize-winning economist James M Buchanan asserted that people willing to give credence to the Myth and Measurement studies were “camp-following whores”. For economists to advance such heretical claims about the minimum wage, Buchanan argued, was the equivalent of a physicist arguing that “water runs uphill” (which, I must note, is not uncommon in man-made plumbing and irrigation systems). High-pitched public denunciations like Buchanan’s were just the tip of the disciplinary iceberg. More than a decade later, Card recalled that he subsequently avoided the subject, in part because many of his fellow economists “became very angry or disappointed. They thought that in publishing our work we were being traitors to the cause of economics as a whole.”

There were some shortcomings in Card and Krueger’s initial work, but their findings inspired droves of economists to start conducting empirical studies of minimum-wage increases. Over time, they developed new statistical techniques to make those studies more precise and robust. After several generations of such studies, there is now considerable agreement among economists that, in available historical examples, increases in the minimum wage have not substantially reduced employment. But this newer consensus is far short of the near-unanimity of the 1980s. There are prominent dissenters who insist that the field’s new tolerance for minimum wages is politically expedient wishful thinking – that the data, when properly analysed, still confirms the old predictions of neoclassical theory. And every new study from one side of the debate still generates a rapid response from the other team, in both the specialist academic literature and the wider media.

What has returned the minimum wage to the foreground of US politics is not the slowly shifting discourse of academic economists, but the efforts of the Fight for $15 and its new brand of labour activism. The traditional template for US labour organising was centred on unions – on workers pooling their power to collectively negotiate better contracts with their employers. But in the past four decades, the weakening of US labour law and the loss of jobs in industries that were once bastions of union strength have made traditional unions harder to form, less powerful and easier to break, especially in low-wage service industries.

These conditions have given birth to what is often called “alt-labour”: a wide variety of groups and campaigns (many of them funded or supported by traditional unions) that look more like activist movements. Campaigns such as the Fight for $15 often voice support for unionisation as an ideal (and their union backers would like the additional members), but in the meantime, alt-labour groups seek to address worker grievances through more public means, including the courts, elections and protest actions, including “wildcat” strikes.

In November 2012, some 200 non-unionised workers at fast-food chain restaurants in New York City walked off the job and marched through the streets to broadcast two central demands: the ability to form a union and a $15 minimum wage. (At the time, New York’s minimum wage was $7.25, the same as the national minimum.) The marches also sought to emphasise the fact that, contrary to persistent stereotype, minimum-wage jobs are not held exclusively, or even primarily, by teenagers working for pocket money or job experience; many of the participants were adults attempting to provide for families. The march, the largest of its kind in fast-food history, was coordinated with help from one of the US’s largest and most politically active unions, the Service Employees International Union. Soon the SEIU was helping fast-food workers stage similar walkouts across the country. The Fight for $15 had begun.

As the campaign gathered steam – earning widespread media coverage, helping secure minimum-wage increases in many cities and states, and putting the issue back into the national political conversation – the media turned to economists for their opinion. Their responses illustrated the extent to which the old neoclassical consensus had been upended, but also the ways in which it remained the same.

The old economic consensus insisted that the only good minimum wage was no minimum wage; the new consensus recognises that this is not the case. Increasingly, following Card and Krueger, economists recognise that monopsonistic conditions, in which there is little competition among purchasers of labour, are more common than once thought. If competition among low-wage employers is not as high as it “should” be, wages – like those of fast-food workers – can be “unnaturally” suppressed. Therefore, a minimum wage is accepted as a tweak necessary to correct this flaw. For economists, the “correct” minimum wage is the one calculated, on the basis of past studies, to give the average worker more money without significantly reducing the number of available jobs.

But this meant that almost no economists, even staunch defenders of minimum-wage increases, would endorse the central demand of the Fight for $15. A hike of that size, they pointed out, was considerably more drastic than any increase in the minimum wage they had previously analysed – and therefore, by the standards of the field, too risky to be endorsed. Arindrajit Dube, a professor at the University of Massachusetts, and perhaps contemporary economics’ most prominent defender of minimum-wage increases, cautioned that $15 might be fine for a prosperous coastal city, but it could end up incurring dangerously high costs in poorer parts of the country. Alan Krueger himself came out against setting a federal target of $15, arguing in a New York Times op-ed that such a high wage floor was “beyond the range studied in past research”, and therefore “could well be counterproductive”.

Of course, these economists may be right. But if all minimum-wage policy had been held to this standard, the US federal minimum wage would not exist to begin with – since the initial jump, from $0 to $0.25, was certainly well “beyond the range studied in past research”.

Almost exactly a year after fast-food workers first walked off the job in New York City, launching the Fight for $15, the country’s first $15 minimum wage became law in SeaTac, Washington, a city of fewer than 30,000 people, known mostly (if at all) as the home of Seattle’s major airport, Seattle-Tacoma International. It was an emblematic victory for “alt-labour”: for years, poorly paid airport ground-crew workers had been trying and failing to form a union, stymied by legal technicalities. With SEIU help, these workers launched a campaign to hold a public referendum on a $15 wage – not expecting to win, but in the hope that the negative publicity would put pressure on the airlines that flew through SeaTac. But in November 2013, the city’s residents – by a slim margin of 77 votes – passed the country’s highest minimum wage.

That same day, a socialist economist named Kshama Sawant won a seat on Seattle’s City Council. Sawant had made a $15 minimum wage a central plank of her campaign. Afraid of being outflanked from the left in one of the most proudly liberal cities in the US, most of her fellow council candidates and both major mayoral candidates endorsed the idea, too. (At the time, the city’s minimum wage was $9.47.) On 2 June 2014, the city council – hoping to avoid a public referendum on the matter – unanimously approved the increase to $15, to be phased in over three years, with future increases pegged to inflation.

The furious Seattle minimum-wage debate of last summer was ostensibly about the $15 rate. But the subject of those competing studies was actually the city’s intermediate increase, at the start of 2016, from the 2015 minimum of $11, to either $13 – for large businesses with more than 500 employees – or $12, for smaller ones. (Businesses that provided their employees with healthcare were allowed to pay less.)

When a group of researchers at the University of Washington (UW) released a paper analysing this incremental hike in June 2017, their conclusion appeared to uphold the predictions of neoclassical theory and throw cold water on the Fight for $15. Yes, low-wage Seattle workers now earned more per hour in 2016 than in 2015. But, the paper argued, having become more expensive to hire, they were being hired less often, and for fewer hours, with the overall reduction in hours outweighing the jump in hourly rates. According to their calculations, the average low-wage worker in Seattle made $1,500 less in 2016 than the year before, even though the city was experiencing an economic boom.

Some of the funding for the University of Washington researchers had come from the Seattle city council. (The group has released several other papers tracking the minimum wage’s effects, and plans to release at least 20 more in the years to come.) But after city officials read a draft of the study, they sought a second opinion from the Center on Wage and Employment Dynamics at the University of California, Berkeley – a research group long associated with support for minimum-wage increases. The Berkeley economists had been preparing their own study of Seattle’s minimum wage, which reached very different conclusions. At the city’s request, they accelerated its release, so it would come out before the more negative UW paper. And after the UW paper was released, Michael Reich, one of the Berkeley study’s lead authors, published a letter directly criticising its methods and dismissing its conclusions.

illustration for long read about the minimum wage: people trapped inside a dollar sign
 Illustration: The Project Twins

It was around this point that the op-ed salvos started flying in both directions. The conditions for widespread, contentious coverage could hardly have been more perfect: supporters of the Fight for $15 and its detractors each had one study to trumpet and one to dismiss.

Conservatives leaped to portray liberals as delusional utopians who would keep commissioning scientific findings until they got one they liked. Some proponents of the Fight for $15, meanwhile, scoured the internet for any sign that Jacob Vigdor, who led the UW study, had a previous bias against the minimum wage.

Critics of the UW study pointed out that it had only used payroll data from businesses with a single location – thus excluding larger businesses and chains such as Domino’s and Starbucks, which were most likely to cope with the short-term local shock in labour costs (and, plausibly, to absorb some of the work that may have been lost at smaller businesses). The Berkeley study, on the other hand, relied solely on data from the restaurant industry, and critics contended this did not fully represent the city’s whole low-wage economy.

But on one point, almost everyone agreed. Both studies were measuring the one thing that really mattered: whether the higher minimum wage led to fewer working hours for low-wage workers, and if so, whether the loss in hours had counteracted the increase in pay.

This approach revealed a fundamental continuity between the post-Card and Krueger consensus and the neoclassical orthodoxy it had replaced. When Roosevelt pushed for America’s first minimum wage, he was confident that capitalists would deal with the temporary price shock by doing what capitalists do best: relentlessly seeking out new ways to save costs elsewhere. He rejected the idea that a functioning economy simply must contain certain types of jobs, or that particular industries were intrinsically required to be poorly compensated or exploitative.

Economies and jobs are, to some extent, what we decide to make them. In developed economies like the US and the UK, it is common to lament the disappearance of “good jobs” in manufacturing and their replacement by “bad” low-wage work in service industries. But much of what was “good” about those manufacturing jobs was made that way over time by concessions won and regulations demanded by labour activists. Today, there is no natural reason that the exploding class of service jobs must be as “bad” as they often are.

The Fight for $15 has not notched its victories by convincing libertarian economists that they are wrong; it has won because more and more Americans work bad jobs – poorly paid jobs, unrewarding jobs, insecure jobs – and they are willing to try voting some of that badness out of existence.

This willingness is not the product of hours spent reading the post-Card and Krueger economic literature. It has much more to do with an intuitive understanding that – in an economy defined by historically high levels of worker productivity on the one hand, and skyrocketing but unevenly distributed profit on the other – some significantly better arrangement must be possible, and that new rules might help nudge us in the right direction, steering employers’ profit-seeking energies towards other methods of cutting costs besides miserably low pay. But we should not expect that there will be a study that proves ahead of time how this will work – just as Roosevelt could not prove his conjecture that the US economy did not have an existential dependence on impoverished sweatshop labour.

Last November, I spent several days in Seattle, mostly talking with labour activists and low-wage workers, including fast-food employees, restaurant waiters and seasonal employees at CenturyLink Field, the city’s American football (and soccer) stadium. In all of these conversations, people talked about the higher minimum wage with palpable pride and enthusiasm. Crystal Thompson, a 36-year-old Domino’s supervisor (she was recently promoted from phone operator), told me she still loved looking at pictures from Seattle’s Fight for $15 marches: proof that even the poorest workers could shut down traffic across a major city and make their demands heard. “I wasn’t even a voter before,” she told me. In fact, more than one person said that since the higher wage had passed, they were on the lookout for the next fight to join.

The more people I talked to, the more difficult it was to keep seeing the minimum-wage debate through the narrow lens of the economics literature – where it is analysed as a discrete policy option, a dial to be turned up or down, with the correct level to be determined by experts. Again and again, my conversations with workers naturally drifted from the minimum wage to other battles about work and pay in Seattle. Since passing the $15 minimum wage, the city had instituted new laws mandating paid sick and family leave, set legal limits on unpredictable shift scheduling, and funded the creation of an office of labour investigators to track down violators of these new rules. (One dark footnote to any conversation about the minimum wage is the fact that, without effective enforcement, many employers regularly opt not to pay it. Another dark footnote is that minimum wage law does not apply to the rapidly growing number of workers classified as “independent contractors”, many of whom toil in the gig economy.)

It was obvious in Seattle that all these victories were intertwined – that victory in one battle had provided energy and momentum for the next – and that all of these advances for labour took the form of limits, imposed by politics, on the latitude allowed to employers in the name of profit-seeking.

Toward the end of my visit, I went to see Jacob Vigdor, the economist who was the lead author of the UW study arguing that Seattle’s minimum wage was actually costing low-wage workers money. He told me he hadn’t ever expected to find himself at the centre of a national storm about wage policy. “I managed to spend 18 years of my career successfully staying away from the minimum wage,” he said. “And then for a while there it kind of took over my life.”

He wanted to defend the study from its critics on the economic left – but he also wanted to stress that his group’s findings were tentative, and insufficiently detailed to make a final ruling about the impact of the minimum wage in Seattle or anywhere else. “This is not enough information to really make a normative call about this minimum-wage policy,” he said.

The UW paper itself is equally explicit on this front, something its many public proponents have been all too willing to forget. But it wasn’t just pundits who took liberties with interpreting the results: in August 2017, the Republican governor of Illinois explicitly cited the paper when vetoing a $15 minimum-wage bill. That same month, the Republican governor of Missouri also cited the UW study, while signing a law to block cities within the state from raising their own minimum wages. Thanks in large part to efforts of business lobbyists, 27 states have passed “pre-emption” laws that stop states and counties from raising their wage floors. (Vigdor has since acknowledged, on Twitter, that it was disingenuous for the governors to cite his study to justify their “politically motivated” decisions.)

Much like my conversations with low-wage workers across the city, talking to Vigdor ultimately left me feeling that, when examined closely, the minimum-wage discourse playing out in the field of economics – and, by extension, across the media landscape – had startlingly little direct relevance to anything at all other than itself. I mentioned to Vigdor that, walking around Seattle, I’d seen a surprising number of restaurants advertising an immediate, urgent need for basic help: dishwashers, busboys, kitchen staff. This had motivated me to go digging in state employment statistics, where I learned that in 2016 and 2017, restaurants across Seattle recorded a consistent need for several thousand more employees than they could find. How did this square with the idea that the higher minimum wage had led to low-wage workers losing work?

“That’s a story about labour supply,” Vigdor said. “Our labour supply is drying up.” Amazon and other tech companies, he said, were drawing in lots of high-skilled, high-wage workers. These transplants were rapidly driving up rents, making the city unlivable for workers at the bottom of the economic food chain, a dynamic exacerbated by the city’s relatively small stock of publicly subsidised low-income housing.

These downward pressures on the labour supply, Vigdor pointed out, were essentially independent of the minimum wage. “The minimum wage [increase] is maybe just accelerating something that was bound to happen anyway,” he said.

This was not the sort of thing I had expected to hear from the author of the study that launched a hundred vitriolic assaults on the $15 minimum wage. “A million online op-ed writers’ heads just exploded,” I said.

Vigdor laughed ruefully. “Well, we’re going to be studying this for a long time.”

A few days earlier, I met with Kshama Sawant, the socialist economist who had been so instrumental in passing the $15 wage. She was eager to make sure I had read the Berkeley study, and that I had seen all the criticisms of the UW study. But her most impassioned argument wasn’t about the studies – and it was one that Roosevelt would have found very familiar.

“Look, if it were true that the economic system we have today can’t even bring our most poverty-stricken workers to a semi-decent standard of living – and $15 is not even a living wage, by the way – then why would we defend it?” She paused. “That would be straightforward evidence that we need a better system.”

 Follow the Long Read on Twitter at @gdnlongread, or sign up to the long read weekly email here.