Economics is a compelling subject. It helps us understand the world, people’s decisions, the development of nations, and essentially everything else that matters in our world. Econometrics is how we empirically analyse our economic findings; it helps us analyse trends and aids in drawing conclusions to pressing problems. Econometrics, however, can be really, really challenging, and that is talking from shared experience across many students in many universities. So, if you feel a little lost in your complex econometrics classes, don’t fear, in this series we are going to try and make intermediate econometrics a little clearer.
Data
We typically deal with two kinds of data when participating in econometrics: time series and cross-sectional data. Time series data is data which is collected at regular intervals over time, and it allows us to track trends over this course of time. Cross-sectional, on the other hand, is a snapshot of a collection of data points, which allows us to compare the data at the same instance of time.
Variables
in any econometric analysis, or any experiment for that matter, there are two main variables; the independent(sometimes called an exogenous) and dependent variable (sometimes called an endogenous variable).
The independent variable is the variable that the experimenter changes or controls so that they can observe the effects on the dependent variable, which is the one that is measured (because it’s result is ‘dependent’ on the independent variable).
Regressions
Let’s propose a model; it seems there are random scatterings around the graph. How do we possibly determine the trend in the data? The answer is a simple straight line. This is essentially what a regression is. The line does not do a perfect job at explaining all variations in the data, but it does a fair job at summarising the relationship between variables.
But how is this straight line determined? Well, it essentially sets a straight line which has the smallest sum of squared residuals. The program you use will draw a line through the data and calculate the squared value of the distance of the line to the data point; this is called the squared residual. The squared residual is essentially the error that is present in the model, the difference between the line and the data point is squared to take the absolute value of the difference and to emphasise the error values of big differences. Then the sum of all these squared residuals is taken to get the total squared residual value of this line. The program will then repeat this process until it finds the straight line which will have the smallest sum of squared residuals. This process is called the order of least squares (OLS)
Correlation
When carrying out econometric analysis, we might want to know if two variables we are analysing are associated – another way of saying correlated. This metric demonstrates the degree to which these given variables move in value together.
The correlation coefficients are typically denoted in econometric analysis by r. The coefficient r is measured between -1 and 1, such that:
- r = 0: no correlation
- r < 0: negative correlation – if one variable increases, the other decreases.
- r > 0: positive correlation – if one variable increases, the other also increases.
A warning, however! Correlation does not equal causation! This is a very important point to remember, as often people believe that just because things are correlated, it does not mean that one thing specifically causes the other. For example, the number of fresh lemons imported to the US from Mexico is highly correlated with the number of highway deaths. Does one of these cause the other? I think not…
Economic Convergence
In economics, convergence is the idea that poorer countries grow faster than richer ones in per capita income. This is because there’s more room to grow. When you’re starting from a low base, basic improvements such as infrastructure, education, or technology adoption can generate big gains.
There are two types of convergence on the track. Absolute convergence is the optimistic scenario: all countries, regardless of their baggage, eventually reach the same income level. But in reality, we usually see conditional convergence, where countries only catch up if they share similar fundamentals like savings rates, education levels, population growth, and institutional strength.
Because understanding convergence tells us a lot about where future growth is likely to happen. It shapes investment flows to development policy. in emerging markets, it’s not just about growth, it’s about how fast and under what conditions they can narrow the gap.
