Why and how could we measure the Matthew effect for countries

A quarter of a century after the appearance of the classical work of Merton (1968) on the Matthew effect in science, we have unambigously measured it for countries. With this measurement technique, countries will be able to determine their position in the stratified rank order of scientific nations. The fact that every country belongs to one of two categories receiving fewer citations than expected or receiving more citations than expected suggests that the world of science can be divided into two worlds which we have called, respectively, Left World and Right World. This Matthew effect based affiliation of a country to either world depends only weakly on the field under investigation. We have introduced a measure relative national loss (win) of citations as a basis for ranking countries. According to this measure and to the form of the rank distribution, a low absolute (but high relative) national loss of each of the many Left World countries is equal to a high absolute (but low relative) national win of each of the few Right World countries. A significant finding is mixed membership of the two worlds. Big countries do not automatically belong in the Right World, neither do small countries automatically belong in the Left World. The challenges of the (now measured) stratification of countries in the world(s) of science are at least twofold. First, on the basis of our findings, national science policies might be checked for ways of improving a country's scientific performance especially of Left World countries. Secondly, the Matthew effect itself needs to be explained and understood. The latter is no easy task : the effect is undoubtedly a collective phenomenon which is deeply embedded in the citation behaviour of individual scientists in millions of papers in thousands of scientific journals. In the paper we describe the method of detecting the Matthew effect for countries at the macrolevel of science where it can be unambigously measured. However, the Matthew effect has its roots in the microlevel of science. The country rank distributions (according to citations per paper) of the individual journals are governed by freedom : they do not necessarily reproduce the macrolevel distribution. Based on our findings to date, our conviction is that the Matthew effect for countries penetrates the whole stratification system of science, from the micro- to the macrolevel, but it is unambigously expressed and measurable only at a high macrolevel. Examples of the Matthew effect for countries (MEC) for the Scandinavian countries are given.