2.1 The therapeutic relationship as a key factor

Empirical psychotherapy research, across different schools of thought, agrees that the therapeutic relationship is a central factor in the effectiveness of psychotherapy (Grawe, Donati & Bernauer, 1994).

In an academic essay, Edward Bordin (1979, p. 253) describes how the effectiveness of therapy depends, at least in part, if not entirely, on the strength of the therapeutic alliance. He identifies three components that constitute this relationship:

• agreement on the therapeutic goals (goals),

• consensus on the tasks and methods within the therapeutic process,

• the emotional quality of the relationship and the development of an affective bond.

Whilst the first two levels (goals and tasks) can be described more as cognitive processes, the bond between the two primarily requires emotional qualities such as mutual trust and acceptance.

‘[…] when attention is directed towards the more protected recesses of inner experience, deeper bonds of trust and attachment are required and developed.’ (Bordin, 1979, p. 254, abridged)

Trust should therefore characterise every therapeutic relationship. However, a healing encounter in psychotherapy requires more: a transformative depth, mutual trust and a bond that can only arise through a jointly developed intersubjective relational framework, for it is only this framework that makes it possible to direct attention towards protected inner experiences and spaces of retreat.

Bordin’s hypothesis has been empirically confirmed over the following decades. In 1991, Horvath and Symonds presented the first systematic meta-analysis on the correlation between the quality of the therapeutic alliance and therapeutic success.

In a large-scale meta-analysis (295 studies involving over 30,000 patients, with primary research conducted between 1978 and 2017), Flückiger et al. (2018) likewise confirmed the link between the therapeutic relationship and therapeutic success, with an aggregated effect size of r = .278 (95% CI [.256, .299]). Another notable finding of this study is that this also applies to internet-based psychotherapy, and that digital formats do not significantly influence this correlation.^{1Footnote 1 The fact that communication takes place via a digital medium does not alter the fact that a physical subject remains involved at both ends. Chapter 4 returns to this distinction.}

With his ‘Contextual Model’, Wampold (2015, pp. 270–272) provides a theoretical framework that underpins these findings. He describes three pathways of action, all of which are relationship-dependent: the ‘real relationship’ as an authentic interpersonal connection, the formation of expectations through a shared understanding, and the joint performance of healing actions. Wampold makes the surprising observation that neither adherence (following the manual) nor competence (competent implementation of the technique) correlates significantly with therapeutic success. It is therefore not the specific method or technique, but the quality of the relationship between therapist and client that is the decisive factor for therapeutic success.

The strength of this correlation proves to be remarkably stable: Flückiger et al. (2018) show that the correlation remains largely constant across the therapeutic schools, assessment instruments and evaluator perspectives examined. It is precisely this invariance that contradicts the interpretation that the relationship is merely a by-product of a particular method.

In a further meta-analysis (60 samples, 6,061 participants), Flückiger et al. (2020) examine the obvious objection that the correlation might in fact be based on other factors: the clients’ initial distress or the extent to which the technique was implemented faithfully according to the manual and competently. When these factors are statistically controlled for, the association between therapeutic alliance and success remains stable (r = .23 to .31). It cannot, therefore, be attributed either to favourable baseline conditions or to the technique being implemented effectively.

There is therefore exceptionally strong empirical support for the effectiveness of the therapeutic relationship. How it works, however, remains a matter of debate: Horvath and Symonds (1991, p. 147) had already noted a ’significant lack of unanimity with regard to how the alliance operates’. Almost three decades later, Flückiger et al. (2018, p. 14) still describe the question of causality as ’important and controversial’. Wampold (2015, p. 275) even refers to the findings on adherence and competence as a ’mystery’.