“Electricity inflation” is a neologism coined by the website Phastidio. It describes the current and future state of the Western and global economies. It may also be a variant of stagflation (a combination of inflation and recession), which raged in the 1970s, but potentially more difficult to manage. At the time, stagflation was the result of the oil shock; this time, however, it is the result of a technological revolution.
A consequence of the 1970s phenomenon was that it ended tolerance for a minimum level of inflation, which in those years was accepted as the price to pay for growth. In fact, Nobel laureate in Economics Samuelson stated in Newsweek in February 1971: “Creeping inflation is the malaria of the modern mixed economy. In fact, just like malaria, inflation makes life difficult and does not heal on its own. However, unlike malaria, there does not seem to be a cure for creeping inflation that is better than the disease itself.” In essence, inflation was viewed as a necessary evil, almost as if it were a price to be paid for growth. That narrative was rapidly challenged by stagflation.
While stagflation was a positive opportunity for academia to rethink important aspects of economic theory, it was, conversely, a terrible time for economic policymakers, who found themselves forced to navigate between “inducing” a recession to control prices or accepting the resulting inflation. It was a veritable “devil’s crossroads”, which was later resolved, triggered by a geopolitical shock in the Middle East. But today, as fate would have it, the phenomenon of cost-driven inflation is once again emerging from the history books – where we had hoped to consign it – to threaten a new shock that compounds the existing geopolitical ones. This new threat is being produced by technology and an inflationary surge driven by electricity demand fueled by the data centers required for AI, with the risks of supply bottlenecks and a potential recession.
Unfortunately, the shock comes at a terrible time, namely, in a world where the “geopolitical costs” of the effect of the closure of the Strait of Hormuz, along with speculation about its duration, are having an increasingly heavy impact on prices. The investment bank Goldman Sachs has highlighted the importance of this issue, in a February 2026 report titled “The Macroeconomic Spillovers From AI Electricity Demand.” The study aptly notes that no choice, including that of AI, comes without a cost.
The report paints a rather pessimistic picture, especially when read in the current context. It is worth noting, however, that the argument Goldman Sachs makes retains some optimism; rather, it is at least more optimistic when projected into the future. In other words, the Lower Manhattan investment bank believes that over time, both the costs of electricity inflation and its redistributive effects (inflation has a greater impact on the budgets of low-income households) will be offset by the greater GDP benefits progressively generated by AI. The problem is, as analyst Seminerio argues, that the impact of this “technological-industrial leap” on energy bills could easily offset the anticipated benefits, thereby negating its positive effects on GDP.
Theories About Electricity Inflation
Goldman Sachs proposes an analysis of electricity inflation that differs from a typical monetary approach. Of course, ignoring money when explaining inflationary processes would be misleading (about its role, see M. Fratianni, F. Spinelli, Storia monetaria d’Italia, ETAS, 2001), in the sense that money is as necessary for inflation’s persistence as oxygen is for fire: without it, it goes out. That said, for explanatory purposes, price growth can be broken down into its various phases (for an analysis that is excessively non-monetary but interesting, see G. Vidal, W. C. Marshall, L. P. Rochon, “Structural Inflation and Conflict Inflation Today,” Moneta e credito, December 2025).
The New York investment bank attempts to do this with electricity inflation, identifying one cause – not the only one, but a significant one – as the financial needs of utilities. Goldman Sachs states:
We expect utility capital expenditure to increase significantly in response to the supply-demand imbalances we described above. Since utilities generally earn a pre-set return on investment from invested assets that is typically financed through higher prices (we discuss this further below), higher capital expenditure is likely to push up electricity inflation.
In short, pressure on prices, which is certainly linked to the increased demand for electricity from data centers, depends – and will continue to depend – on the financial needs of energy producers, who in turn tend to pass their costs on to consumer prices, thus ultimately to inflation. As mentioned, this has distributive consequences that deserve attention. It is also interesting that Goldman Sachs addresses the issue of electricity inflation in a way that recalls theories about the causes of 20th-century stagflation, developed by the scholar and trade unionist C. Levinson (Capitale, inflazione e imprese multinazionali, ETAS/Kompass, 1971).
Levinson’s thesis is that the paradox of simultaneous recession and inflation that emerged as a result of the first major oil crisis could be explained – given the rigidity of industrial prices in oligopolistic markets – by examining the financial choices made by producers to hedge against market uncertainties. Consequently, Levinson argues that to understand stagflation, one must analyze the relationship between cash flow (the ability of firms to generate positive cash flow even as demand declines) and inflation. Thus, if demand slows, prices rise and the volume of goods supplied decreases. Here we can draw an analogy with N. Kaldor’s model (“Inflation and Recession in the World Economy”, Economic Journal, vol. 86, 1976), which interprets the pricing policies of the industrial sector – an oligopolistic sector distinct from that of raw materials – which at the time was less financialized and more sensitive to the market. Provided, of course, as Levinson argues, that demand is inelastic in the face of higher prices. The scholar clarifies his thinking:
Cash flow maximization helps explain the neologism stagflation – ‘stagnation-inflation’ or ‘recession-in-inflation syndrome’ – as it underlies most of the price inflation being created today (op. cit., p. 173).
Here, a clear similarity emerges between Levinson’s approach from the 1970s and Goldman Sachs’ view on the formation of electricity inflation. Essentially, both relate to the use of prices to finance the expansion of investment in a sector (the electricity sector, in this case). That said, it is unlikely that the New York investment bank fully shares Levinson’s non-monetary approach.
Among other things, from the perspective of post-oil-crisis economic theory, Levinson’s approach has the serious limitation of underestimating the rebound effect that expectations have on inflation once it has taken hold. In this case, as the New York investment bank points out with this chart, this is further facilitated by the demand-driven momentum generated by the strong growth of the data center economy.
Goldman Sachs elaborates further on the topic of electricity inflation:
We expect power supply to increase only slowly to meet higher demand, as generation capacity takes a long time to be brought online. … There are several bottlenecks to faster supply growth. … In addition, the rush to increase generation capacity is running up against shortages of both specialized equipment like gas turbines and qualified labor, which takes several years to train.
Consequently, Goldman Sachs expects electricity prices to have a negative economic impact. It is worth emphasizing once again that, according to Goldman Sachs, this has distributive effects, particularly to the detriment of the most vulnerable social groups. This is because electricity costs account for a significant portion of their household budgets.
Therefore, the impact will be greater on this social group in terms of both income and spending. It is true that Goldman Sachs is referring to the United States; however, this is also of great concern to us, both in terms of development and geoeconomics (the geographical distribution of data centers). These are the inevitable costs of the AI revolution.
The Geoeconomics of Data Centers
The fact is that, like in the U.S., the geographical distribution of data centers is asymmetrical. This means that price hikes – along with the risk of supply bottlenecks – will disproportionately affect areas with the highest concentration of data centers. In this regard, as analyst Seminerio (Phastidio) points out, Italy is experiencing a real boom.
According to the ESG News website, demand for high-voltage connections exceeds that of 2023 by a factor of 13. The sequence is: AI accelerates digitalization, and this in turn drives energy demand. Geographically, this occurs asymmetrically. Consequently, the related energy demand is concentrated primarily in Lombardy and Piedmont, followed at a distance by Emilia-Romagna, Lazio, Apulia, and Veneto.
In fact, the Digitalization and Decarbonization Report 2025 from the School of Management at the Politecnico di Milano states that AI is undoubtedly a decisive lever for maximizing both the efficiency and competitiveness of the “Italian system.” At the same time, however, all of this translates into a new source of energy demand. The problem is that this increase in demand requires careful public policy.
This is because, if left unchecked, it can lead to both supply bottlenecks and, additionally, an increased source of CO₂ emissions. Furthermore, as the cited studies suggest, the geographical concentration of data centers (in Italy as well as in the U.S.) poses problems related to both inflation and energy equity. For example, looking at the Italian Peninsula, it is possible that areas with less digital infrastructure (concentrated mainly in the South) will bear the brunt, because they will suffer the costs of network upgrades without reaping direct benefits.
Above all, Italy faces an inevitable inflationary risk, partly due to its uncompetitive energy mix. In short, the green transition is not only an ecological challenge but also an economic one. Since inflation acts as a “reverse Robin Hood” tax, it is wise to pay close attention to it. Because it is the near, and already present, future.
In 1921, Lenin remarked that “communism was Soviet government plus electrification”, that is, a utopia founded on technology. Well, to paraphrase the Bolshevik leader, today one could say in some ways that AI is a utopia of government – even a warlike one – founded on data centers and their demand for electricity. Because prices, as he taught (Lineamenti di una storia monetaria d’Europa, Einaudi, 1981), are an important barometer of social change. This holds true for electricity prices as well.
L’articolo Electricity Inflation. The New Shock of the Digital Economy proviene da ytali..
4uMagazine
