Saudi corruption crackdown

If you are to search the word corruption on any search engine at this moment, you are most likely going to see headlines covering Saudi Arabia’s corruption crackdown. Well, 3 days ago a royal decree was issued and ordered to detain dozens of powerful princes, former ministers, and businessmen that are believed to have been involved with corruption. Saudi officials closed down private plane airport, and the Saudi Monetary Agency has ordered banks to freeze private bank accounts of the detainees. 

Corruption: “A form of dishonest or unethical conduct by a person entrusted with a position of authority, often to acquire a personal benefit. Activities of corruption can include bribery, and embezzlement. “

The move was long over due, too long have Saudi been plagued with corrupt public leaders that aimed to benefit themselves and their relatives at the expense of society. The costs of corruption is high, especially in shaping people’s incentives and expectations.  It reduces the trust people have in the political and economic systems they inhabit. This leads to a loss of faith in the institutions and leaders they follow. Which creates the wrong incentive for people, one that is  not aligned with prosperity and sustained economic growth.


Figure 1 is from the Journal of Economic Perspectives; the paper is called “Eight Questions about Corruption” by Jakob Svensson. The figure depicts a regression line of corruption on per capita income. There is a strong negative connection between the relationship between the well being of countries and corruption. Even though measuring corruption can be difficult, its impact can spread to other factors other than just income.

Corruption can also cause large wealth inequalities and labor market failures. The case can be particularly strong in Saudi. I just do not have the data to test this.  Essentially, Saudi Arabia’s labor market decisions is predominantly influenced by connections rather than qualifications. If you know the right person in Saudi, you are most likely going to be hired. This reality is seen by many of us, and it causes the wrong incentive. If people are getting hired based on their connections rather than their qualifications, this can be an indication that education, experience, and training does not matter as much as it should at determining wage levels and job-skill matchmaking. Which could send all kinds of wrong labor market signals.

On an international level the move is welcomed and feared. The corruption crackdown is believed to have effected oil markets. Oil prices hit a two-year high on Monday.

Figure2-Oil prices hit two-year highs on Monday following bullish news coming from OPEC and the Middle East, further stressing the tremendous impact of geopolitical risk on oil markets.

Figure 3 below depicts the Saudi Riyal  exchange against the US Dollar. The announcement of the arrests was on the news on November 4th. We can clearly see signs of fear against the currency that persisted for more than a day.

saudi riyah
Figure 3- Saudi Riyals for every 1 US dollar.

As Saudi approaches its vision in 2030, and Aramco’s IPO in 2018. It is going to need to signal to investors that corruption is not welcomed.  Investors need assurance that their property rights and investments are going to be protected from any entity. No matter how big the family name is or how deep their pockets go, corruption does not have a place in the new Saudi Arabia.






Wage determinants II: Gender & Race

In my last post, I briefly looked at differences between average wage earnings between Males and Females for 2015. I use the Census Population Survey (CPS) for March 2015 . Again, the average annual wages between males and females is shown in the table below. On average males make $50,344 and females make $42,671.


In this post we are first going to look at racial & gender differences in average annual wages.


We first begin by looking at females among races. The table above shows that Nonwhite females in this sample on average make 82 cents compared to white females.

Now, let us look at the case of white males and nonwhite males.


From the table above, we see roughly a 10% difference between white males and on white males. Nonwhite males earn $46,198 whereas white males earn $51,151.45.

All these average comparison tells a story about wage differentials. The most important thing to question is, does a person’s endowment plays a role in his/her earnings potential. Answering that question requires more empirical investigation. I can employ a multivariate regression analysis to try estimate a model of earnings. If you would like to learn the statistics behind it, click Here!

I construct two models.

Model 1: Is trying to estimate the average effect of Experience, and Experience Squared ON Wages.

Model 2= everything in Model 1 + Female, and Nonwhite ON Wages.

A brief summary of these variables would give you a better idea of what they represent.

  • Experience is basically ( Age – Education – 6), we assume that people are not working during education and everyone started school at age 6. So computing experience was suggested by the CPS.
  • Experience Squared is basically Experience x Experience. Why? Well, human capital theory suggests that experience diminishes. Meaning that after a certain point of years of experience the incremental gain from that experience is smaller than the previous. For example, after 35 years of experience you are probably going to get old and your skills and knowledge will become less significant and obsolete. The inclusion of experience squared is important. Since experience can not be forever linear. As the graph below shows. Experience has a nonlinear effect on wages.


  • Education is years of schooling. For example, 12 years = high school, 16 years= Bachelors, etc..
  • Females are just a binary variable that takes the value of 1 if female or 0 if male. For example, if an individual in the data is a female, the female variable will take the value of 1.
  • Nonwhite are also a binary variable that takes the value of 1 if nonwhite or 0 if white.

The table below is the output from the regression.


Model 1: Education and Experience (Column 1)

  • 1 more year of experience increases wages by 3.53%
  • 1 more year of education increases wages by 10.9%

Model 2: Model 1+ Female and Nonwhite (Column 2)

  • 1 more year of experience increases wages by 3.43%
  • 1 more year of education increases wages by 11.3$
  • Being a female reduces wages by 20.8%
  • Being a nonwhite reduces wages by 7.47%

These findings suggest that there other factors that affect earnings other than the conventional Education and Experience. The implication here is serious, that genetic endowment at birth might play a significant role for determining wages. In other words, your productive capacity given by your education and experience is not enough to determine your wage. Other factors such as gender and race might have a substantial role.






Wage Determinants I: Gender Wage Gap


In all forms and colors of societies, the pursuit of acquiring a job is a prominent one. Jobs are necessary in the limited time we have in this world. However, jobs are not all equal and it highly influences what kind of standard we can provide for ourselves and our loved ones. It is because of the wage that the job pays that determines your living standards. How much you are paid is a signal of the value society places on the skills/service you provide.

Little would argue against the fact that, education level and experience are the two most important factors that contribute to someones wage is , but does it stop there? Or are there other characteristics that influences wages as well?

In this post and future posts I will be looking at data from  from the census population survey (CPS) for March 2015 for the United States. Our interest here is trying to consider characteristics that can explain wage differentials. The reason this is important is because it is crucial to know whether characteristics that are endowed to us at birth might be detrimental to determining our wage.

We will be focusing this post by looking at averages, and I will emphasize why looking and comparing averages gives an unclear picture of reality. I will begin by providing some summary statistics for this sample.

Table 1.

Table1 BP

Table 1 tells us that the average hourly wage for our sample is $23/hour, the average age is 39, the average years of education is 13, and average experience is 27 years.

Table 2 (below) tell us more information about the sample. For example, there are 150,907*0.443= 66867 females and so on (Column 5). Column 5 displays how many people are in the data based on gender, race, marital status, the number of children under 18 a person has,  whether if someone is from the south or north, and whether he/she is a member of a union, and race.

Table 2.



For the purposes of this post we are going to examine Gender  wage differentials. We are interested to see whether being a female or male affects your earnings.

Table 3.


If we look at the table above,  it seems that being a female is associated with a lower average hourly wage than that of a male. The female earns on average $21 where the male earns $25. In other words, females earn 0.84 cents for every $1 a male makes. But this is a broad statement, isn’t it ? Think about it! we are not really specifying anything about this sample. Could it be that  men are just more educated than females, and that is why they receive a higher wage? We need a more accurate look to start making statements. Let us consider the sample again but this time lets limit our search to males and females that hold Bachelor, Graduate, and High school degrees.

Let us start off with the examining the average wage between males and females at the bachelor degree level.

Table 4.


In table 4, we can see that females with bachelor degrees earn $27 compared to males with bachelor degrees who earn $33 . This is even worse than the average we first computed (table 3). For Bachelor degree holders  females earn 81 cents on the dollar compared to a male.

Let us continue by looking at the average wage between males and females at the Graduate level, (i.e. +18 years of schooling).  Table 5 below surprisingly displays output on the wage differential between males and females and finds that it shrinks at the masters and doctorate level.

Table 5.


The wage differential now is much smaller at this level of education when comparing between genders. Females earn 97 cents for every $1 a male earns.

Let us now look at High school educated males and females  (12 years of schooling). It is here where we find the largest wage gap .  Females earn 75 cents for every $1 a male earns at this level of education.

Table 6.



This is important to realize, first for policy makers and for people with false perception of reality and no data to back up their assumptions. Gender wage differential does exist, but it is highly dependable on many factors. As I have demonstrated it seems that among High school and Bachelor degree holders the female/male wage gap is largest and starts to get smaller as the level of education gets higher.

It is vital to mention that the magnitude of the gap is not very accurate. Since there are other unobserved characteristics that might be causing these numbers and this data and method (looking at averages) does not take these into consideration. For example, it does not take majors the types of different degrees pursued by either males and females into account. Nor does it consider whether females might be more inclined to attain certain jobs that might be causing a large supply of females in a specific labor market. These issues could potentially be behind  lower wages, regardless of having the same level of education.

Consider the graph below as an example. Many females find it satisfying working in educational institutions because the work time suits their preferences and their ability to work and finish before 3 pm,give or take, which allows them to   be home with their families. This preference is shared by many females and sometimes causes unintended market consequences.

If a large number of females have the same preferences in jobs  that have flexible hours. Then this causes a flood of females in that labor market. In the simple graphical representation (Figure 1.), we can imagine that a flood of females in a certain job can cause the wage rate to go down.

Figure 1.

A large influx of female labor in specific jobs causes the supply to increase (shift right Labor supply to (2)), and that under perfect market conditions should decrease the wage rate, (W1 to W2).

The implication here is not to say that gender wage discrimination does not exist, and that some other factors are causing a lower wage rate for females. What I am trying to portray is from a policy stand point,  it might be more ideal to create more jobs that are more favorable to women’s preferences, in the effort to induce females to disperse to a wider scope of jobs. Rather than enforce wage floors that goes against market forces. Because enforcing prices can cause unintended consequences, in one extreme a higher unemployment.


Saudi Population 2017: Pyramids & Employment

It is indisputable that Saudi faces problems when it comes to employment and Foreign workers. According to the General Authority for Statistics, Quarter 3 for 2016 the reported unemployment rate is 5.7%. But before we take that seriously, we need to first be clear on what does unemployment actually mean? it sounds trivial, but it’s not.

Unemployment is reported only for those who have been actively looking for work, in other words it only captures those, who for the past 4 weeks, have been actively seeking work.

Therefore, the unemployment rate becomes this limited view of what true unemployment is. This is not something that only Saudis do, most countries around the world report unemployment rate this way. Discouraged workers are not reported in the unemployment rate. The General Authority for Statistics reports roughly 10.6 million individuals outside the labor force 7.8 mil for Saudis and 2.7 mil for Non-Saudis. Around a third of the population is outside the labor force, this includes elderly and children. Therefore,  a population breakdown will be helpful in explaining some of this contrast.

 A population pyramid,  is a graphical illustration that shows the distribution of various age groups in a population(typically that of a country or region of the world), which forms the shape of a pyramid when the population is growing.

Now let us look at the population pyramid for Saudi’s in Saudi Arabia (below). I repeat this is only for Saudis. We can see that pyramid looks normal and growing, equally divided between Males and Females. Most of the population is concentrated at lower three age brackets, that is (0-4), (4-5), and (20-24).

Saudi Population 2017

We can say that Saudi is a young population, which also means that there will be a continuous rise for employment needs to satisfy a younger demographic.

This shape is usually considered normal for most growing countries, because it translates that there is a large, young, and able to work population. That can support the elderly in the form of retirement and pension programs. Problems occur for countries when they have a shrinking young population and growing elder population. In such a case, not enough taxes is being collected to support programs to sustain the elderly. This is not the case for Saudi, but check out the pyramid for Japan (Below Pyramid).

Japan population

Lets focus our attention to Saudi again, but this time on the Non-Saudi population pyramid below , it tells us something about the 12.1 million foreigners in Saudi. We see a completely different shape and dynamic of the population for Non-Saudis. Males are the largest population, and the concentration of age groups exists in ages (35-39) and (40-44). The Non-Saudi population is highly concentrated in the middle, which distorts the image of pyramid. Does this image have any implications for employment?

Saudi Population Non 2017

Well if you look closer, you will begin to realize that some of the concentration in age groups by foreigners exceeds that of Saudis. I mean look at the graph above, for the (35-39) age group for Males there is about 1.4 million foreigners, compared to 0.7 million Saudis in that age group, almost double.  The table below gives a clearer picture, it has three columns. The first is age groups, second is foreign population for every 1000 Saudis (males) and the third column is foreign population for every 1000 female Saudis.

They are computed as follows:


The above computation above is to be interpreted as, for every 1000 Male Saudi’s in ages 0-4 there are approx. 259 Non-Saudis in that age group. I do this for all age groups and the table above are the numbers. Now after examining the table we see a red highlight for certain age groups. I do that because in these age groups Foreigner population exceeds Saudis. For example, for ages (40-44) there are 2041 Male foreigners for every 1000 Male Saudis in that age group, more than double.

Here is the total population pyramid for both Saudis and Non-Saudis.

Total Saudi Population

What does this mean? Well it means that Saudis face high competition now and in the future in the labor market. However, not only from themselves but also from foreigners. Ironically, it also means that Foreigners are more employed than Saudis in Saudi Arabia. As of Quarter 3 2016, there are approximately 5 million Saudis employed and 7 millions Non-Saudis employed.  In other words, for every 1000 Saudis employed there is 1,465 Non-Saudi’s employed in Saudi Arabia.

Saudi GDP II : Productivity & Capital Stock

In my last post I discussed Saudi’s simple GDP dynamics. Yet we did not discuss anything other than oil fluctuations and its relationship with growth rate of GDP and the importance of looking at growth rather than levels of GDP. Nor did we consider the important distinction between nominal and real GDP. Where real GDP is adjusted for inflation. In this post we do. We also make some adjustments to our GDP variable, looking closely at GDP per person to give us a clearer picture about the well-being of individuals. Below is the inflation adjusted gross domestic product for Saudi Arabia from 1970-2014.Rplot05.png

Factors of production

Playing around with more R and some data I provide some interesting visuals that asks important questions.  Before we do that, we need to refresh our memory about a certain element that constitutes the production process in a country. When talking about an economy’s output of goods and services, it is agreed that it depends on its quantity of inputs. These inputs are called the factors of production. An economy’s ability to turn these factors/inputs into output is represented by a production function.

The two most important inputs in production are capital and labor. Capital refers to the set of tools/machines/equipment that workers use, and Labor is the time people spend working. Another important element exists in the production process is the Total Factor Productivity (TFP). TFP is the portion of output that is not explained by the amount of input used in production. It determines how efficient inputs are utilized in production. The graphic below illustrates the process of converting inputs into outputs.


Consider three economies, that have the same level of inputs 100 Labor and 100 Capital, but they differ in productivity. The table below shows how TFP plays a role on output. Country A has a TFP of 1 and correspondingly its output is 10,000. Country B has a TFP of 1.2, suggesting it can convert its inputs more effectively and can produce 2,000 more output than country A. Country C has a TFP of 1.6 and therefore is the most efficient country that yields the highest output.


Total Factor Productivity (TFP)

Now that we have a solid introduction to these factors let us consider again the Saudi case. Using the Federal Reserve data of St. Louis that acquired the TFP data on Saudi Arabia from the “Next generation of the Penn World tables” .(TFP is indexed to USA =1). We plot the Productivity level on the left, and Productivity change on the right, of Saudi Arabia against time (1970-2014).TFP

Now, let’s plot the productivity change on top of Real GDP change. Consider the graph below, we can see below that Real GDP moves close with productivity. The rough estimated correlation is 0.6.Rplot03

Plotting the GDP per person against TFP (Below). We see that as productivity goes up we expect to see higher GDP per person. In other words, the more productive Saudis are, the more incomes they will earn.RGDP Per capita with TFP

Capital Stock 

Let us shift our focus to another aspect of the production function, Capital stock for Saudi Arabia. The capital stock is simply the amount of capital stock in Saudi Arabia across time. Consider the left graph that depicts capital stock levels and right graph capital stock change.

Capital stockCapital stock has been increasing since the 70s with the similar story of the dip in the 80’s, that followed a continuous rise. Now let us examine how changes in capital stock goes with output or Real GDP. Below we plot real GDP per person against change in capital stock. We see a somewhat linear relationship; the estimated correlation coefficient is

These visuals tells us that our production function story is relatively true, that is increases in capital and productivity are associated with higher incomes for people. We will refrain from discussing the labor input for another post.

To conclude

Let us see how it all adds up together. Below we plot real GDP per person on top of TFP (left graph).We see that there has been a close association from 1971-2000: the correlation between TFP and Real GDP per person is 0.95 (1971-2000).


This association departs in 2000, we see that after 2000 productivity did not catch up with the rise in GDP. This tells us something about the structure of Saudi’s economy, which is a natural resource dependent country. We also plot to the right the change in capital stock on top of TFP, we see that up to 90’s there was a close association between changes in capital stock and TFP: from (1971-1990) the correlation coefficient  is 0.95. The departure of the associations entails an interesting story. Despite the rapid increase in capital stock, Saudi’s productivity has remained within [1.0-1.5] range. Whether that is normal standards or not we can clearly see that dynamics of production have changed.



SAUDI GDP: Using R visualization

There is an important distinction to be made when anyone examines Gross Domestic Product (GDP). Before we go deeper let’s clarify what it means. Gross domestic product is the monetary value of all the finished goods and services produced in a country in each year. It entails all private and public consumption, investments, adding exports and subtracting imports. Simple equation illustrates,

GDP = C + I + G + (EX-IM)

GDP is an important indicator of economic health of a country, used by many as proxy for standard of living. Is it the full picture? what about the pulse of the economy?

Note: arguments that relate how GDP does not capture standard of living is for another post.

I will consider the case for my country Saudi Arabia. The graph below shows the GDP level for Saudi Arabia across time (in millions), specifically 1970-2016.SAUDIGDP1970-2016

We see a rise from 1970 level compared with 2016 level, a large dip in the 80s, and a somewhat steady continuous rise. Consider now the growth rate of GDP. The growth rate of the economy is the percentage change of the GDP from one year to the next. Which explains how fast an economy is growing.

Gdp Growth 1960-2017.png

The graph above shows the growth rate of 46 years of Saudi Arabia. This graph does does not look as consistent as we thought it is by checking the first graph, when looking only at GDP level. Growth rate tells us a different story about the pulse of the economy, one that is far more interesting than the GDP level, where the only story is during the 80’s, which we can relate to in the growth picture. In that time frame Saudi’s GDP year on year declined by 20%.

Moreover, considering the case for Saudi we can see that it is far from being consistent or stable. However, if one would look at the level of GDP Saudi starting in 1970 compared to 2016 we can safely say that on average the growth rate for 46 years was 3.7%. Yet that is far from the truth now isn’t it.

Saudi GDP.png

As we know Saudi’s GDP stems from its oil production, then there must be a considerable effect from the oil price fluctuations. Oil prices are very volatile, check the two comparisons below. The left graph depicts the price of oil since 1986, where the right graph entails the change of oil prices year on year. We see that oil is very volatile across time.


Now let’s see how oil price fluctuations looks with Saudi GDP growth. Below we can see that there exists some sort of lag effect from oil prices on GDP. By lag I mean it might be that last year oil price change effects the following year in GDP growth.Oil GDP growth 86 vs Oil price change.png

There exists an intimate relationship between the change in oil price and GDP growth for Saudi Arabia. When one looks at GDP level we do not see the whole picture of the pulse of the economy. For a natural resource driven country we see plenty of volatility from its reliance on oil as main source of income. Yet we can conclude that in the long run (46 years) Saudi has grown on average 3.7% per annum.

Update* June 20th 2017

I acquired data that explains this relationship better. As the graph below shows, the Real GDP Growth of the Saudi Economy and Oil Sector growth. They exhibit a 0.77 correlation which indicates the intimacy mentioned previously.

Saudi RGDP growth and oil sectory growth.png

The white tax on undeveloped land in Saudi Arabia has been implemented.

According to a recent post by ARGAM ““, the white tax on undeveloped land has been implemented.

I am in full support of this policy, since the government is trying to incentivize owners of land to develop their lands in order to reduce the shortage in the housing market that is propping up house prices. Which also safeguards fair competition and it combats
monopolistic practices in the housing market (since paying a 2.5% tax on property that does not generate cash flow seems to be a like a drain on a landowner’s account balance).

There has been substantial amount of literature and research on the issue of land inequality, such that a small percent of people own a large percent amount of land. This inequality should not be taken lightly. Many publications in top economic journals have discussed the implications on society and the economy when high levels of land inequality exists. Some findings even suggest that land inequality by itself promotes stagnation in economies that seek to develop into industrialized ones.

The ideal situation would be that house prices decline, then the incentive of owning a piece of land falls and that in turn, causes people with wealth to dwell into other businesses that can generate fair returns and economic activity.

At a time where Saudi is facing rapid structural changes, those who have exploited the less fortunate finally must pay up.