Stylized Facts

about former Yugoslav republics economies

Real Sector Tourism

Analysing tourism time series

Tourism demand in six states of the former Yugoslavia

This is the first in a series of blog posts analysing and discussing tourism time series in the former Yugoslav republics, now six independent states.


The aims of these blog posts are the following:

  • To describe main features of tourism time series
  • To model series using popular time series models
  • To forecast tourists demand
  • To compare forecasting performance of different time series models

The dataset under examination comes from the Eurostat database ( and respected national statistical offices. It shows monthly domestic and foreign tourists arrival and overnight stay in thousands.

The following list gives for each country: name, (acronym), the period covered and the website:

Though we have longer time series (since 2001 or even earlier) not all of the time series are homogeneous due to the changes in definitions. For instance, in case of Montenegro the graph in Figure 1. vividly illustrates the change in the time series since Montenegro became independent country in 2006. The definition of domestic/foreign tourists changed and suddenly tourists from Serbia became foreign tourists while before they were registered as domestic tourists. Therefore there is a huge drop/increase in the level of domestic/foreign tourists arrival and overnight stay since January 2007.

Montenegro tourism time series
Figure 1. Montenegro tourism time series

Unfortunately the time series for domestic and foreign tourists were not updates before 2007 to reflect change in the definition. However, the total tourists arrival and overnight stay time series still could be used in the whole period.

In case of Slovenia the change in methodology in 2018 has been explained in the following document:

We have extended both time series (tourists arrival and overnight stay) after December 2017 with the data obtained using the new methodology because the difference is not so significant, as noticed in the document: “First estimates thus show that the impact of changed coverage and methodology on the increase in arrivals and overnight stays in the first three months of 2018 is about 10 percentage points.

Dataset description

Dataset is available for download in two formats: csv and RData.

Column name in csv file uses the following format: ZzzYX, where Zzz is the acronym for the country (BiH, Cro, Mac, Mon, Ser and Slo), Y takes values D (domestic) and F (foreign), and X takes values A (arrival) and O (overnight stay). For example, SerDA is the column name for domestic tourists arrival in Serbia. Period covered in the csv file: January 2001 to July 2018.

Those familiar with R could download and use RData file. This file contains time series in different formats, but the time series used directly in the next blog posts have names in the following format: Zzzn, where Zzz is the acronym for the country (BiH, Cro, Mac, Mon, Ser and Slo) and n takes values 1 (arrival) or 2 (overnight stay). For example, Ser1 is a multivariate time series of tourists arrival with three columns, i.e. series: Total, Domestic and Foreign. In the next blog posts line graphs for each country tourism time series will be presented and discussed.


Your email address will not be published. Required fields are marked *

Director of Wellington based My Statistical Consultant Ltd company. Retired Associate Professor in Statistics. Has a PhD in Statistics and over 35 years experience as a university professor, international researcher and government consultant.