following the wave of protests in Chile, Colombia and Ecuador, the issue of the high concentration of income in Latin America once again occupied the headlines. I note, however, that there is an inconsistency between the trends detected in the behavior of inequality and the manifest discontent.

The indicator used most often in international databases for measuring inequality is the Gini coefficient (so named by those who proposed more than a century ago). How much closer is the Gini of the number zero, the lower is the inequality. With a Gini equal to 0.46, Latin America is —indeed— the most unequal region. But something that often escapes the everyday reading on the topic, inequality in Latin america fell in the last century on a scale rarely seen in history (from which data are available, of course). Around the year 2000, the Gini coefficient was equal to 0.54, or 17% higher than the latest of 0.46. A fall of this size means that, for example, in Brazil —the country's most unequal region— the income received by the richest 10% went from being close to sixty times higher than the 10% more low, less than forty times higher. Inequality declined in each country of the region, including the three where the protests have been intense. So far this century, in Chile the Gini coefficient decreased from 0.53 to 0.47 and in Ecuador fell from 0.53 to 0.44 (based on data from SEDLAC, World Bank and the University of la Plata). If the inequality experienced a fall unprecedented in the recent period how can we explain the social discontent and its virulence? I propose some explanations that are not necessarily exclusive but, on the contrary, they are dynamic that is reinforcing.

In the countries of South America, the end of the boom of raw materials has resulted in a fall in the rate of growth of per capita income and in some countries is in open recession. Therefore it should not surprise us that the popular vote in the presidential elections has favored the opposition in almost all cases, whether it were more to the left or to the right of a person who was in Government when times became hard. It is a vote of protest against the loss of purchasing power, unemployment and the erosion of benefits from the treasury. To this we should add that, in several countries, the inequality reversed its trend of the previous decade and began to climb. This has happened in, for example, Argentina, Brazil, Chile and Paraguay. The less economic dynamism, combined with increasing inequality has resulted in an increase in the incidence of poverty, just when the capacity of the state to provide mechanisms of compensation has been weakened. Combinations of this type feed on the discontent because the population experiences an intense frustration. The palpable progress during the first decade of the century was not sustained.

however, perhaps the incongruity also resides in how we measure inequality. That is to say, if we were measuring it right, since there would be no paradox: the discontent go hand in hand with growing inequality. The failures in the measurement can have two origins: the indicator or the data. It is possible that the selected indicator is not adequate to capture the frustration of the society against the inequality. The Gini coefficient (and all the other indicators of income concentration that is conventionally used measure of the differences relative to the best what causes resentment is the growth of absolute differences. If in a country all the income increase in the same ratio, the Gini coefficient of the country would be the same before and after such growth. However, in terms of purchasing power, that part of a higher income, will benefit most from this growth, even one who starts from a less income.

What has happened to the differences of income in absolute terms? Take, for example, the case of Chile, a country that —in the wake of the protests since last October and its unexpected virulence— has become a focus of special attention. While the income received by the chilean richest 10% declined to be close to thirty-three times higher than that of the 10% lowest in the year 2000 to twenty times higher in 2017, the differences in absolute terms grew markedly. During the same period, the amount of income received by the richest 10% in comparison with the poorest 10% grew up in nothing more nor nothing less than a 50% (and by 45% when compared to the difference in the income of the richest 10% and the inhabitant's means).

that Is, even when the poorest sector improved its situation, the richest group was able to expand each time more their consumption of luxury at the same time that the poor and the middle classes continued to face difficult situations caused by a social contract where the State withholding almost everything. In a thought-provoking column published on VoxEu, Sebastian Edwards underlines how Chile fares poorly on virtually all dimensions of the Index for a Better Life proposed by the Organization for Economic Cooperation and Development (OECD), and how these factors have contributed to the collective discomfort even when the inequality measured in relative terms it has fallen. I think, however, that the unrest responds to the deficiencies captured by this index —especially because they were considered unjustified in a country that is already part of the group of high-income countries— combined with a significant increase in inequality in absolute terms.

The other factor that may be leading to erroneous interpretations about the recent events has to do with the quality of the data used to measure inequality. The sources of typical information are the household surveys. A known limitation of such surveys is that households tend to declare less income than you actually receive, especially the procentes of capital income. Following this, both the extent of inequality as the trend may be miscalculated. When you correct the surveys and to remove this bias, the results can be very different. By way of example, see the work of Brazil and Chile in the World Database on Inequality of the School of Economics of Paris (developed under the leadership of the late Anthony Atkinson, Thomas Piketty and Emmanuel Saez). With the corrected data, the Gini coefficient in Brazil is not only is quite high, but that's practically not a drop was observed from the year 2000. In the case of Chile, the share of income captured by the richest 1% not only is much higher with the corrected data, but that the proportion increases instead of decreasing. From these exercises it is claror that in order to measure the inequality fully it is essential to have access to information on tax and other administrative sources that allow to calculate the income —especially that of the population in the higher strata—well. While this does not happen (and many governments are reluctant to share it), we will have a look partial and biased to the degree of inequality and its evolution over time. This leads to a misdiagnosis of the causes and consequences of inequality and public policy recommendations incomplete and wrong.

Nora Lustig is a professor of economics at Tulane University where he holds the chair Samuel Z. Stone Professor of Latin American Economics and is the founding director of the Institute of Commitment to Equity.