Özet:
This study assesses the International Digital Economy and Society Index (I DESI), which measures the progress of the digital economy, clusters countries, and
forecasts their scores. First of all, because the initial criteria weights are chosen
subjectively according to EU policy orientation, they were compared to new ones,
generated using entropy method characterized by its objectivity. Then, the ranking
model of the I-DESI is assessed by using the Entropy-based, the TOPSIS-based, and
the Entropy-based TOPSIS models. Ranks agreement among the four models is
tested using Kendall W. Afterword, in order to check the similarities of the EU
countries, hierarchical and K-mean clustering methods are performed. At the end,
forecasting of the I-DESI, by using Gompertz II model, is performed in order to
compare Turkey’s forecasts by dimension to the forecasts of the I-DESI-45 countries
average. The findings demonstrate that the Entropy method gives the highest weights
to "Connectivity" and "Integration of Digital Technology and Business" dimensions,
whereas the I-DESI scoring model gives it to "Connectivity" and "Human Capital".
The Entropy method, on average, gives nearly double the weight to the "Use of
Internet Services by Citizens" than to the "Digital Public Services", whereas
the I-DESI scoring model gives both the same weight. Also, a very strong correlation
exists between the I-DESI model and the TOPSIS-based and the Entropy-based
models. While a moderate positive one is with the Entropy-based TOPSIS model.
The four ranking models agree to a considerable extent. The differences found
between the scoring model and the entropy means that countries (mostly non-EU)
who are doing well in “Use of Internet Services” dimension, like Turkey, get
penalized in the I-DESI computation as the subjectively fixed weight is almost half
what the entropy generated. The correlation found between the four ranking models
means the I-DESI model, even including the use of subjective criteria weights and
viii
aggregating using the score model is a similar model of scoring and ranking when
compared to objective multi-criteria decision methods (Entropy and TOPSIS). The
Kendall agreement between the four models means that they generate almost the
same ranks. Recommendations could be used to adapt strategies for digital
competitiveness. Such strategies should include: Scoring higher and faster in the I DESI in the future, requires giving more importance to two dimensions: Connectivity
and "Human Capital", because they weight together 50%. Added to that, maintaining
the rising performance of Use of Internet Services dimension, this strategy
orientation will prove effective and straightforward: (i) Expanding the fixed
broadband and increasing internet speed, (ii) Boosting trainings for people to acquire
basic and advanced IT skills (word, spreadsheets, coding) and to increase the number
of ICT graduates from educational institutions, (iii) Increasing the use of technology
related to banking and shopping transactions and the number of internet users in rural
areas in particular, (iv) Enhancing e-government services to comparable leading
countries and making it possible to complete each step of key services completely
online.