The economic impact of the blackout in detail

In this article, we expand on the Brief Note assessing the economic impact of the blackout on 28 April, published on our website a few days after the entirety of mainland Spain was left without electricity. Based on aggregated and anonymised data on payments and cash withdrawals carried out with cards issued by CaixaBank, we have estimated the economic impact and taken a cross-section of that impact by sector and autonomous community region.

The analysis of CaixaBank data reveals that the blackout on 28 April caused a 34% fall in consumption spending by Spanish households that day, as a result of the contraction of consumption via the main payment methods: in-person card spending, cash withdrawals and e-commerce.1 However, in the days that followed, a significant rebound effect was observed: spending increased above normal levels, suggesting that some of the lost consumption was shifted to the following days.2 Overall, we estimate that over half of the spending not carried out on the day of the blackout was recovered in the next two days, reducing the net impact to a 15% drop in the spending expected for that Monday. This figure is the result of comparing the pattern of in-person card spending, cash withdrawals and online consumption on the day of the blackout and on the days that followed with the usual spending pattern on Mondays, and with the spending observed in areas where there were no power cuts.

From a macroeconomic perspective, the blackout had a limited effect on economic activity. According to our estimate, it caused a 15% drop in household consumption. Given that this component represents approximately 55% of GDP, this contraction would amount to an 8% reduction in the day’s GDP.3 At the quarterly level, the net effect on GDP in Q2 is estimated to be less than 0.1 percentage points, or less than 400 million euros.4 This figure suggests that, while the blackout had a significant economic cost in the short term, its impact on quarterly growth will be marginal.

The impact of the blackout on household consumption was different for each payment method on Monday 28 April and the two days thereafter. We estimate that the impact of the blackout on in-person card spending carried out by Spaniards in mainland Spain was 42%. This result is obtained by comparing the differing patterns of in-person spending on the 28th throughout the mainland compared to the rest of the country.5 Given that in- person spending usually follows a very similar pattern in the two territories, we use this differential as an approximation of the impact on in-person spending.6 In the two days after the blackout, in-person spending by card on the mainland was greater than the benchmark level indicated by the regions that did not experience the power cut.7 Taking this rebound into account, the impact of the blackout is reduced to an estimated drop of 18% in consumption on the mainland on the day of the blackout.

E-commerce spending by Spaniards fell compared to a normal day, both on the mainland and in the rest of the country, in the latter case probably because servers located on the peninsula were down. Taking into account the fall in both territories and apportioning each region according to its relative weight in GDP, the fall in online spending generated by the blackout is around 54% in Spain as a whole. The recovery in the two days after
the blackout places its net impact at a fall of 28%.

Cash withdrawals by Spaniards on the mainland fell 34% relative to the average level on Mondays in the months
of March and April 2025.8 For this payment method, we take the usual average level of cash withdrawals in March and April as a benchmark, since on the day of the blackout there was an increase in cash withdrawals outside the mainland, for precautionary reasons, which does not allow us to make a comparison between the two territories. In the two days following the blackout, cash withdrawals were also higher than usual, in part to normalise cash holdings. In this case, we assume that cash spending recovered in line with that observed in
the case of in-person spending, which places the net impact of the blackout at a fall of 15%.

Thus, taking into account the estimated decrease in spending with the three payment methods described, together with the fact that expenditure via transfers and direct debit charges did not suffer a material impact, we estimate that total spending throughout Spain on the day of the blackout was 34% below what could be expected for that day.9 In the two days after the blackout, the rebound in consumption compensated for almost half of the decline observed that day, bringing the net fall in total spending across Spain to 15% compared to the expected level.

Zooming in on the impact of the blackout by the hour, by economic sector and by autonomous community region reveals stark differences across the country. The pattern of hourly consumption on an ordinary day shows two peaks: around 12 noon and between 6pm and 7pm. The pattern of in-person card spending on the mainland was practically the same as the average for Mondays in March and April in the hours prior to the blackout. However, from 12:30pm onwards, there was a sharp drop in consumption via this payment method, and it did not recover in the remainder of the day.

The pattern of hourly consumption also allows us to understand the differential impact between the various sectors. The differences between sectors can be explained both by the magnitude of the initial fall and by the extent of the recovery observed after 6pm. In the case of essential goods, the impact was 34%, the lowest among all sectors, since the initial fall was more moderate and because in the evening spending stabilised at a higher level than in other sectors.10 On the other hand, spending on leisure and on accommodation and food services suffered the biggest decline, at 52%, partly because there was no recovery in the evening.

It should be noted that we cannot observe cash payments in business establishments, so the actual impact observed in the different sectors could have been lower.

The blackout caused a widespread drop in consumption in all autonomous communities, although its impact varied.11 As we can see in the last chart, on the 28th, consumption fell by 30-40% in most regions. The rebound in the following two days compensated for part of the drop in consumption observed on 28 April in all autonomous communities, albeit to differing degrees. Some regions in the central and north-eastern part of the country recovered more than 60% of their consumption in the following two days, while Galicia, for example, recovered just 30% of the consumption lost that day.