At the Munich Security Conference 2017, authors from the MSC, McKinsey & Company and the Hertie School of Governance published a joint paper entitled "More European, More Connected and More Capable". The paper (download here) also included an analysis in which the number of different weapon systems in the EDA states + Denmark (=EU28) is compared with those in the USA and the efficiency of armaments in the EU is derived in comparison with the USA. One of the core statements refers to the number of different types of battle tanks. One can read, for example, that the EU would allegedly operate 17 different types, while the USA would manage with a single type.
The source of this statement is always the study "The Military Balance" published annually by the British research institute IISS (International Institute for Strategic Studies).
The general statement that the EU would afford several types for the same purpose and thus lose efficiency to the US is obvious. However, the statement repeatedly made in discussions and lectures that the EU allegedly had 17 types of tanks and the USA only one is wrong. It seems that neither security policy experts, security policy authors nor high-ranking officials and soldiers in the BMVg have taken a critical look at this ratio, despite the widespread recognition that a large number of EU states use the German bestseller Leopard 2 and that several EU states have no tanks at all. This alone should have been enough to critically question the 17:1 ratio.
Following the logic that one can only compare the same with the same and that only one "type" is given for the U.S. Armed Forces, the comparison can only be between Main Battle Tanks (MBT). Wheeled tank destroyer or armoured infantry fighting vehicles, of which the U.S. armed forces also have systems in service, must therefore not be counted, as otherwise the number 1 on the US side would not be correct. The same applies to versions of a battle tank, since the US Armed Forces use the Abrams both in the version M1A1 and in different M1A2 variants. Thus only types of battle tanks (Leopard 2, Leclerc,...), but not individual versions or local "derivatives" of the same tank type (of which there are some differences in the T-72 series, e.g. the M 84 in Croatia), should have been counted. If this were done, we would have five "different" main battle tanks in use in the Bundeswehr alone with the versions Leopard 2 A5, A6, A6M, A7 and A7V.
In relation to different types, the EU 28 has interpreted very strictly eleven and not 17 different MBTs in service (without Cyprus and Greece there would only be six). If, on the other hand, you were to count each version individually, it would be considerably more than 17, but on the US side there would also be several versions. Which of the individual versions of the T-72s or Leopard 2 series represent individual types, cannot be answered clearly. To do this, it would first have to be decided which modifications automatically result in a new type. The result is thus: The ratio 17:1 is definitely not correct.