In this project we have selected all those individuals who had some sort of income from employment over the years 2007 to 2018, using data from the Muestra Continua de Vidas Laborales or MCVL (Continuous Work History Sample), which is based on their relationship with Social Security each year (employment, unemployment, pensions). For the studied population, the annual employment income for the period 2007 to 2018 (Graph 1) is calculated by adjusting for the effect of price rises to provide a measure of income with stable purchasing power (real wages).
This analysis is performed by calculating the median and the quartiles. The median is the value that leaves half of the data on each side. Similarly, the quartiles divide our population into four equal parts. The first quartile marks the boundary of the quarter of the workforce with least income. And the third quartile marks the point beyond which we find the 25% of workers with most income.
Given that the median value in 2007 was 18,919 euros, half of the workforce had an annual income below this sum, and the other half earned more (Graph 1). We can see that the median reached a maximum of 20,216 euros gross in 2009, the minimum being recorded in 2017 (18,478 euros). The first quartile stood at around 13,000 euros, while the third quartile was at about twice this figure.

If the trend in annual income is represented as an index that takes value 100 in 2007 (Graph 2), we find that in 2008 and 2009 income grew, and that higher incomes (above the third quartile) did so to a greater extent. Thus, in 2009, whereas the third quartile was 7.6 percentage points higher than the 2007 level, the first quartile rose by only 3.8 points. When wages started to fall, the effect was greater for lower wages. The drop that occurred as of 2010 caused workers in the third quartile to revert to 2007 wage levels by 2012, while in the same period those in the first quartile had lost 2.7 percentage points. After 2012, the third quartile group had lower income only in the period 2016-2017, standing 2.3 points below their 2007 income. However, the income of the lowest-paid group fell progressively every year, with the result that by 2017 their annual employment income was seven points lower than in 2007. In the last year studied, 2018, a slight wage recovery was observed in the lowest-wage group, although they still fell short of the 2007 income level.

If we consider the trend of the median, we can see that in the first two years of the crisis the median real wage rose by 2.6% in 2008 and 4.1% in 2009. At the time there was much debate on the causes of this increase, because at the same time as wages were rising, the number of jobs was plummeting. Suffice to say that in 2008 more than 1.3 million jobs were destroyed.
It was unclear to what extent this wage increase corresponded to the destruction of low-quality jobs with low wages and the maintaining of jobs with medium and high wages that therefore contributed to an increase in the median wage. It was suggested as an alternative that the rigidity of collective bargaining prevented wage adjustments, and that companies consequently opted to reduce their employment levels (Cuadrado et al., 2011).
Despite this lack of initial response to the crisis, wages decreased as of 2010, especially in 2012, when they fell by 3.4%. It is important to bear in mind that the labour reforms passed in 2009, 2010 and particularly 2012 changed substantial aspects of collective bargaining, which contributed to this cut in wages (Pérez Infante, 2015; Malo, 2015).
In the wake of the strong impact of the crisis, in 2014 the Spanish economy began to recover jobs. However, wages did not pick up. From 2013 to 2015, variations were minor, median annual income remaining around 19,000 euros (in real terms). But in 2016 and 2017 it dropped again by more than 1% per annum. Overall, the accumulated decrease caused the median wage in 2017 to be, in real terms, 8.6% lower than that corresponding to 2009. In 2018 there was a slight increase of 0.3%.
On analysing the interannual variation by quartiles we might get the impression that the drop in wages affected all workers in much the same way (Graph 3). However, we must take into account that although the trend is similar, when growth occurs, it is greater in the third quartile (the workers who earn most) than in the first (the workers who earn least). Thus, in 2008 and 2009, wages in the first quartile rose by 1% and 2.7% respectively. The increases for third-quartile wages were larger (2.6% in 2008 and 4.9% in 2009). The same happened in 2013 and 2015, i.e., wage increases were larger for workers with higher wages than they were for low-income workers. In 2018, however, the small rise that occurred had more impact on lower-paid jobs (first quartile), with an increase of 0.9%.

Sex-disaggregated data show that the trend is similar in men and women, but women’s wages are substantially lower than men’s. Thus, whether we consider the median or the quartiles, women’s wages stand at around 80-85% of men’s, which reflects the well-known gender pay gap (Graph 4).
