par Viadere, Elise 
Président du jury Wolff, Guntram
Promoteur Cantillon, Estelle
Publication Non publié, 2026-09-25
Président du jury Wolff, Guntram
Promoteur Cantillon, Estelle
Publication Non publié, 2026-09-25
Thèse de doctorat
| Résumé : | Prices coordinate economic activity by translating costs into incentives. When they accurately reflect the social costs of individual choices, private decisions can support socially efficient outcomes. The more consequential cases arise when this alignment breaks down. Prices may fail to capture costs that vary across time, location, or consumers. Consumers may be unable to adjust even when prices are informative, and the firms that set prices may have objectives that differ from social efficiency. Therefore, what a price signal achieves depends on its accuracy, but also on the response it induces, on how those responses interact in aggregate, and on the incentives of the intermediary that designs it. This dissertation studies these questions in retail electricity, a market that displays all three problems in an especially clean form.Unlike most goods, electricity must be produced and consumed at the same instant, and its marginal cost changes sharply from one hour and one location to the next. The residential electricity consumer, meanwhile, typically pays a price that does not change within the day at all. This gap has always been a distortion, but for a long time it carried little cost: generation was stable and largely dispatchable, so the cost of averaging over the day was low, and residential demand was inelastic. That condition no longer holds. Intermittent renewables widen the changes in the underlying marginal cost of electricity, and the electrification of transport and heating creates new uses of electricity whose timing can be shifted through the day. The flat electricity price used to be a relatively minor distortion. Today, however, it affects many of the outcomes that matter for the energy transition: what it costs to run the system, how heavily the network is used, and who gains from decarbonisation and who pays for it.This makes the timing of electricity demand increasingly important. If consumers can shift some consumption away from costly or congested periods, they can help the system adapt to these new conditions. This responsiveness to price, called \textit{flexibility} or demand response, has become valuable to the electricity system, and unlocking it has become a central aim of energy policy.The price paid by an electricity consumer is made of several components. The energy component pays for the electricity supplied and is set by the retailer, which buys electricity on the wholesale market and resells it to consumers. Network tariffs pay for the transmission and distribution infrastructure and are set within a regulatory framework. Taxes and levies add a further component determined by public authorities. The price signal faced by the consumer reflects a combination of market and regulatory choices. These different components can all be used to incentivise flexibility. Three questions stand out. Can consumers respond to the signal at all? How accurate does the signal need to be? And can the retailer shape that signal to its own advantage? The chapters that follow take up these questions in turn.The contribution is threefold. First, the chapters measure the value of flexibility by the response it produces: when consumption is rescheduled, by how much, and what this does to system cost. Second, this dissertation asks who gains from flexibility. Access to the technologies and incentives that make flexibility possible is unevenly distributed, and those who create the value of flexibility are not always those who capture its benefits. Third, it studies the institutions and market power that stand between the price and the consumer: regulators set the network tariffs and retailers set the electricity prices. These choices shape both how much flexibility is incentivised and how its gains are distributed.The first chapter analyses network-tariff adjustments for energy-sharing communities and tests their effects in a case study of a Belgian pilot. Energy communities pool locally generated renewable electricity among their members, who pay a reduced price for what they consume locally. Several European Union member states support these communities by lowering their electricity network tariffs. The chapter asks two questions. Does a community need the support to break even? And are the adjustments cost-reflective: does the community's lower tariffs match the lower network costs it actually causes? The distinction matters because the network must still be financed, at a time when the energy transition requires substantial investment in its expansion and reinforcement. When a network-tariff reduction is not cost-reflective, the shortfall must be recovered from other network users. Results show that support can be necessary for financial viability. Even a community with high self-consumption may not break even without network-tariff adjustments. However, in the community studied, the support did not affect consumption patterns. The energy community had neither real-time feedback on network constraints nor significant electricity consumption it could reschedule. The network-tariff adjustments lowered the community's costs without lowering the network costs the community imposes, and the difference is borne by other consumers. Therefore, the chapter states that support aimed at ensuring a community’s financial viability should be kept distinct from incentives intended to reward measurable network benefits.The second chapter asks what is lost when real-time pricing is simplified. Real-time pricing is the efficient benchmark, since it exposes consumers to prices that reflect hourly marginal costs. However, it has proven difficult to deploy at scale. Time-of-use pricing offers a simpler and more practical alternative: instead of prices changing every hour, the day is divided into a small number of fixed periods, such as peak and off-peak hours. The question is whether this simpler price structure remains a good alternative to activate flexibility. The answer depends on a feature usually left out in existing research. Flexible consumer tasks run for several consecutive hours, so the consumer chooses not an hour but a schedule. Modeling the decision this way separates three things: whether the price incentivizes accurately, how much the consumer gains from shifting, and what her shift does to system costs once many consumers respond simultaneously. From this perspective, time-of-use pricing is a good alternative only under specific conditions. It preserves the direction of the real-time price signal while weakening its strength, and it can concentrate demand in the same hours. Simulations of the Belgian market show that this coordination can lower system costs, but can also create new demand peaks and raise costs for consumers who do not respond.The third chapter, written with Leticia Pieraerts, asks how a monopolist retailer designs electricity contracts. Wholesale electricity is more expensive in some hours than in others, but retailers decide how much of this cost difference is reflected in the prices paid by consumers. We call this the contract's \textit{pass-through rate}. A flat price contract passes none of the difference through, while a fully dynamic contract passes it through in full. Consumers do not value this exposure equally. A flexible consumer can respond to a larger price difference by shifting consumption toward cheaper hours, while a less flexible consumer finds it more costly. The monopolist retailer can exploit this difference when designing its contracts. Pass-through becomes a screening device: the retailer gives the most flexible consumers the full, cost-reflective signal and deliberately weakens it for the less flexible, so that their contract does not attract the flexible type and the rent left to that type remains small. This distortion is not a consequence of private information alone. Marginal-cost pricing would give every consumer full pass-through, and perfect competition would do the same. The compression of the price signal arises because the retailer has market power and uses contract design to limit the rent of flexible consumers. The resulting distortion falls on the less flexible, while most of the surplus created by flexibility accrues to the retailer. As more consumers become flexible, a larger share of the gains can be captured by the retailer rather than passed on to consumers. The chapter studies how much of the price signal reaches consumers and who keeps the gains.The three chapters point to the same conclusion. Prices do not activate flexibility on their own. What activates it is the surrounding set of institutions and incentives that determine who is able to respond, how, and who is paid for responding. Three requirements follow for any policy meant to encourage retail electricity market flexibility. Policies should be cost-reflective, with support tied to measurable reductions in system or network costs and price signals sufficiently granular that they neither create arbitrary transfers nor induce shifting that provides no system benefit. Policies should be designed considering that consumers differ in their ability to respond, in the risk they can bear, and in the technology they own. And policy makers should assume that intermediaries will act strategically. Therefore, these chapters point to market design as a central determinant of how flexibility is activated, allocated, and valued. |
