Measuring the concert audience with the same survey methodology, same questionnaire design, same sampling allows the analysis of changes in 11 years and allow for comparisons accross Europe.
Simplified Hungarian concert demography
The Hungarian concert audience had been shrinking with little variation over the last 11 years. In Hungary, mainly the initial visits declined. The decline has several factors, but the most important one is the dramatic decline in the young population. The changing demography is only a part of the explanation. In the same time, cinamas, with a similar demographic profile, have significantly increased their audience in Hungary.
The linear simplification of the relationship between age and visiting frequency or visiting probability is typical for almost all European countries. We can desribe the size of the audience with two parameters:
In the diadict example, a Nordic country has high intial visiting frequency and the decline is almost non-exisiting over age, like in Scandinavia or the Netherlands. With 10 million inhabitants this results in 20 million tickets sold.
The Southern country in the example has similar initial visiting frequency, not unlike Croatia, but the decline is quick with age. This results in 10 million tickets for 10 million adults.
The Western country has a somewhat lower initial visiting frequency and no decline, resulting in 15 million tickets sold. The Eastern country, similar to Hungary, has a low initial visit rate and relatively quick decline, resulting in 5 million tickets sold for 10 million inhabitants.
The standardized CAP surveys allow cross-country comparisons, and allow modelling. What would happen, if music education would be at the level Denmark or Austria? How would that change the concert audience of Hungary? What would happen if Hungarians had a similar middle-aged profile like Danes or the Dutch?
For understanding the change in market forces, the ICET framework and the standardized CAP surveys (with the supplementary music professional surveys) are essential tools. It is extremely important to follow the same methodology, survey methods, sampling, etc. to allow comparison for historical data (starting from 2007) and international data.
Of course, the true relationship is not linear, although the most important factors are indeed the age and the initial visiting probability. CEEMID uses a wide range of statistical and machine learning models to find relevant forecasting tools and explanations to changes.
Hungarian relationship with loess smoothing