The general impression is thus that, despite their potential problems of subjectivity, they seem to be capturing very closely related phenomena. Treisman elaborates "…perceptions – even if not matched by reality – can have powerful effects. Mauro (1995) found that perceived corruption reduced growth via its effect on investment but found much weaker evidence of effects going via other paths, such as increased costs or reduced productivity. Other studies have confirmed that higher perceived corruption reduces foreign direct investment. Soares (2004) found that high
perceived corruption discourages victims from reporting crimes to the police. So perceptions clearly matter." For these reasons, this report proposes that EU regional and country level differences in quality of government be measured to the greatest extent through perception-based indicators – and, in particular, through an EU-wide survey of households in which regular “users of government” evaluate them according to this definition of QoG