Regulating public and private interactions in public sector digitalisation through procurement

As discussed in previous entries in this blog (see here, here, here, here or here), public procurement is progressively being erected as the gatekeeper of the public interest in the process of digital technology adoption by the public sector, and thus positioned as digital technology regulator—especially in the EU and UK context.

In this gatekeeping role, procurement is expected to ensure that the public sector only acquires and adopts trustworthy technologies, and that (private) technology providers adhere to adequate technical, legal, and ethical standards to ensure that this is the case. Procurement is also expected to operate as a lever for the propagation of (soft) regulatory tools, such as independently set technical standards or codes of conduct, to promote their adoption and harness market dynamics to generate effects beyond the public sector (ie market-shaping). Even further, where such standards are not readily available or independently set, the procurement function is expected to formulate specific (contractual) requirements to ensure compliance with the overarching regulatory goals identified at higher levels of policymaking. The procurement function is thus expected to leverage the design of public tenders and public contracts as tools of digital technology regulation to plug the regulatory gap resulting from the absence of binding (legal) requirements. This is a tall order.

Analysing this gatekeeping role and whether procurement can adequately perform it is the focus of the last part of my current research project. In this latest draft book chapter, I focus on an analysis of the procurement function as a regulatory actor. The following chapter will focus on an analysis of procurement rules on the design of tender procedures and some elements of contractual design as regulatory tools. Combined, the analyses will shed light on the unsuitability of procurement to carry out this gatekeeping role in the absence of minimum mandatory requirements and external oversight, which will also be explored in detail in later chapters. This draft book chapter is giving me a bit of a hard time and some of the ideas there are still slightly tentative, so I would more than ever welcome any and all feedback.

In ‘Regulating public and private interactions in public sector digitalisation through procurement: the clash between agency and gatekeeping logics’, my main argument is that the proposals to leverage procurement to regulate public sector digitalisation, which seek to use public sector market power and its gatekeeping role to enforce standards of technological regulation by embedding them in public contracts, are bound to generate significant dysfunction due to a break in regulatory logic. That regulatory logic results from an analysis of the procurement function from an agency theory and a gatekeeping theory perspective, which in my view evidence the impossibility for procurement to carry out conflicting roles. To support this claim, I explore: 1) the position of the procurement function amongst the public and private actors involved in public sector digitalisation; 2) the governance implications of the procurement function’s institutional embeddedness; and 3) the likely (in)effectiveness of public contracts in disciplining private and public behaviour, as well as behaviour that is mutually influenced or coproduced by public and private actors during the execution of public contracts.

My analysis finds that, in the regulation of public-private interactions, the regulatory logic underpinning procurement is premised on the existence of a vertical relationship between the public buyer and (potential) technology providers and an expectation of superiority of the public buyer, which is thus (expected to be) able to dictate the terms of the market interaction (through tender requirements), to operate as gatekeeper (eg by excluding potential providers that fall short of pre-specified standards), and to dictate the terms of the future contract (eg through contract performance clauses with a regulatory component). This regulatory logic hits obvious limitations when the public buyer faces potential providers with market power, an insufficient offer of (regulated) goods and services, or significant information asymmetries, which result in a potential ‘weak public buyer’ problem. Such problem has generally been tried to be addressed through procurement centralisation and upskilling of the (centralised) procurement workforce, but those measures create additional governance challenges (especially centralisation) and are unlikely to completely re-establish the balance of power required for the effective regulation by contract of public sector digitalisation, as far as the provider side is concerned.

Parking the ‘weak public buyer’ problem, my analysis then focuses on the regulation of public-public interactions between the adopting public sector entity and the procurement function. I separate them for the purposes of the analysis, to point out that at theoretical level, there is a tension between the expectations of agency and gatekeeping theories in this context. While both of them conceptualise the relationship as vertical, they operate on an opposite understanding of who holds a predominant position. Under agency theory, the public buyer is the agent and thus subject to the instructions of the public entity that will ultimately adopt the digital technology. Conversely, under gatekeeping theory, the public buyer is the (independent) guarantor of a set of goals or attributes in public sector digitalisation projects and is thus tasked with ensuring compliance therewith. This would place the public buyer in a position of (functional) superiority, in that it would (be expected to) be able to dictate (some of) the terms of the technological adoption. This conflict in regulatory logics creates a structural conflict of interest for the procurement function as both agent and gatekeeper.

The analysis then focuses on how the institutional embeddedness of procurement exacerbates this problem. Where the procurement function is embedded in the same administrative unit or entity that is seeking to adopt the technology, it is subjected to hierarchical governance and thus lacks the independence required to carry out the gatekeeping role. Similarly, where the procurement function is separate (eg in the case of centralised or collaborative procurement), in the absence of mandatory requirements (eg to use the centralised procurement vehicle), the adopting public entity retains discretion whether to subject itself to the (gatekeeper) procurement function or to carry out its own procurement. Moreover, even when it uses centralised procurement vehicles, it tends to retain discretion (eg on the terms of mini-competitions or for the negotiation of some contractual clauses), which also erodes the position of the procurement function to effectively carry out its gatekeeping role.

On the whole, the procurement function is not in a good position to discipline the behaviour of the adopting public entity and this creates another major obstacle to the effectiveness of the proposed approach to the regulation by contract of public sector digitalisation. This is exacerbated by the fact that the adopting public entity will be the principal of the regulatory contract with the (chosen) technology provider, which means that the contractual mechanisms designed to enforce regulatory goals will be left to interpretation and enforcement by those actors whose behaviour it seeks to govern.

In such decentred interactions, procurement lacks any meaningful means to challenge deviations from the contract that are in the mutual interest of both the adopting entity and the technology provider. The emerging approach to regulation by contract cannot properly function where the adopting public entity is not entirely committed to maximising the goals of digital regulation that are meant to be enforced by contract, and where the public contractor has a concurring interest in deviating from those goals by reducing the level of demand of the relevant contractual clauses. In the setting of digital technology regulation, this seems a likely common case, especially if we consider that the main regulatory goals (eg explainability, trustworthiness) are open-ended and thus the question is not whether the goals in themselves are embraced in abstracto by the adopting entity and the technology provider, but the extent to which effective (and costly or limiting) measures are put in place to maximise the realisation of such goals. In this context, (relational) contracts seem inadequate to prevent behaviour (eg shirking) that is the mutual interest of the contractual parties.

This generates what I label as a ‘two-sided gatekeeping’ challenge. This challenge encapsulates the difficulties for the procurement function to effectively influence regulatory outcomes where it needs to discipline both the behaviour of technology providers and adopting entities, and where contract implementation depends on the decentred interaction of those two agents with the procurement function as a (toothless) bystander.

Overall, then, the analysis shows that agency and gatekeeping theory point towards a disfunction in the leveraging of procurement to regulate public sector digitalisation by contract. There are two main points of tension or rupture with the regulatory logic. First, the regulatory approach cannot effectively operate in the absence of a clear set of mandatory requirements to bind the discretion of the procurement function during the tendering and contract formation phase, as well as the discretion of the adopting public entity during contract implementation phase, and which are also enforceable on the technology provider regardless of the terms of the contract. Second, the regulatory approach cannot effectively operate in the absence of an independent actor capable of enforcing those standards and monitoring continuous compliance during the lifecycle of technological adoption and use by the public sector entity. As things stand, the procurement function is affected by structural and irresolvable conflicts between its overlaid roles. Moreover, even if the procurement function was not caught by the conflicting logics and requirements of agency and gatekeeping (eg as a result of the adoption of the mandatory requirements mentioned above), it would still not be in an adequate position to monitor and discipline the behaviour of the adopting public entity—and, relatedly, of the technology provider—after the conclusion of the procurement phase.

The regulatory analysis thus points to the need to discharge the procurement function from its newest gatekeeping role, to realign it with agency theory as appropriate. This would require both the enactment of mandatory requirements and the subjection to external oversight of the process of technological adoption by the public sector. This same conclusion will be further supported by an analysis of the limitations of procurement law to effectively operate as a regulatory tool, which will be the focus of the next chapter in the book.

New Paper (in Spanish) on Agency Theory and Conflicts of Interest in Public Procurement

I have written a paper in Spanish (a rare occurrence) for the special issue on law and economics of the Revista de Economía Industrial, which I have now posted on SSRN as 'La Aplicación de la Teoría de Agencia a la Prevención de Conflictos de Interés en la Contratación Pública bajo la Directiva 2014/24' [Application of Agency Theory to the Prevention of Conflicts of Interest in Public Procurement Under Directive 2014/24] (September 22, 2015). Revista de Economía Industrial, número monográfico sobre Análisis Económico del Derecho. http://ssrn.com/abstract=2663947. Its abstract is as follows:

ABSTRACT
Law and economics analyses of public procurement have provided important contributions regarding contract design, particularly from the perspective of bidders’ incentives, as well as facilitated the formalisation of studies on collusion and corruption in settings of strict transparency obligations and rigid demand planning requirements. This paper does not focus on any of those facets of economic analysis of public procurement, but rather on the less developed application of agency theory to the activities of the public buyer. Building upon the contributions by Trepte (2004) and Yukins (2010), this paper explores the varied dimensions in which public procurement gives rise to agency problems, as well as some of the solutions to remedy them developed in Directive 2014/24.

RESUMEN
El análisis económico del derecho de la contratación pública ha dado lugar a importantes contribuciones relacionadas con el diseño de los contratos desde el punto de vista de los incentivos de los licitadores, así como a avances en la formalización de estudios relacionados con la colusión y la corrupción en escenarios sujetos a estrictas normas de transparencia y a rígidos ejercicios de planificación de la demanda. Este artículo no se fija en ninguna de estas facetas, sino en la menos desarrollada aplicación de la teoría de agencia a las actividades del comprador público. Basándose en las contribuciones de Trepte (2004) y Yukins (2010), el artículo explora las varias dimensiones en que la contratación pública se ve afectada por problemas de agencia, así como algunas de las soluciones que la nueva Directiva 2014/24 ha desarrollado para tratar de remediarlas.

The full paper can be accessed here: http://ssrn.com/abstract=2663947.