This paper shows that there is a discontinuity in the representation of Argus, the guardian of Io: while in the earliest literary source, the Aegimius ([Hes.] fr. 230 Most), and in the sixth century iconography (LIMC V 664.1, 667.31) we find the conception of Argus as a two-faced monster with four eyes (three according to Pherecydes fr. 66 Fowler), all fifth-century and later sources depict Argus as a giant with thousands of eyes dappling his entire body (Bacch. 18.19-25; Aesch. Suppl. 305; [Aesch.] Pr. 568, 677, etc.). The time period in which this sudden change is observed suggests the following hypothesis: the image of the myriad-eyed cowherd was imported from Achaemenid Iran. Around the turn of the fifth century Greek craftsmen of all kinds had access to the court of the Persian king and it is precisely at that time that a considerable Iranian influence on Greek philosophy and literature can be detected (as shown, above all, by W. Burkert and M. L. West). It is argued that the source for the many-eyed representation of Argus in the fifth century was the Iranian deity Miθra who had a cult in Persepolis: Mithra’s standing epithet is ʻhe who has ten thousand eyesʼ. In Avestan texts Miθra is said to be all-seeing and ever awake, just like Argus, and his vigilance is repeatedly emphasized. Another epithet of Miθra is ʻlord of wide pasturesʼ; he is the quintessential guardian. Miθra is associated with starry sky just like Argus. Given the prominence of the cow in Zoroastrianism, it becomes possible to propose a scenario in which a Greek in the sixth century could have acquired an incomplete picture of a Persian triad corresponding to Zeus, Argus and Io, where Argus matches Miθra, the myriad-eyed cowherd.
Political reforms are often designed in a gradual manner, even though it would
be more e¢ cient to make use of available complementarities and implement the
reforms as "big bang."We show that this ine¢ cient behavior can be explained by
a simple citizen candidate model where voters are time-inconsistent. The result is
not only that gradualism arises in equilibrium, but that it is second-best. Grad-
ualism is ine¢ cient compared to the rst-best outcome where complementarities
are not lost and the reform is carried out at the optimal point in time. However,
it is welfare-enhancing compared to the case where gradualism is not an available
option, because then the reform would have been delayed even further. We also
show that when voters are sophisticated (i.e. aware of their time-inconsistency),
they elect an agenda setting politician who is more patient than the median voter
in order to commit to a reform schedule where the negative consequences of the
time-inconsistency are less severe.
Selection by quotas is an important policy measure in the affirmative action tool box. However, quotas may come with unintended side effects, for example by causing uncooperative behavior in the group formed with quota-based selection rules. In the laboratory I measure the impact of a quota on group cooperation, and examine the underlying mechanisms. Two groups are created by randomly assigning participants to either an orange or a purple group. In the unrepresentative quota treatment, orange participants are chosen as members of a selected group by performance on a simple unrelated math task whereas purple participants are chosen based solely on the quota. I compare contributions in a public good game in this unrepresentative quota treatment to behavior in a control treatment, where the orange and purple participants are treated symmetrically and all members of the selected group are chosen based on performance on the unrelated math task. My results show significantly less cooperation in the quota treatment and I furthermore find that this tendency is observed in both the meritocratically chosen orange participants and the quota-advantaged purple participants, and regardless of the color of the matched player. The reduced cooperation remains even when participants are given a rationale for the unrepresentative quota, e.g., by appealing to a fairness argument. The negative effect on cooperation from the unrepresentative quota disappears when selection is done completely randomly instead of on the basis of performance.