Revisiting the Phillips Curve in Saudi Arabia Context
Background on the Phillips Curve
In his paper on the relationship between unemployment and inflation, later known as the Phillips Curve (1958), Professor Phillips looks at the correlation between the two macroeconomic indicators between 1861 and 1957 in the United Kingdom. Specifically, the author’s objective is to empirically test the hypothesis that variation in wage rates can be explained by changes in the unemployment rate, which is an indicator for the demand of labor.
Phillips begins his paper with a discussion of three main factors that provide evidence for a non-linear relationship between the inflation rate and the unemployment rate. First, during economic booms where demand for labor is high and with a few job seekers; firms are likely to raise wages above existing wage rates to attract talented workers. Second, during economic downturns when labor demand is low and a high unemployment rate; workers are often hesitant to accept wages below current market rates. Third, changes in retail prices can affect prevailing wage rates, through the cost of living adjustments that firms make to maintain their workforce. More specifically, the author argues that with the exception of periods with high increases in import prices, wage rates can be largely explained by the unemployment rate.
Furthermore, the author introduces two new concepts that explain changes in the wage rate. First, Phillips discusses demand-pulls, whereby wage rate changes occur due to aggregate demand increases outpacing supply or firms’ production capacity. This is because as total demand rises, firms attempt to satisfy this demand by hiring more labor; which in turn means more people have employment and more disposable income that puts further pressure on prices, thus ultimately leading to higher inflation. Second, the author explains how cost-push factors relate to changes in the wage rate, whereby increases in production costs cause a lower supply of goods, while demand remains the same. This causes firms, according to Phillips, to pass on the burden of higher production costs onto consumers, thus leading to higher inflation.
Critiques of the Phillips Curve
Since the publication of the Phillips curve in 1958, it has drawn attention and criticism from several prominent economists starting with Samuelson and Solow (1960), and followed by Friedman, who criticized the Phillips’s model and argued that it only applies in the short-term. The critique of the Phillips curve amongst economists further intensified during the mid-1970’s when both unemployment and inflation rates remained high for several years. Therefore, this economic phenomenon, later referred to as stagflation, violated the assumptions stated in the Phillips curve where there is a constant tradeoff between the unemployment rate and the price level.
However, more recently, empirical work by Professors Akerlof, Dickens, and Perry (1996) suggests that there still remains a moderate tradeoff between the two economic indicators. More specifically, the authors found that workers tend to tolerate real wage cuts more than they would accept nominal wage reductions. For instance, a worker is more likely to accept and be satisfied with a wage increase of 3% when inflation is 4%, than accepting a wage cut of 3% when inflation is 2%. This behavior occurs although the impact on real wages (i.e. adjusted for inflation) is the same.
The Phillips Curve in Saudi Arabia
Below I look at the relationship between the two macroeconomic indicators in my country Saudi Arabia, between 2003 and 2019. I used youth unemployment rate given some data limitations and given that most of the unemployed population (roughly 91%) in Saudi Arabia is composed of younger individuals who never held a job. There are periods where there appears to be a moderate inverse relationship between unemployment and inflation rates, but it is often inconsistent across the years.
Between 2005 and 2007, the two indicators moved together steeply downwards until the Great Recession in 2008, when consumer prices began to rise sharply. I think one plausible explanation for the steep increase in consumer prices between 2007 and 2008 is that it has been driven primarily by oil prices which peaked at $165 in May 2008, and because the country relies heavily on imports for consumption, it eventually drove the price of consumer goods upwards.
Between mid-2008 and 2010, oil prices decreased by over a 100% to $50 and this time drove the inflation rate downwards, as the cost of a primary production input, oil, has declined.
Interestingly, consumer prices began to increase again due to another oil boom early in 2010 and the unemployment rate moved in the opposite direction, thus providing evidence for the Phillips curve. I believe; however, that the steady decline in unemployment was driven by the government’s stimulus package at the time. With rising oil prices and instability in the region, the government responded with a large welfare package that hired roughly 540,000 individuals (mostly below 35 years old) between 2011 and 2013. To put this figure in perspective, this is equivalent to employing over 5% of the country’s population within a two year period. Despite the introduction of several wage subsidy and on-job-training programs; the unemployment rate began to increase again starting in 2013 due to a combination of demographic factors and the expiration of subsidized working arrangements. Furthermore, young females and males began to enter the workforce at unprecedented levels at 200,000–250,000 individuals (70% of which were college graduates) per year. At the time, the private sector had 9 million workers, but only 5 million of which were suitable for those with a HS degree or more. Therefore, this essentially meant that private firms had to create approximately 5% (250K/5 million) more of the existing jobs to handle the new entrants into the labor force. However, the private sector was heavily dependent on both government projects and wage subsidy programs and I estimate that it was on average, only adding 30,000–50,000 nonsubsidized jobs.
One may then question why youth unemployment rate declined sharply from 2015 onwards. I believe one explanation is that as oil prices crashed in December 2014 by approximately 90%, the government responded in three ways that ultimately led to lower unemployment.
First, it reduced its spending on welfare programs and did not renew spending at the same scale. For instance, the Teachers Wage Subsidy program (it employed an average of 105,000 teachers in the private sector for a five-year period) between 2012 and 2017 was let to expire. This caused many workers to decide between lower wages or unemployment, and most choose the former. Second, the government significantly cut wages for public sector workers, who then represented 80% of employed Saudis. As a result, the public-private sector wage differentials declined and the public sector, albeit still paying higher hourly rates, became less enticing to Saudi youth. Third, the government reduced public sector hiring and many ministries imposed hiring-freezes that are still in place today.
The three events in turn led to many young Saudis to adjust their expectations of wage rates in the private sector and to reduce their reservation wage to levels that they would not have accepted in the pre-2014 oil crash. While I will need more data to further test the hypothesis above. However, one piece of evidence to support it comes from aggregate data on the private and public sector employment ratios. Specifically, in Q2, 2016, the ratio of native workers in the public sector was 55%, while the remainder were in the private sector. Within four years by Q1, 2020, the ratio of employed nationals in the private sector increased significantly to .61 or 61%, while the remaining 39% are in the public sector.
