European Journal of International Relations 10(1) Table 1 Possible Iconographic Choices Goal Post-Materialist Actor Traditional Materialist (or Post-Modern) State State symbols in Historical heads Representatives of classical/antique of state,generals, 'official culture' garb,classical other statesmen (e.g.Francis Scott gods (e.g.Athena,(e.g.George Key,Edward ‘Columbia),or Washington, Teller) purely ornamental Douglas figures MacArthur) Society/Classes Classical/antique Imagery of real- Imagery of real- imagery of mass, world mass,class, world mass,class, class,sector or sector or region sector or region region representatives at representatives at representatives work (e.g.farmers play (e.g.kids (e.g.allegorical in the fields) playing baseball) figures 'Industry', 'commerce'） Individual Historical non- Historical non- Historical non- state actors from state actors who state actors who classical/antique made significant made significant era or social or contributions to representations of economic the sciences or individual-level contributions the arts and virtues (e.g. (e.g.Margaret letters (e.g.Albert Aristotle,St Paul, Sanger,Henry Einstein,Charles 'Faith') Ford) Ives,Herman Melville) obvious centrality of human figures in currency iconography over time(the uninhabited scenes in the euro are a clear anomaly to this tradition,to be discussed in Section 4),and second,the relative ease of identification and interpretation of such human figures-a necessary component of a project covering many decades and countries.The decision to focus on human figures at this stage of the research should not,however,be taken to imply that they are the only interesting or relevant aspects of currency iconography for the overall theoretical issues spelled out above.Rather,the human figures simply serve as a reasonable proxy for the overall message of the banknotes they grace. 10 Reproduced with permission of the copyright owner.Further reproduction prohibited without permission
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
Hymans:The Changing Color of Money The database built on the basis of the Standard Catalog data can be described as follows.For the current EU 15 plus East Germany,I collected information on all regular,mass-produced,central bank-issued banknotes (not Treasury Bills or other interest-bearing notes).I made a separate entry in the database for each banknote that has a separate catalog entry number in the Standard Catalog.+I recorded separately every human figure(named and unnamed)on the front or back of the notes.5 Then I located biographical information about any named individuals pictured on the banknotes,including their historical importance and date of birth.This work was greatly facilitated by a massive online biographical index (Inter- nationaler Biographischer Index,1999).Thanks to this and other sources,I currently lack identifying biographical information for only 15 images(14 different historical figures),mostly from Spain and Portugal from around the turn of the 20th century. The database covers 1368 separate banknotes from the 15 current EU member states plus East Germany.Of those 1368 separate banknotes,1174 depict human figures.On the 1174 banknotes with human figures,I counted 1424 such human figures.Of these,I was able to track down the biographical data for 1409 of them,and it is these 1409 images that serve as the key data set for my empirical analysis.The 194 banknotes without human figures mainly come from one of three sources-the period before 1920 when many central bank printing facilities were still in their infancy, the German hyperinflationary period of the 1920s and 1930s,or very small denominations. Table 2 breaks down these totals by country,noting also the start date of each banknote series in the data set.The far-right hand column is the key one;it represents the set of 1424 images that form the basic data set on which I perform my analysis.This includes many banknotes with only very slight differences between them.But it may surprise those expecting iconographic stickiness'to note that in most countries iconographic change has been more common than iconographic stasis.Overall,out of the 1424 images in the data set,846 or 59%depict a different person than the previous note at the same denomination.There is variation across countries on this score,but in the data set only Great Britain has had a strong tendency to issue new banknotes featuring the same human figures as before.Apart from Great Britain's 15%score on this measure,no other country scores lower than 34%,and only two,Ireland and Sweden,score lower than 40%.Thus it would appear that there is little if any inherent obstacle to iconographic change.Given this finding,even if the imagery does not change it is reasonable to assume that this stasis was the result of a conscious choice. 11 Reproduced with permission of the copyright owner.Further reproduction prohibited without permission
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
European Journal of International Relations 10(1) Table 2 Country-by-Country Summary of Banknotes in Data Set No.with Total Human Country Starting Date No.of Human Figures Banknotes Figures Austria 1825 115 104 123 Belgium 1851 75 75 98 Denmark 1875 55 40 43 Finland 1922 77 48 50 France 1800 98 91 118 GDR 1948 27 12 12 Germany 1876 157 99 120 (incl.FRG) Great Britain 1855 134 132 153 Greece 1841 100 96 129 Ireland 1928 44 44 79 Italy 1896 70 70 74 Luxembourg 1929 16 16 16 Netherlands 1904 64 48 53 Portugal 1847 120 115 125 Spain 1874 151 135 147 Sweden 1874 65 49 84 Totals 1368 1174 1424 Testing the Hypotheses I coded the images on banknotes along the two Meyer/Inglehart-inspired axes-locus of actorhood and nature of goals.The basic coding scheme follows the examples from Table 1.In this section I first present the overall statistical results,and then I discuss those results. Before proceeding to the results,however,I should note some important coding decisions.Two decisions in particular had a significant effect on the overall results-the decisions to code purely ornamental human figures and classical gods as depicting state actorhood.The reasons for these decisions are as follows.In the case of ornamental figures,they are ornamenting state- produced banknotes that themselves are a symbol of state.6 Meanwhile, classical gods,as the rulers of the universe,clearly evoke the functions of state.?There were other coding ambiguities as well,but these were essentially irrelevant to the overall contours of the results.8 Table 3 answers the question of whether there has been secular change in the types of actors depicted in currency images.The results are shown in 12 Reproduced with permission of the copyright owner.Further reproduction prohibited without permission
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
Hymans:The Changing Color of Money Table 3 A Secular Change in the Locus of Actorhood? Actor Epoch of Currency Issue State Society Individual Total N Pre-1920 418 49 56 523 (80%) (9%) (11%) (100%) 1920-49 255 171 51 477 (53%) (36%) (11%) (100%) 1950-79 116 39 101 256 (45%) (15%) (39%) (100%) 1980-on 53 11 89 153 (35%) (7%) (58%) (100%) Total N 842 270 297 1409 (60%) (19%) (21%) (100%) Pearson's chi-squared (6):364.808;p =0.000 terms of four 30-year periods.The utility of this periodization was identified inductively.This table shows that indeed there has been a secular shift from the state to the individual as the locus of actorhood.Depictions of the state start out clearly dominant in the pre-1920 period and subsequently decline smoothly.Beginning around 1920 there is a boom in depictions of society; this then fades away after around 1950 with the dramatic rise of depictions of individuals.The results of the chi-squared test show that we can reject the null hypothesis that the row and column data are unrelated. Table 4 answers the question of whether there has been a secular change in the nature of goals depicted by currency images.This table shows that indeed there has been a secular shift.The use of classical images(represent- ing traditional goals)starts out dominant and then gradually vanishes into thin air.The use of historical images(representing materialist goals)begins strongly in the 1920s;it subsequently holds its ground in percentage terms down to the present day.This persistence represents a minor anomalous finding for the 'cultural shift'hypothesis,which would expect representa- tions of materialist goals to decline in the more recent period.Meanwhile the use of cultural images (representing post-materialist goals)begins strongly in the 1950s and then soars in the current period.The results of the chi-squared test show that we can reject the null hypothesis that the row and column data are unrelated. 13 Reproduced with permission of the copyright owner.Further reproduction prohibited without permission
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
European Journal of International Relations 10(1) Table 4 A Secular Change in the Nature of Life Goals? Goal Post- Epoch of Traditional Materialist Materialist Currency Issue (Classical) (Historical) (Cultural) Total N Pre-1920 402 91 30 523 (77%) (17%) (6%) (100%) 1920-49 239 211 27 477 (50%) (44%) (6%) (100%) 1950-79 45 131 80 256 (18%) (51%) (31%) (100%) 1980-on 3 69 81 153 (2%) (45%) (53%) (100%) Total N 689 502 218 1409 (49%) (36%) (15%) (100%) Pearson's chi-squared (6):525.726;p-0.000 As might be expected given the preceding results,the two key axes Meyer and Inglehart identify-the locus of actorhood and the nature of goals- work in tandem.In all,65%of images depicting the state also feature classical themes;56%of images depicting society also feature historical themes;and 69%of images depicting individuals also feature cultural themes.Overall,61%of the images are located down the left-to-right diagonal of Table 1,while a Cohen's kappa test shows that only 39%would have been expected there from random chance.The probability of this relationship occurring by chance is 0,and the kappa score is 0.41. Overall,the cultural shift'hypothesis is strongly supported by these findings.A qualitative look at the data also reinforces this conclusion.Here I offer a holistic appreciation;the interested reader can also consult the appendix to this article,which reprints images of Dutch banknotes to show the 'cultural shift'as it played out in that case (which is also discussed in Section 4 later). Consider some of the more popular images from different epochs: In the period before 1920,popular images include mythical female symbols of states in flowing robes,such as Great Britain's 'Britannia'(78 images),Sweden's 'Svea'(22 images),Germany's'Germania'(18 images), Austria's 'Austria'(13 images),Italy's 'Italia'and 'Roma'(3 images)and Portugal's 'Lusitania'(1 image).Also making a strong showing are 14 Reproduced with permission of the copyright owner.Further reproduction prohibited without permission
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission