The radial electric field E-r(x, t), and particularly its gradient, has been invoked by various theories and empirical models as a crucial parameter 'per se' for determining the transition to high confinement regimes, such as the onset of an internal transport barrier (ITB) in the plasma core and of the H-mode pedestal at the plasma edge. This idea, however, does not consider the basic fact that in most experiments the transition to a steady-state higher confinement regimes is produced by applying sufficient additional heating onto a given target density and current profile. In order to test this ansatz on a more routine basis, we have developed here an analytical approximation to the neoclassical calculation of the radial electric field, adapted for the 2D toroidal geometry of JET to describe all collisionality regimes (banana, banana-plateau, Pfirsch-Schluter) and to include averaging over the potato orbits. An analytic calculation of the error bars on E-r (x, t) has also been developed, which has allowed us to compare and successfully benchmark our calculations with the results of neoclassical codes such as JETTO and NCLASS. We are then able to demonstrate a striking similarity in the shape of E-r (x, t) in steady-state L-mode, H-mode and ITB plasmas when normalizing E-r (x, t) with respect to the total heating power flux. This clearly indicates that, experimentally, there is no direct causality relation between changes in E-r (x, t) and steady-state improved confinement, as these are brought about together by changes in the power deposition profile. Only two cases do not satisfy this general rule. First, localized and rapid transients (i.e. occurring on time scales much shorter than the momentum and energy confinement time) could be linked to non-neoclassical changes in E-r (x, t), possibly due to turbulence suppression mechanisms. Second, when comparing H-mode plasmas with forward and reversed ion del B-drift direction, we demonstrate the role of prompt fast ion losses in affecting E-r(x, t), most likely due to the different edge flows and their cascading effect onto the core plasma.