Adjustments should be made for lower costs of living in Southern countries and higher in Northern countries. Particular attention should be given to employees moving to or from Switzerland and the Czech Republic as they are outliers. Central European countries face similar costs of living and salaries. Please note the salary ranges are based of the average labor pool with relatively low job switching friction. Skilled and specialized labor may face country specific industry differences.

This model is to support judgement, not replace it. An employee transferring from the Czech Republic to Switzerland should be compensated greater than if they were to transfer to Portugal. Should one support the opposite position for an individual then specific and significant arguments would be required.

Problem
Determine salary ranges for European branches of company using provided datasets.

Data
Provided data for the exercise was taken from the Organization for Economic Co-operation and Development (OECD). This exercise was restricted to given data. Industry specific and non-average labor pool data would provide better estimates and finer grained results. The provided data consisted of the consumer price index and average salary levels computed from 2014 through 2018 observations for 12 countries. Oddly included with this data for the European market was Australia and the United States.

I excluded Australian and United States data from the already thin dataset. How CPI is measured varies across countries. Harmonized Index of Consumer Prices has a more precise measure, but is exclusive to G7 countries and not included with the given data. I am making the assumption that the error in measure is similar between neighboring countries, but dissimilar between continents. An additional assumption is that the switching costs for an employee moving between European nations is significantly lower than between AUS or the US and Europe. Removing AUS and the US changes the fit to linear.

Model
A linear OLS model is used. One could use the average salary as stated and call it a day. However by using OLS we are able to control for cost of living by country. This helps account for the opportunity cost of hiring someone in a neighboring country by standardizing to a real wage. I am ignoring any differences in relative output of employees by country. Simple regression models in which average salary is controlled by cost of living are tested to find the best fit, with a linear model providing the greatest precision. This provides a simple model to predict average salary for each country.

However, the data is sparse and does not account for variance around the average. A prediction interval range is created around the point estimate to provide managers with leeway in offering compensation. The prediction interval is a basic estimate of what the future average wage will be by country. I make use of a 90% prediction range. This provides a range around each countries adjusted average salary. The range varies between $5.8k and $5.1k plus or minus from a each countries adjusted average wage.

The salary ranges provided by this prediction range may be used as a reference for manager to support arguments, not replace them.