Information Sector Survey 2000
Part two: costs of academic books

Part one: Chances for the internet

This is the fourth annual survey by the Information Program (formerly Center of Publishing Development) of the Open Society Institute-Budapest that has inquired into the costs of producing academic books in the ex-communist countries of East and Central Europe and Central Asia.

The inquiry was for "expenses: breakdown of costs of academic books (%), net cost, without VAT. An academic book is a book used for teaching a course at university level." For the response a blank table was offered with the following categories:

• paper
• printing and binding
• authors' fees and royalties
• publisher
• wholesaler/distributor
• bookshop
• other, if any

The answers only fitted into the table with some difficulty (see below). However, the last box was only used in two cases. The Czech answer reported 1% for transport and 2% for editorial expenses. The one from Russia suggested indicated 3% for marketing costs. These will be observed at the next survey: for 1999 they were included in the expenses of the publisher.

Reliability of data

Admittedly, the replies rarely draw on local fact collections. In most cases they reflect opinions and assumptions prevailing in certain professional circles. Besides, although they are supposed to give an aggregated answer about each country, the survey relies on one reply only per country. Reliability is often undermined by basic errors by the respondents, e.g. the percentages fail to add up to 100%. However, there is no optimal structure for expenses, and with all these reservations the survey still provides an interesting insight into the circumstances among which academic books are produced and distributed in these countries after decades of command economy. The responses may point at extremities or imbalances and data from earlier years enables us to detect trends in these conditions.
By February 2001, replies from 20 countries had arrived. Since, in Survey 2000 (covering 1999), only in 11 of 20 cases did the percentages given to this question add up to exactly 100% a choice had to be made about whether to take into account the 11 formally correct answers or to correct the rest. The second option was tried, whereby the constituent figures were adapted to a reconstructed total of 100%. The test showed that both within themselves and compared to previous years the 'corrected' figures made sense and therefore their full range was included in the conclusions. The data that provides grounds for the greatest doubts is not connected to this error of addition.

The effects of correction are shown in two examples:

Original data:

Corrected data:

Structure of expenses in 1999

The question aims at the structure of the expenses and not at the actual selling price. The figures in red stand for the highest and the ones in bold for the lowest amount in a column.

First, about striking extremities
Costs of paper are below 10% of the gross amounts (i.e. net, VAT-free bookshop price) in three countries. One may think of low weight newspaper, which is indeed often used for academic books, too, in the poorer countries. The confusing thing is that for all three countries one finds neighbours reporting double this proportion. The amounts of 40% or above (Kosovo and Tajikistan) reflect the extreme difficulty in getting materials in particularly troubled zones.

As to printing and binding, it is embarrassing to see that the two extreme positions are occupied by practically neighbouring Tajikistan and Kazakhstan, where the latter seems less convincing.
Academic authors will flock to Azerbaijan if it is confirmed that a 1/3rd share of revenues goes to their account (previous surveys indicated a similar share). As many as 5 countries indicate 1/20th for the same data, i.e. authors' fees and royalties.
The interpretation of the expenses of the publisher posed difficulties: as a Yugoslavian publisher's share is nearly twenty times higher, than that of the neighbouring Hungarian publisher's (some respondents may have had the profits of the publisher in mind).
The next column is even more problematic, with the disappearance of the former monopoly state distributors (with their huge central warehouses) the functions of the wholesaler are not very clear or visible in the post communist world. That is why 5 countries evaded the answer by combining the two phases of commerce. With the remaining, it is also revealing that in two places wholesale is believed to take a bigger share from the expenses than retail. These two countries, Azerbaijan and Croatia, report the lowest (in fact unrealistically low) shares for bookshops, toppled by Kazakhstan alone where some serious misunderstanding could only result from such tiny percentages for distribution.
Most of the cited extremes are less symptomatic of the facts than of the level of awareness behind them, the detection of which is also one of the objectives of the OSI-CPD surveys.

Averages and trends

The earliest data from our surveys comes from 1996. The 12 countries reporting in that year are all present in the latest survey, too, enabling two kinds of comparison: the 12 in 1996 and 1999, as well as all 12 in 1996 and all 19 in 1999. (In calculating the means the population sizes of countries were disregarded.)

Most data indicates relative stability across the three-year time span. This is certainly so with the two blocks of production and distribution where changes are limited, both within the same 13 countries and also if the comparison involves all 20 countries reporting in 1999. In fact the minor shift has happened in an unexpected direction.
The 40% production costs were the result of the legacy of the state economy: low technological levels, high unit costs and the lack of materials due to the collapsed supply systems. Symptoms, which should indeed disappear in the face of nearly total private ownership, increased globalisation, etc. The almost 45% for production, exactly coinciding for the 12 and 19 countries in 1999, is a serious warning that not all conditions are necessarily improving. Closer examination will reveal more exact explanations.

One would also expect that - no matter how it is deplored by both professionals in the trade and intellectuals - the costs of linking the publisher and the customer, in other words the expenses of book distribution, will inevitably rise higher than the 30% indicated in 1996. Indeed, on western experiences one would not be surprised to find the opposite distribution of costs between production and distribution, namely the former exceeding the latter. Again, the data contradicts this, at least in the 3 years perspective, and show the average expenses of distribution falling below 30%.
Columns 4 and 5 show more uncertainties, which may stem less from the transparent character of the data than from the more public 'hard' data of production and distribution. Most publishers are more likely to give exact information on the latter than on their authors' or their own expenses. Authors' fees and royalties going above 10% should be welcomed in the light of the very limited share they had under the previous regime.

The last two tables compare the five sub regions. Due to the small numbers, in the case averages of two countries only, the variance is great. Yet, the logic is clear. The area closest to and most influenced by the west, i.e. East-Central Europe, shows the greatest resemblance to western standards. Curiously, however, this was more apparent in 1996 than in 1999, because the data here shows a reverse trend with production costs going up and those of distribution falling!
Respondents from Eastern Europe (the same 3 countries each time) show the highest internal consistency in their data in time. On the other hand, it needs further inquiry to find out if the great variances in data from Central Asia and the Caucasus are due to real changes or rather to survey techniques: lessons for the next survey.