Total confusion: How can Hungary's ailing tourism keep breaking records?

Let's kick off with three quotes:

1. "In fact, this year's peak season has seen 17 million overnight stays, which is ahead of last year and even 2019."
2. "More than six million guests visited Hungarian accommodation establishments in the three summer months, according to the National Tourism Data Supply Centre (NTDSC). This year's summer traffic is 1 percent higher than last year's peak season, and 6.3 percent higher than in 2019."
3. "The World Athletics Championships in Budapest in August were a huge success, boosting the Hungarian capital's visitor numbers significantly. The number of overnight stays by foreign visitors from Australia and Oceania increased by more than 120 percent, while those from Africa and Asia increased by almost 60 and 50 percent respectively compared to August 2022."

The first two remarks were made by Zoltan Guller, President of the Hungarian Tourism Agency, who assessed the country's summer tourism performance. The MTÜ is the government-appointed operator and supervisor of the National Tourism Data Supply Centre (NTDSC or NTAK in Hungarian), which has been collecting data on the traffic of tourist accommodation since 2019.

The third quote comes from Áron Kincses, President of the Central Statistical Office (KSH), who spoke at the same event about the impact of the World Athletics Championships in August. The statements are on shaky ground on several points, and the absurdity of the presentation of tourism results is illustrated below using these examples.

Before explaining why the above statements are extremely strange, and how the data can be used to paint a completely arbitrary picture, we present the following figure for illustrative purposes. It shows the evolution of guest nights in commercial accommodation establishments:

The graph shows the following main trends in tourism performance:

• Domestic tourism recovered last year from the downturn suffered in the wake of the coronavirus crisis, with traffic returning to 2019 levels by 2022, but has contracted slightly again this year. The primary reason for this may be the fall in real wages, but a pick-up in travel abroad may also have contributed to the decline in domestic tourism.
• The travel activity of international tourists has still not reached pre-crisis levels, as the rapid and sharp fall due to Covid has not yet been followed by a full correction. However, continued this year, albeit with still less momentum.

If you feel that this evaluation does not coincide with the message of the quotes at the beginning of the article, it's not your fault.

(As we will not discuss this later, we note for now that, in our view, this assessment correctly describes the situation and main trends in domestic tourism, despite various methodological and statistical limitations. 2023 brought a moderate stall in the correction after the coronavirus outbreak, but overall there is nothing dramatic in tourism traffic.)

This is how messing around with numbers becomes child's play

To understand why there is so much confusion about communication around tourism, it is worth looking at the data. It is possible to get lost in the confusion of tourism figures. Performance is measured by different indicators, in different breakdowns:

• Guests, nights. The two main indicators are guests and guest nights. These are broadly in harmony, but there are minor differences. (For example, in August this year, 3.8% fewer guests spent 0.8% more nights in domestic accommodation than in the same month last year.)
• Foreign, domestic. We distinguish between foreign and domestic tourists, with Budapest being the main destination for the former and the countryside for the latter.
• Regions. Accordingly, there is also a breakdown by main tourist destinations (Budapest, Lake Balaton, Mátra-Bükk, etc.).
• Data sources, KSH vs. NTDSC. The main source of confusion and "creative interpretation" is the two different data sources. The KSH has statistics based on its own data collection going back years. In 2019, the National Tourism Data Supply Centre (NTDSC) system was set up and started collecting data in parallel with the KSH data collection. But the NTDSC does not provide data to the public. Therefore, the NTDSC data are only be available from 2022 onwards, after which the KSH replaced its own survey with the NTDSC figures and now publishes them month by month.
• Commercial versus private. Prior to 2019, the KSH collected data on only one of the two broad types of "tourist accommodation", namely "commercial accommodation", which includes hotels, boarding houses, camping sites, community hostels, and bungalow complexes. From 2019, the NTDSC also started to monitor the turnover of "private and other accommodation" (which accounts for about one third of total tourist accommodation turnover), giving a broader picture of the performance of the tourism sector. However, it is also true that the extended data are only be available from 2022 onwards.

If this seems confusing at first, you may find the table below helpful:

In other words, we have virtually no meaningful time series for the entire period under review. Firstly, about a third of the accommodation statistics are not available for before 2021. Secondly, the data that are available (commercial accommodation) are also compiled using a new methodology, so the data for different periods can only be compared with great uncertainty.

A good example for this is that the number of nights spent in commercial accommodation is significantly higher in the NTDSC system than in the KSH data collection. We know this from the fact that from the beginning of 2021 to May 2022 we have both sets of data, where the difference is significant, 17%.

This is also the reason why our graph presented above is of limited use, as it only includes the number of nights spent at commercial accommodation establishments, and that too with a methodological break. However, the public cannot know any more than this, because (as indicated above) the NTDSC data for 2019 to 2021 are not publicly available.

Portfolio has attempted to obtain the missing data from the KSH and the Hungarian Tourism Agency (HTA or MTÜ in Hungarian). Without giving any explanation, KSH simply replied rather bluntly that "the data are available from 2022". It did not elaborate why, if such data existed from 2019, it considered it appropriate to take them from the NTDSC only from 2022.

The response from the HTA was not overly helpful either. "According to the current legislation, official statistics on the tourism sector are available on the website of the KSH. The HTA does not perform statistical data communication tasks." - they wrote.

It is true that the HTA does not publish its data in a structured form, but we can see from the quotes at the beginning of this article that it nevertheless prefers to refer to its own unpublished 2019 data as the peak period for tourism, which it says the sector has gloriously outperformed this year.

Communication confusion is basically the result of the above:

1. The professional organisations carrying out the evaluations refer to data that are not known to the public and their comparability with recent figures is highly questionable.
2. There is often a sense of selectivity in communication. Among the vast array of numbers (by country, by destination, by sending country, by type of accommodation, etc.), there is always one that shows a favourable picture, and these are often highlighted, with little attempt to present a complete, objective picture.

Wild Interesting claims - Should we split hairs?

In light of the above, let's see what problems arise with the quotes at the beginning of the article.

"In fact, this year's peak season has seen 17 million overnight stays, which is ahead of last year and even 2019."

The cited data can cause head-scratching and raised eyebrows for various reasons.

1. In June-August 2022, 17.145 million tourism nights were recorded, according to the KSH, and 17.102 million in the same period of 2023. This means that this year's season was not stronger than last year's, but a tad weaker. Moreover, the difference in this case cannot be caused by a different database, as the KSH used the NTDSC data for 2022-2023.
2. Even more controversial is the claim that the summer of 2023 was stronger than 2019 in terms of overnight stays. As shown above, there is no public and consistent time series for this period. What data is available makes it very difficult to understand what would have made this year's peak season stronger than the pre-Covid peak. In fact, according to the figures of the KSH, the number of nights spent in commercial accommodation in the summer months in 2023 was 11% lower than in 2019 (10.9 vs 12.3 million). As the 2019 KSH data is still based on an old methodology, the explanation may be that the non-published NTDSC data used by the HTA shows fewer guest nights for 2019, so there could be an increase this year compared to that. However, as we have seen above, it is the NTDSC figures that are higher in 2021-22. So it's a complete mystery what the HTA is basing this summer's tourism peak on. The only way to resolve this enigma would be to make the data on which the HTA's assessment is based public, so that it can show why the assessment differs sharply from what can be inferred from publicly available data.

The confusion around the first quote is also evident in the second.

"More than six million guests visited Hungarian accommodation establishments in the three summer months, according to the National Tourism Data Supply Centre (NTDSC). This year's summer traffic is 1 percent higher than last year's peak season, and 6.3 percent higher than in 2019."

A 1% annual increase is in line with the official figures (a 7% fall in domestic traffic was offset by a 15% rise in foreign traffic). However, the comparison with 2019 brings up the problems we discussed above: we do not have a public, complete time series up to 2019, and what we do have shows a very different picture and gives a very different overall picture.

In fact, compared to 2019, the number of guests staying in commercial accommodation this summer was 9% lower this year. How this translated into a 6.3% increase is a mystery. Especially since the observation that the old methodology of the KSH reports a lower number of guests than the NTDSC (at least between January 2021 and May 2022 there was a 9% difference between the two data sources), which makes the discrepancy between the 6.3% increase and the 9% decrease seem even greater.

You can guess, of course. The NTDSC's unpublished figures, or more precisely the picture painted by the officials who refer to them, may be much more attractive than the official figures of the KSH for two technical reasons.

One is that private accommodation, for which no data are available before 2022, performed so poorly in 2019 that the growth in private accommodation grossly overcompensates for the contraction in commercial accommodation. However, this is unlikely.

Firstly, in the recent years there has been no spike in private accommodation that would be reminiscent of the "Zimmer Frei" times, a surge in the 1990s after the change of regime. Secondly, this segment would have had to grow enormously to have such an impact on the performance of the sector as a whole, given its small weight. Thirdly, it is totally unrealistic that the two major accommodation types should face such a different cycle in tourism.

The other possibility is that the NTDSC system has much smaller numbers of commercial accommodation units than the tables of the KSH. Theoretically, however, the NTDSC system is more complete and should therefore include larger numbers - as the lessons from the sub-data quoted above confirm.

It's only speculation, but we wonder that the 2019 figures are lower because the NTDSC system simply could not provide sufficient data in the first year of its roll-out. (This possibility would be strengthened if KSH did not want to replace its own survey with NTDSC data retroactively to 2019 because it sees that there were problems with the data supply at the beginning of the period - but this is of course also only speculation.) If this is the case, however, it would be a huge professional mistake to refer to these 2019 figures as a benchmark and talk about record breakings this year.

Our third remark we took as an example is as follows:

"The World Athletics Championships in Budapest in August were a huge success, boosting the Hungarian capital's visitor numbers significantly. The number of overnight stays by foreign visitors from Australia and Oceania increased by more than 120 percent, while those from Africa and Asia increased by almost 60 and 50 percent respectively compared to August 2022," said Áron Kincses, President of the Central Statistical Office (KSH).

This quote, unlike the two above, is not a bad example of incomplete data being treated as confidential. It is a case of how, out of a vast array of numbers, there are always some that present a favourable picture, and these are highlighted with a predilection, with little attempt to show a complete, objective picture.

What is really about the capital's tourism that was buoyed by the Athletics World Championships in August? Well, it's that in August this year, 14.1% more nights were spent in tourist accommodation in Budapest than at the same time last year.

It's not 120%, but it's still an increase, you might say. However, the year-on-year increase was 13.2% in June and 12.3% in July, so August, even with the effect of the World Athletics Championships, does not stand out from the summer months in terms of tourism performance.

The last month of the summer was exactly what the main trends described at the beginning of this article suggest: this year foreigners are spending more nights in Hungary, while locals spend fewer, but as tourism in the capital is foreign-dominated, there is an increase here, but not in the countryside. Of the 12 tourism regions, the capital, the capital's surroundings and Győr-Pannonhalma saw an increase from last year, while the other nine saw a decline.

In other words, the final tourism figure does not at all support the KSH President's claim that the sporting event significantly boosted the Hungarian capital's visitor numbers.

So what do we make of the huge growth numbers in the quote? We add them to the list of extreme, atypical, extraordinary examples that are so misleading. It is not surprising that in the case of an athletics world championship, more people come from Oceania, where very few come from, and even a broadly defined athletics contingent would make a significant difference to the numbers. But the weight of the countries and regions listed above in our total foreign tourism is only 8-12%.

Unfortunately, we do not see data for the capital, but for the country as a whole, broken down by sending country, but these are still suitable for drawing cautious conclusions. In the peak season (the three summer months), foreigners spent 12.4% more nights in Hungary than in the base period.

Within this, the annual increase in August is smaller, at 11.4%. The picture may be slightly more positive in the capital, but given Budapest's predominant pattern of inbound tourism by foreigners, there is certainly no evidence of explosive growth in August. Among the major sending countries, the number of nights spent by visitors from France, the Netherlands and Slovakia in August was roughly the same as a year earlier.

the athletics world championships did not have such an impact on tourism in the capital that it had a spectacular effect on August's overall performance.

We don't have to look very close at the HTA to find similar examples of communication that over-emphasise the good results and keep the negative ones under wraps. This organisation is really fond of commenting on the monthly tourism data with a gung-ho attitude.

There are no lessons, except that statistics should be treated with caution. If they are published and interpreted in a way that does not meet the required professional standards, they may sooner or later lead to a serious loss of credibility. And statistics that become discredited would be a serious loss to society, they would disappear as a common basis for discussion and professional debate, they would seriously impair our understanding of the processes taking place in the country, and they could lead to bad economic and economic policy decisions.

Cover photo: Harun Ozalp/Anadolu Agency via Getty Images