Skip to content

Search

Economic Bulletin – October 2024

Capa Economic Bulletin — October 2024

Click on the infographic for the short summary. Learn all about the economic projections in two minutes.

Sumário acessível_BE out 2024_EN

 

The Portuguese economy: 2024–26

Projections for the Portuguese economy: 2024–26

Economic activity in Portugal is expected to grow by 1.6% in 2024, 2.1% in 2025 and 2.2% in 2026 (Table I.1.1). Inflation is projected to drop to 2.6% in 2024, to stand at figures consistent with price stability over the following years. These projections result in the Portuguese economy continuing to converge towards European income levels and in an inflation differential of approximately zero vis-à-vis the euro area (Box 1 – External environment, financing conditions and policies). Compared with the June issue of the Economic Bulletin, growth was revised downward in 2024 (-0.4 p.p.) and 2025 (-0.2 p.p.).1 Inflation was revised slightly upwards in 2024 and downwards in 2025.

  1. Projections of Banco de Portugal for 2024-26 | Annual rate of change, in percentage (unless otherwise stated)

 

Weights 2023

EB October 2024

EB June 2024

2023

2024 (p)

2025 (p)

2026 (p)

2023

2024 (p)

2025 (p)

2026 (p)

Gross domestic product (GDP)

100.0

2.5

1.6

2.1

2.2

2.3

2.0

2.3

2.2

Private consumption

61.8

2.0

2.5

2.3

1.9

1.7

2.0

1.9

1.8

Public consumption

16.8

0.6

1.0

0.9

0.8

1.0

1.0

0.9

0.8

Gross fixed capital formation

20.1

3.6

0.8

5.4

5.1

2.5

3.3

6.1

5.0

Domestic demand

99.1

1.7

1.9

2.6

2.3

1.4

2.1

2.5

2.3

Exports

47.3

3.5

3.8

3.3

3.4

4.1

4.2

3.7

3.4

Imports

46.4

1.7

4.5

4.4

3.7

2.2

4.3

4.3

3.5

Employment (a)

 

1.0

1.1

0.6

0.9

0.9

1.0

0.8

0.8

Unemployment rate (b)

 

6.5

6.4

6.4

6.4

6.5

6.5

6.6

6.6

Current and capital account (% of GDP)

 

1.9

4.2

4.1

4.0

2.7

4.4

4.4

4.5

Trade balance (% of GDP)

 

1.2

2.5

2.1

2.1

1.2

2.4

1.8

2.1

Harmonised index of consumer prices

 

5.3

2.6

2.0

2.0

5.3

2.5

2.1

2.0

Excluding energy and food

 

5.4

2.6

2.3

2.3

5.4

2.3

2.3

2.3

GDP deflator

 

6.9

4.5

2.9

2.7

7.1

3.8

2.7

2.4

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. The cut-off date for the projections occured on September 30. For each aggregate, this table shows the projection corresponding to the most likely value, conditional on the set of assumptions. (a) According to the national accounts concept. (b) In percentage of the labour force.

The Portuguese economy is expected to maintain balanced growth over the projection horizon. External inflationary pressures are projected to remain subdued (Box 1 – External environment, financing conditions and policies). The momentum in disposable income will continue to reflect favourable labour market developments, with rising employment and wages, and the impact of fiscal measures. The gradual transition to lower interest rates and EU fund inflows are expected to support stronger investment growth. External demand for Portuguese goods and services will accelerate over the projection horizon, but developments in exports are projected to be constrained by the fading out of the post-pandemic recovery in services, in particular tourism-related services.

Recent developments in activity have been weaker than expected, while an acceleration is projected towards the end of the year. Quarter-on-quarter GDP growth declined in the second quarter, to 0.2%, after a dynamic start to the year (0.6%) (Chart I.1.1 – Panel A). Domestic demand accelerated, exports grew at a robust pace and there was a marked increase in imports, especially of services. In the third quarter, growth is expected to remain low according to available short-term indicators, including an activity indicator based on qualitative business surveys (Box 2 – An indicator for economic activity in Portugal based on business confidence surveys). Estimated developments reflect a slowdown in exports and private consumption. The projection of higher GDP growth in the fourth quarter and start of 2025 is partly due to a recovery in exports related to accelerating external demand and more buoyant tourism. Private consumption should also accelerate in line with improving household confidence and developments in disposable income.

After an increase in the second quarter, inflation declined to 2.3% in the third quarter and is expected to stand at 2.4% at the end of the year (Chart I.1.1 – Panel B). The recent volatility in inflation was unexpected and largely reflected idiosyncratic effects on the prices of accommodation services. The rate of change in the HICP excluding energy and food and the volatile tourism-related components remained at around 2.5% between the beginning of the year and the month of August. Inflation is expected to fluctuate around 2% over the projection horizon.

Economic activity is sustained by private consumption and exports in 2024 and will accelerate in 2025–26, reflecting buoyant investment. Exports of goods and services (net of import content) will continue to make an important contribution to growth over the projection horizon, although lower than in 2023 (Chart I.1.2). Investment is projected to increase significantly in 2025–26. The contribution of private consumption over the period is consistent with a stabilisation of its share in GDP, in real terms, and an increase in the saving rate.

  1. Quarterly GDP and inflation projections

Panel A – GDP – Quarter-on-quarter rate of change | Percentage

Panel B – HICP – Year-on-year rate of change | Percentage

Sources: Banco de Portugal and Statistics Portugal. | Note: The dashed lines correspond to the projected values in the EB of June and October 2024.

Private consumption is expected to grow moderately compared to the momentum of real income in 2024 – reflecting factors conducive to an increase in savings – and more in line in 2025–26. Private consumption is projected to grow by 2.5% in 2024, 2.3% in 2025 and 1.9% in 2026 (Chart I.1.3). Real disposable income is expected to grow by 6.6% in 2024, a historically high rate resulting from the favourable labour market – in terms of employment and real wages –, the increase in pensions and other transfers and the impact of the personal income tax reduction. The saving rate will increase to 11.5% in 2024 (8.0% in 2023), a historically high level in a non-recession context (Chart I.1.3). In the following years, real disposable income growth is projected to be more contained (1.9% on average), reflecting a deceleration in the wage bill and the fading effects of the fiscal measures included in the projection. This is advisable to ensure public finances remain balanced, having benefited in recent years from high Social Security surpluses. The household saving rate is expected to stabilise at close to 11%.

  1. GDP rate of change and contributions of components (net of import content) | Percentage and percentage points

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. For information on the methodology for calculating import-content net contributions, see Cardoso and Rua (2021), “Unveiling the real contribution of final demand to GDP growth”, Banco de Portugal Economic Studies, Volume VII, no. 3.

  1. Private consumption, real disposable income and savings rate | Chained data in volume, in millions of euros and as a percentage of disposable income

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. The gray bars mark the years when economic activity contracted.

The rise in the household saving rate in 2024 reflects the usual smoothing of consumer spending in response to an increase in income, but is compounded by the regime of positive interest rates, which contrasts with the figures close to zero that characterised the pre-pandemic decade. Higher interest rates increase the opportunity cost of consumption by making the return on financial investments and the repayment of loans more attractive. They also weigh on consumer spending of indebted households and spending using credit. Savings have also increased in other European countries (Box 1). Subdued consumption growth in 2024 can also be explained by distributional effects on disposable income, given that its momentum results in a considerable part from components that benefit groups of households with a lower marginal propensity to consume. In particular, a relevant contribution is expected from other income (excluding wages and transfers) – comprising mixed income, rents and capital income – as well as from the reduction in direct taxes, the impact of which is more marked for higher-income households.2 In addition, the recent surge in inflation has reduced the purchasing power of households’ financial assets, which may encourage savings to restore the real value of this wealth. Higher savings may also reflect increased household caution in consumer spending after the extreme shocks of recent years, including the pandemic and the significant rise in the cost of living. A deterioration in public savings is also expected to contribute to the precautionary strengthening of household savings. The maintenance of a high saving rate in 2025–26 is supported by the upward trend in savings intentions over the next 12 months reported in the European Commission’s Consumer Survey (Chart I.14). This trend is evident for households in the highest income quartiles, which account for the largest share of savings. However, uncertainty about the persistence of the factors listed above entails risks of a higher-than-projected increase in consumption.

  1. Savings intentions in the next 12 months – by income quartile | Balance

Source: EC (Consumer Survey). | Note: Balance of extreme responses to the following question put to consumers: In the next 12 months, do you think you will be able to save some money? (++) Very likely, (+) Likely, (-) Not likely, (--) Not at all likely.

Investment will slow down this year, but increased momentum is expected in 2025–26, in light of easing financial conditions, a better global outlook and the stimulus provided by EU funds (Table I.1.1). Growth in gross fixed capital formation (GFCF) in 2024 is expected to be low (0.8%) and concentrated in the public sector. Corporate investment and household investment in housing have been affected by tightening financial conditions. Nevertheless, developments in the ratio of GFCF to GDP in recent years compare favourably with the euro area (Chart I.1.5). The acceleration in investment in 2025–26, to an average growth rate of 5.2%, reflects continued robust growth in the public component and the recovery of GFCF in the corporate sector, where there continues to be a need to increase and modernise the capital stock, particularly in areas such as digitalisation, adoption of new technologies and energy transition. Housing investment is also projected to recover gradually, supported by declining interest rates, disposable income growth and buoyant migration flows, but constrained by the availability of labour in the construction sector.

  1. Share of GFCF in GDP in Portugal and the euro area | Percentage

Sources: Banco de Portugal, Statistics Portugal, Eurostat and ECB | Notes: (p) – projected. Shares based on chained volume data. The projections for the euro area's total GFCF coincide with those of the ECB's projection exercise released on September 12 (see “ECB staff macroeconomic projections for the euro area”, September 2024). The ECB does not release projections for business GFCF in the euro area.

Average export growth in 2024–26 is expected to be similar to 2023 (3.5%), driven by a slowdown in the services component and an acceleration in goods (Chart I.1.6). This recomposition reflects, on the one hand, the normalisation of global consumption patterns as the post-pandemic effects on services demand dissipate and, on the other hand, the acceleration in external demand, mainly impacting goods. The average growth projected for goods exports is 3.1% in 2024–26, following a decline of 1.5% in 2023. Tourism, albeit slowing down, will continue to outgrow total exports (5% on average in 2024–26). Other services are also projected to decelerate from 2023, growing at an average pace of 3.2% over the projection horizon. The non-tourism services component grew the most in cumulative terms compared to 2019, reflecting the momentum in IT, management consultancy, research and development, architecture and engineering activities, among others (Chart I.1.6).3 Projected developments in exports of goods and services are consistent with continued market share gains. In the first half of 2024, goods exporters continued to gain market share in EU markets in nominal terms, with gains extending to most product groups.

The change in the composition of exports, with a lower contribution from services, and the acceleration in GFCF imply growth in activity more dependent on imported goods, reflected in an acceleration of imports compared to 2023 (Table I.1.1).

  1. Exports of goods and services | Index 2019 = 100

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. The indicator of external demand for the Portuguese economy consists of an average of imports of trading partners, weighted by their share in Portuguese exports.

The economy’s net lending is expected to increase to an average of 4.1% of GDP in 2024–26. The goods and services balance is projected to increase from 1.2% of GDP in 2023 to an average of 2.2% of GDP in 2024–26 This improvement reflects a lower deficit in trade in goods and a higher surplus in services (Chart I.1.7). In the case of goods, the higher growth of imports in terms of volume relative to exports is more than offset by terms of trade gains, which are more significant in 2024. In the first half of the year, gains reflected developments in the relative price of non-energy goods – unlike in the previous year when they were also related to the fall in oil prices – and were broad-based across product groups. These terms of trade gains are part of a long trend and indicate structural changes in the export sector, with a redirection towards higher value-added products.4 The projected increase in the services surplus is also expected to continue a trend: between 2019 and 2023, the balance of services increased by 2.2 p.p. of GDP, with contributions of 1 p.p. from tourism and 1.2 p.p. from other services (Chart I.1.7). The income and capital account balance is projected to improve due to increased transfers of EU funds from 1.2% of GDP in 2023 to an average of 2.5% of GDP in 2024–26. The projections for the external balance imply that the external indebtedness ratio will continue to decline.

  1. Current and capital account | As a percentage of GDP

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. The breakdown of the balance of services between travel and tourism and other services is not available over the projection horizon, since imports of services are projected in aggregate terms.

In the labour market, employment is expected to continue to grow and the unemployment rate to remain low in the context of an increase in the participation rate and significant immigration flows. Employment is projected to increase by 1.1% in 2024, 0.6% in 2025 and 0.9% in 2026 (Table I.1.1) The labour force – which represents the potential for job creation – is expected to increase further. Immigration, resulting in an increase in the participation rate, has been a key part of employment growth. Between 2019 and 2023, the number of employees registered with Social Security increased by 14.4%, with foreign individuals contributing nearly 10 p.p.5 These immigrants have fulfilled labour needs in most sectors, in particular agriculture and fishing, industry, construction, trade, accommodation and food services and administrative services (Chart I.1.8).

Wages per worker in the total economy are projected to slow down over the projection period, in line with declining inflation expectations. Real wages are expected to increase by 4.6% in 2024 (after 3.5% in 2023), with gains being more moderate and aligned with productivity growth in 2025 and 2026 (Chart I.1.9). Average growth in productivity per worker over the projection horizon (1.1%) is expected to exceed the average observed over the period 2015–19 (0.5%) or a longer period (0.8% in 2000–19). These developments are projected to reflect the impact of changes in the Portuguese economy, such as improvements in the skills of the population, job creation in high and medium-tech industries and knowledge-intensive services and increased digitalisation.

  1. Employees registered with Social Security – change between 2019 and 2023 by sector and nationality | Thousands of individuals

Sources: Social Security microdata (Banco de Portugal calculations) and DGAEP. | Notes: Employees of working age (16-74), living in Portugal and earning at least the equivalent of one day's pay in the company each month were considered. In “Memorandum item”, Caixa Geral de Aposentações subscribers are included in the change of the number of workers in public administration, education and health sectors. Taking these individuals into account (for whom a breakdown by nationality is not available), the change in employees in the sector is around half that obtained when considering only individuals registered with Social Security (68,000 vs. 124,000).

  1. Nominal and real wages and productivity per worker - rate of change | Percentage

Sources: Banco de Portugal and Statistics Portugal. | Notes: (p) – projected. The private consumption deflator was used to calculate real wages.

Inflation is projected to fall to 2.6% in 2024 and stabilise at 2% in 2025–26 against a background of decelerating wage costs and moderate external pressures. In 2024 the decrease is expected to reflect a lower contribution of all main components except for energy (Chart I.1.10 – Panel A). Services prices are projected to continue to grow more than goods over the projection horizon (Chart I.1.10 – Panel B). Developments in services prices are an indicator of domestic inflationary pressures and persistent inflation. While the prices of most goods are determined in international markets and fluctuate with global supply and demand conditions, a large share of services are less tradable and are labour intensive; as a result, their prices are more dependent on the degree of tightness in the labour market and on wage developments. A model that breaks down inflation in Portugal into the contribution of its determinants points to the preponderance of external price shocks in the rise and subsequent decline in inflation in the post-pandemic period but suggests a greater influence of labour market pressures in the more recent period (Box 3 – Drivers of inflation according to the Bernanke-Blanchard model). In the euro area, inflationary pressures have also been decreasing, with inflation closer to the ECB’s price stability objective (Box 4 – The direction of inflationary pressures in the euro area).

The GDP deflator is projected to grow by 4.5% in 2024 (6.9% in 2023) and slow down to 2.9% in 2025 and 2.7% in 2026 (Table I.1.1). The slowdown in 2025-26 is a reflection of lower domestic inflationary pressures stemming from unit labour costs.

  1. HICP and components

Panel A – HICP rate of change and contribution of components | Percentage and pp

Panel B – HICP rate of change | Percentage

Sources: Banco de Portugal and Statistics Portugal. | Note: (p) – projected.

Risks surrounding the projection for growth and inflation are balanced. For activity, downside risks stemming from international geopolitical tensions remain, with the possibility of a worsening of the conflicts in Ukraine, related to Russia’s invasion, and in the Middle East and also reflecting the outcome of the US presidential election, trade tensions with China and proposed policies to increase protectionism. Growth in GFCF may be lower in a scenario of difficulties in fulfilling the RRP targets within the deadlines set. Conversely, there is an upside risk stemming from higher private consumption growth, in response to the projected increase in household income. For inflation, there are downside risks arising from the possibility of the lagged effects of monetary policy being more marked in the short term. These risks are counterbalanced by upside risks associated with shocks to international commodity prices and global supply chains in a context of geopolitical tensions, as well as the momentum in wages and its pass-through to prices.

The resilience of the economy to recent shocks reflects progress in reducing macroeconomic imbalances and other structural weaknesses. The decline in private, public and external debt ratios implied lower vulnerability to the interest rate shock. More exposed household groups have also benefited from measures to mitigate the impact of rising interest rates on mortgages. The labour market has been robust and flexible, with the impact of population ageing being offset by immigration and an increase in the participation rate. There is also an improvement in the skills of the labour force – for the population aged 39 or under, the share of individuals with tertiary education is already close to the European average – which supports the development of higher value-added activities and is reflected in increased productivity and competitiveness gains in external markets.

The Portuguese economy will face important challenges in the near future, linked to technological transformation, the impact of geopolitical changes and climate transition management (Special issue: Climate scenarios for the Portuguese economy). One instrument to address these challenges is the RRP, which requires a determined effort to accelerate the implementation of projects and related reforms (Policy insights: The implementation of the RRP in Portugal). As for fiscal policy, its expansionary stance in all years of the projection horizon, in a context where GDP is above potential, will generate the need for later adjustment in a less favourable phase of the economic cycle. Consequently, the pace of sustained reduction in the public debt ratio should not slow down, as it is a key element for macroeconomic stability and the growth of present and future generations.

  1. External environment, financing conditions and policies

The global economy is expected to continue growing at a moderate pace until 2026. According to the assumptions of the ECB’s September projection exercise, global GDP is expected to grow at an annual rate of 3.1% between 2024 and 2026, with slight upward revisions compared to the June 2024 scenario (Table B1.1). The stability of overall growth reflects a slight recovery in advanced economies, while in emerging market economies GDP is expected to decelerate over the projection horizon.

The ECB’s September projections point to an acceleration of economic activity in the euro area, but to a weaker pace than envisaged in the June exercise. The expected recovery in private consumption in the second quarter of 2024 did not materialise, with the outlook for the second half of the year being revised downwards amid weaker signals from qualitative indicators. The PMI (Purchasing Managers' Index) for the euro area has fallen in recent months, signalling stagnation in GDP in the third quarter, with weak activity in France and Germany (Chart B1.1). The September projection maintains a recovery in euro area economic activity driven by growth in private consumption. On annual average terms, euro area GDP will grow by 0.8% in 2024, 1.3% in 2025 and 1.5% in 2026. These developments are supported by the recovery in real disposable income, an increase in external demand and an easing in financial conditions through lower interest rates (Table B1.1).

  1. Composite PMI – Output | Diffusion index

Source: S&P Global.

The saving rate plays an important role in the projection for the euro area, with an expected decline from its current high levels as consumption patterns return to normal. Following the peaks reached in the pandemic, the saving rate decreased throughout 2021 and 2022, but has been increasing since the end of 2022. In the first quarter of 2024, the saving rate reached 15.4%, above the pre-pandemic average (12.7%). Consumer confidence has also been improving, in the context of a robust labour market, although still standing below its historical average and pre-pandemic levels. This suggests that the recent increase in the saving rate and in saving intentions is not only justified by precautionary reasons. The rise in monetary policy interest rates in the euro area between July 2022 and September 2023 was passed through to interest rates on time deposits. This increase resulted in an incentive to save, with overnight deposits being replaced by time deposits since 2023 (Chart B1.2). Over the projection horizon, the normalisation of monetary policy and the recovery in consumer confidence are expected to contribute to a gradual decline in the saving rate.

  1. Eurosystem staff projection assumptions

 
 

EB october 2024

Revisions from EB June 2024

 

2023

2024

2025

2026

2023

2024

2025

2026

International environment

 
 
 
 

 

 
 
 
 

World GDP

yoy

3.1

3.1

3.1

3.1

0.0

0.1

0.1

0.1

Euro area GDP

yoy

0.5

0.8

1.3

1.5

-0.1

-0.1

-0.1

-0.1

World trade

yoy

0.5

2.3

3.3

3.3

0.1

0.2

0.0

0.1

External demand

yoy

-0.1

1.1

2.8

3.3

0.4

0.0

-0.4

0.1

International prices

 
 
 
 

 

 
 
 
 

Oil prices

aav

77.5

76.5

69.5

66.9

0.0

-1.2

-2.9

-2.3

Gas prices (MWh)

aav

40.6

34.2

41.1

35.4

0.0

3.4

5.7

5.5

Non-oil commodity prices

yoy

-14.5

6.7

0.6

2.5

0.0

-5.0

-3.5

1.6

Competitors' import prices

yoy

-1.6

0.4

2.0

2.3

0.0

-0.4

-0.6

0.0

Monetary and financial conditions

 
 
 
 

 

 
 
 
 

Short-term interest rate (3-month EURIBOR)

%

3.4

3.6

2.5

2.2

0.0

0.0

-0.3

-0.3

Implicit interest rate in public debt

%

2.0

2.2

2.3

2.3

-0.1

-0.2

-0.2

-0.3

Effective exchange rate index

yoy

4.9

2.2

0.5

0.0

0.0

0.4

0.4

0.0

Euro-dollar exchange rate

aav

1.08

1.09

1.10

1.10

0.0

0.8

1.7

1.7

Sources: Banco de Portugal and Eurosystem (Banco de Portugal calculations). | Notes: yoy – year-on-year rate of change, % – in percentage, aav – annual average value, MWh – megawatt-hour. Technical and external environment assumptions, as well as projections for euro area GDP and inflation, coincide with those in the ECB projection exercise released on September 12 (see "Eurosystem staff macroeconomic projections for the euro area", september 2024), which include information up to August 16. International prices are in euros. The technical assumptions for the price of oil, gas and non-energy commodities is based on futures markets. The import price of competitors corresponds to a weighted average of the export deflators of the countries from which Portugal imports, weighted by their share on total Portuguese imports (for more information, see "Trade consistency in the context of the Eurosystem projection exercises: an overview", ECB Occasional Paper 108, March 2010). The evolution of the 3-month EURIBOR is based on expectations implied in futures contracts. The implicit interest rate on public debt is computed as the ratio of interest expenditure for the year to the simple average of the stock of debt at the end of the same year and at the end of the preceding year. An increase in the exchange rate corresponds to an appreciation of the euro. The effective exchange rate of the euro is computed against 41 trading partner countries. The technical assumption for bilateral exchange rates assumes that the average levels observed in the 10 business days prior to the cut-off date are maintained over the projection horizon.

  1. Household deposits | Flows and interest rates | EUR million and in percentage

Panel A – Euro area

Panel B – Portugal

Source: ECB.

Global trade growth is expected to recover over the projection horizon. After growing by 0.5% in 2023, global trade in goods and services is expected to grow by 2.3% in 2024 and slightly above the pace of global activity growth in 2025 and 2026 (3.3%) (Table B1.1). Compared to the June projection exercise, global trade growth was revised upwards in 2024 (0.2 percentage points). This revision is likely related to a frontloading of imports in advanced economies during the second quarter, driven by fears of supply bottlenecks and intensified trade and geopolitical tensions in the second half of the year. The projected recovery in external demand for the Portuguese economy is expected to be slower than that of global trade, with growth of 1.1% in 2024, 2.8% in 2025 and 3.3% in 2026, reflecting more subdued intra-EU trade.

Euro area inflation is expected to decline and converge towards the monetary policy objective by the end of 2025, which is in line with the June projection. Euro area inflation is expected to increase in the last quarter of 2024, driven by base effects in the energy component, followed by a gradual decline until the end of 2026. According to the ECB’s projections, euro area inflation falls from 5.4% in 2023 to 2.5%, 2.2% and 1.9% in 2024, 2025 and 2026 respectively. The inflation measure excluding food and energy is expected to stand at 2.9%, 2.3% and 2.0% in 2024, 2025 and 2026 respectively.

Expectations based on futures contracts point to a larger-than-expected reduction in short-term interest rates compared to the June exercise. On annual average terms, the three-month EURIBOR is expected to fall to 3.6% in 2024, 2.5% in 2025 and 2.2% in 2026 (Table B1.1). The downward revision in the short-term interest rate for 2025 and 2026 reflects expectations of further policy rate cuts by the ECB. The implicit interest rate on Portuguese debt will increase to 2.2% in 2024 and stabilise at 2.3% in 2025–26. These developments reflect the replacement of debt issued in the past at lower interest rates than current issuances.

 
  1. An indicator for economic activity in Portugal based on business confidence surveys

Monitoring a broad and timely set of activity indicators is essential for economic analysis and short-term forecasting. The new composite indicator for the GDP quarter-on-quarter rate of change in Portugal is based on opinion surveys released by the European Commission (EC). This indicator is qualitatively similar to the composite output Purchasing Managers’ Index (PMI), which is widely used for short-term GDP forecasting in several countries, but which is not available for Portugal.6

The activity indicator presented in this Box uses responses from firms to the EC’s business surveys. In May 2023 the sample covered 5,206 firms: 1,801 in industry, 1,808 in services, 728 in trade, and 869 in construction. The EC reports results usually in balances, but it also discloses detailed information by type of response (positive, neutral or negative). To calculate this indicator, the result for each question is reconstructed as a diffusion indicator – similar to the PMI and with a similar interpretation – indicating expansion (above 50), stability (50) or contraction (less than 50).7 Sectoral questions are aggregated based on the weight of each sector of activity in the total economy.8 The combination of questions that proved to be most correlated with GDP in the same quarter aggregates output/business expectations for the following three months from industry, services and retail trade questionnaires. In addition to being timely, as it is available on the last day of the reference quarter, this indicator has the advantage of hardly being revised.9

Chart B2.1 shows the comparison between the activity indicator based on confidence surveys and the GDP quarter-on-quarter rate of change. The indicator shows a high correlation with activity developments (70% in 2004 Q1–2019 Q4 and 66% in 2022 Q2–2024 Q2).

  1. Activity indicator based on confidence surveys and GDP | Diffusion index and quarter-on-quarter rate of change, in percentage

Sources: European Commission and Statistics Portugal (Banco de Portugal calculations). | Note: The scale on the left has been truncated to improve the readability of the graph, with the omitted values indicated in text on the graph.

Chart B2.2 – Panel A assesses the ability of the activity indicator based on confidence surveys to anticipate the sign of the GDP quarter-on-quarter rate of change for several sub-periods. The pandemic period (2020 Q1–2022 Q1) was highly unstable in economic activity and, therefore, it was not included in this analysis. Given the monthly frequency of confidence indicators, the forecast for GDP developments based on the indicator may occur at different points in time in the quarter, corresponding to different degrees of information availability. The outcome of these various scenarios is also displayed in Chart B2.2, which shows a success rate in forecasting the sign of the GDP quarter-on-quarter change close to 90% in the most recent period. This percentage is slightly lower when only one month information is available for confidence indicators, but it stabilises from the release of the second month of information onwards, which occurs around 30 days before the end of the reference quarter for GDP. In terms of its ability to predict an acceleration/slowdown in GDP, the indicator shows a success rate of 89% in the most recent period, when two or more months of the confidence indicators are available. When the availability of confidence indicators is more limited, the success rate is 67%.

  1. Success rate of the activity indicator based on confidence surveys in predicting the sign and an increase/decrease in the quarter-on-quarter rate of change of GDP | Percentage

Panel A – Forecast of the sign of the rate of change

Panel B – Forecast of an increase/decrease in the rate of change

Sources: European Commission and Statistics Portugal (Banco de Portugal calculations). | Notes: The success rate in predicting the sign of the quarter-on-quarter rate of change is calculated as the ratio between the number of quarters in which the sign of the GDP quarter-on-quarter rate of change coincided with the sign of the activity indicator based on confidence surveys minus 50 and the total number of quarters under analysis. The success rate in predicting an increase/decrease in the quarter-on-quarter rate of change is given by the ratio between the number of quarters in which the quarter-on-quarter rate of change in GDP increased (decreased) and the indicator increased (decreased) and the total number of quarters under analysis.

The analysis suggests that this indicator provides useful information on activity developments in the current quarter, but a joint analysis of all available short-term indicators and models cannot be dismissed. A short-term model for the GDP quarter-on-quarter rate of change suggests that each point of increase in the activity indicator corresponds to an increase in the GDP quarter-on-quarter rate of change of 0.13 percentage points, a similar result to that obtained for the euro area based on the PMI.10 On the basis of information of confidence indicators from July to September, this indicator points to a GDP quarter-on-quarter rate of change in 2024 Q3 close to that recorded in the previous quarter (Chart B2.1).

 
  1. Drivers of inflation according to the Bernanke-Blanchard model

What are the primary drivers of inflation behaviour in Portugal in recent times? Using the model recently proposed by Bernanke and Blanchard (2023),11 the dynamics of prices, wages and short and long-run inflation expectations are jointly explained, conditional on exogenous shocks from energy prices, food prices, productivity, supply-side constraints (supply chain constraints) and the degree of tightness in the labour market. Based on the structure of this model, it is possible to quantify the relative role of the various shocks in determining inflation and their transitory or permanent nature.

The model for Portugal has been estimated for 1999 Q1–2024 Q2. Inflation, measured by the rate of change in the HICP, is a function of past inflation, relative energy and food inflation vis-à-vis wages, long-term productivity growth, an indicator of supply chain constraints and nominal wage growth per employee. Labour market conditions and inflation expectations act in the model through the wage growth equation, represented by an expectations-augmented Phillips curve. The role of monetary policy is captured indirectly through its influence on the labour market and long-term inflation expectations. The labour market variable also captures demand-side pressures and the effects of fiscal policy.

The estimated model and the empirical responses to exogenous shocks in the post-pandemic period allow an analysis of the inflation drivers in Portugal in 2021 Q1–2024 Q2 (Chart B3.1).

The increase in inflation until 2022 Q4 was primarily the result of a succession of adverse energy and food price shocks and supply chain constraints. The reopening of economies and the resumption of global activity led to an increased demand for energy commodities and, therefore, to a rise in prices. Constraints in global supply chains reflected a sluggish supply response to increasing demand for goods, coupled with the materialisation of expenditure postponed during the pandemic lockdowns.

With Russia’s invasion of Ukraine in 2022 Q1, the contribution of previously observed effects was amplified until the end of 2022, with a particular focus on energy prices until 2022 Q4 and food prices until 2023 Q2. According to the model, despite the more persistent contribution of food prices, the observed fall in inflation since 2022 Q4 is explained by the reversal of relative price shocks and the easing of supply chain constraints.12

The labour market situation has also influenced rising inflationary pressures, especially in the most recent period. The labour market in Portugal, as in other developed economies, shows a high degree of tightness – reflecting increased aggregate demand and labour supply already at historically high levels – which contributed to the marked rise in nominal wages over the past three years. Real losses in 2022, the year of a sharp rise in inflation, started to be reversed in 2023, also explaining the lagged effect on inflation.

  1. Drivers of post-pandemic inflation | Year-on-year change rate in percentage and contributions in percentage points

Sources: European Commission, Consensus Economics, Eurostat, Statistics Portugal, MTSSS, NYFED (Banco de Portugal calculations). | Notes: The chart shows the decomposition of the year-on-year rate of change in the HICP between 2021Q1 and 2024Q2, based on the estimated model and the implicit impulse response functions. The bars represent the contribution of each exogenous variable and the residuals to observed inflation, net of the contributions of structural conditions. The structural conditions are given by the exogenous variables at their equilibrium or long-term value. The contribution of these bars reflects what inflation would be if the labour market and supply chain constraints variables were in equilibrium (indices at zero), productivity growth was at the long-term average and the growth in the relative prices of energy and food goods was zero. Inflation is measured by the rate of change in the HICP, energy and food prices correspond to the respective subcategories of the HICP, the measure of wages used is basic wages per worker declared to social security, productivity is calculated using GVA per worker, the supply chain constraints index is obtained from the NYFED Global Supply Chain Pressure Index series and the labour market variable corresponds to the Insufficient Labour Force indicator, explained in more detail in the notes to Chart B3.2.

The degree of tightness in the labour market is assessed using the relationship between an indicator of insufficient labour force and the unemployment rate. It is a proxy to the Beveridge curve, which links the unemployment rate to the ratio of job vacancies to the number of unemployed. Chart B3.2 illustrates this relationship and reflects that although the unemployment rate was relatively stable over the past three years, the share of entrepreneurs reporting difficulties in recruiting has risen to historically high levels. In 2024 Q2, the degree of labour market tightness was higher than in the pre-pandemic period despite favourable developments compared to 2022.

The labour market condition contributes to wage pressures with a lag and, therefore, to inflationary pressures. In the estimated model, the contribution of the labour market situation to inflation is the most persistent. Although initially small compared with the contribution of relative price shocks, the tightness of the labour market is the primary driver of inflation in the most recent period, offsetting the negative contribution from energy and food prices (Chart B3.1). It is worth noting, however, that in this period, the inflation rate is close to 2%.

Comparing the results for Portugal with those presented in Bernanke and Blanchard (2024) for the euro area and other Member States until 2023 Q2, it follows that, qualitatively, the factors behind price dynamics are cross-cutting in these economies. Shocks to relative energy and food prices and their subsequent reversal, amid anchored inflation expectations, drove inflation dynamics in recent years.

  1. Labour market tightness and unemployment rate

Sources: European Commission and Statistics Portugal (Banco de Portugal calculations). | Notes: Quarterly data between 1999Q1 and 2024Q2 was used; labour market tightness indicator corresponds to the Insufficient Labour Force (ILF) index, constructed from the percentage of entrepreneurs reporting insufficient labour force as a factor limiting production in the European Commission's surveys, weighting the results for industry, services and construction by the respective weight of each sector's employment in the total economy and then standardised based on the mean and standard deviation; in order to obtain observations since 1999Q1, the indicator was retropolated using the data for the available sectors and adjusting its weighting; the non-linear relationship between the ILF index and the unemployment rate, illustrated by the yellow trend, excludes the period 2010 Q1-2015 Q4; the intersection between the axes is given by the non-linear relationship, implying an unemployment rate of 6.9 per cent.

 
  1. The direction of inflationary pressures in the euro area

The conduct of the ECB’s monetary policy considers a wide range of information. An important part is the analysis of developments in total inflation, complemented by the monitoring of measures of underlying inflation. These measures are mainly aimed at capturing the underlying trend in the prices of goods and services. This Box presents a new tool designed to ascertain the direction of underlying inflation in the euro area. It preserves the properties of a compass, where its needle points towards the mean direction of inflationary pressures, i.e. whether price pressures are accelerating or decelerating.13

The compass is read in a counterclockwise direction, with the East (0º) and West (180º) directions corresponding to a situation where the underlying inflation is 2%. As the needle moves along the first quadrant, price pressures build up, with inflation rising until it reaches a local peak, represented by the 90º angle (North). Transitioning into the second quadrant, ranging from 90º to 180º, inflation remains above the ECB’s price stability target. However, prices are starting to decelerate as this quadrant progresses, until inflation reaches the target in the West direction. In the third quadrant, ranging from 180º to 270º, inflation is below the target. Here, price pressures continue to decelerate as this quadrant progresses, until inflation reaches a local minimum in the South direction (270º). Finally, in the fourth quadrant, ranging from 270° to 360°, prices accelerate. As this quadrant progresses, inflation increases until it reaches the target again (East direction).

The direction is obtained from all granular information for calculating the HICP in the euro area, thus indicating current inflationary pressures. A confidence interval for the mean direction is also shown, in the form of an arc and with a magnitude reflecting the dispersion of the basic sub-items of inflation.

This tool has corroborated a decrease in inflationary pressures throughout 2024. For example, at the beginning of the year, the compass needle was in the second quadrant, signalling that inflationary pressures were decreasing (Chart B4.1). At the beginning of the second quarter, this indication was further corroborated. The latest data suggest that inflationary pressures continued to decline in August, with the needle moving towards the West direction. Notwithstanding the heightened uncertainty compared with previous months given the larger amplitude of the arc, the confidence interval encompasses the West direction, suggesting that inflationary pressures would be consistent with the price stability target.

This uncertainty reinforces the importance of analysing a wide range of indicators to inform the ECB’s policy decisions in the current context.

  1. Inflation compasses

Panel A – Jan 2024

Panel B – Apr 2024

Panel C – Aug 2024

Source: Banco de Portugal. | Notes: The directions of 0° and 180° correspond to an inflation rate of 2%, and the directions of 90° and 270° refer to a local maximum and minimum, respectively. Compasses are calculated based on the ECOICOP-5 classification of the HICP sub-indices. The interval is calculated for a 90% confidence level.

 
 

Special issue

Climate scenarios for the Portuguese economy1

This Special issue presents selected results of the assessment exercise on the macroeconomic impacts of global warming on the Portuguese economy, carried out under the National Roadmap for Adaptation 2100 (RNA 2100). The purpose of RNA 2100 is to assess the Portuguese territory’s vulnerability to climate change in the 21st century, promoted by the Portuguese Environment Agency, partnered with the Banco de Portugal, the Faculty of Sciences of the University of Lisbon, the Directorate-General for Territory, IPMA (the Portuguese Institute for Sea and Atmosphere) and the Norwegian Directorate for Civil Protection.2

The core element of the Roadmap is a set of climate projections for mainland Portugal up to 2100, consistent with three of the scenarios developed by the Intergovernmental Panel on Climate Change (IPCC). The projections provide climate data at a high geographical resolution to support the design of public climate change mitigation and adaptation policies, as well as to contribute to public information and awareness.

The next section provides evidence on the geophysics of global warming and the reasons why this is also an economic problem, while describing appropriate economic policies to mitigate its effects. The global climate scenarios used in the Roadmap are then characterised. By applying these scenarios to the Portuguese territory, it is possible to describe in very general terms Portuguese economic developments in each scenario, using a general equilibrium integrated assessment model (IAM).3 This Special issue includes a box on climate projections for mainland Portugal up to 2100 for the different IPCC scenarios used in the Roadmap.

The global warming problem

The global warming issue has been known since at least the end of the 19th century, when Swedish chemist Svante Arrhenius described the effect that some gases had on the absorption of radiation emitted by the Earth’s surface and on the re-emission back to it, producing a greenhouse effect. Greenhouse gases (GHGs) act as a sort of blanket enveloping the Earth, turning the atmosphere opaque to outgoing radiation. This radiation is emitted at wavelengths generally different from those of sunlight, which pass through the atmosphere if no clouds are present. When GHGs become more concentrated, the effect intensifies, energy retention increases and Earth’s surface temperature rises. This happens until a balance is restored between incoming energy from the Sun (part of it – around 70% – is immediately reflected back into space) and outgoing energy from the Earth.

Of all GHGs, water vapour absorbs and re-emits the most radiation from Earth’s surface. However, its presence in the atmosphere is mainly due to endogenous factors to the climate system itself and not directly to human action. This is not the case for the second most abundant GHG, carbon dioxide, which accounts for ¾ of GHG emissions excluding water vapour. Part of the concentration of this gas in the atmosphere is due to natural factors, but a major component comes from the burning of carbon-rich fossil fuels. Other gases contribute to this phenomenon, such as methane, but do not seem to be as relevant, as they last in the atmosphere for a relatively short period.

Once emitted, carbon dioxide spreads across the atmosphere and remains there for a long time. The time it takes for half of the gas emitted to be removed from the atmosphere through natural processes exceeds one century. The fact that carbon dioxide is a powerful GHG, combined with it remaining in the atmosphere for a long time and being emitted by burning fossil fuels for energy, makes it a major contributor to human-driven global warming.

Global warming is a very complex phenomenon, conditioned not only by the concentration of GHGs in the atmosphere, but also by seasonal fluctuations, geographical variation, human- or nature-induced land morphology changes, and even extraterrestrial phenomena, including the intensity of the Sun’s radiation. Moreover, the fact that the average global Earth temperature is rising does not mean that this increase is evenly spread across the globe, and there may even be localised drops in the average temperature. Global warming also affects precipitation, wind intensity and the frequency of extreme events, such as droughts, floods and storms, in ways that are difficult to predict and quantify. This is why the term “climate change” is often used, rather than “global warming”.

Economic activity is fundamentally linked to the use of energy and, where this energy is produced by burning fossil fuels, the impact of economic activity on the climate becomes apparent. Changes in economic activity will lead to changes in GHG concentration, global average temperature and climate. Does the opposite effect – a change in climate leading to changes in economic activity – also happen? The answer is yes, and for a myriad of reasons. First, economic activity, as a human phenomenon of optimising existing resources to maximise economic benefits, is adapted to the climatic conditions of the surrounding environment. For instance, in cold climates, homes have powerful heating systems and are built with materials suitable for cold temperatures. If the average annual temperature rises, such constructions will become less suitable; for instance, house-building materials could probably be replaced with less expensive ones, and heating systems could prove to be inefficient. The inadequacy resulting from the rise in average temperature results in economic costs that would not have been previously borne, which can be short- or long-lived. The argument applies to many other aspects of economic activity: transport and communication infrastructure, flood prevention barriers, waterways, public buildings, sports resorts, and so on.

There are other ways climate change impacts on an economy’s output. An obvious impact channel is agricultural productivity, given that land, crops and farm machinery are adapted to the baseline situation. Mortality also tends to increase in tandem with temperature in sufficiently warm climates, the opposite being true for sufficiently cold temperatures. Depending on the distribution of the population across regions, this may lead to aggregate economic damages.

Beyond these, there are other less obvious channels. Global warming could lead to a rise in temperature-related extreme weather events. The IPCC considers that it is “likely” that extreme temperature phenomena which would occur once in ten years in a climate without human influence – and which currently, with global warming of 1 °C, are already 2.8 times more frequent – would become 5.6 times more frequent in a 2 °C global warming scenario.4 This broadly linear link means that economic activities sensitive to heat waves will be more frequently affected, thereby increasing the volatility of their returns.5 This temperature increase is associated with a higher frequency of other types of extreme events and adverse phenomena, including floods, droughts and ocean acidification, among others.

From an economic point of view, what has just been described is an overall negative externality.

The economic literature on the fight against climate change distinguishes between two types of measures to mitigate its effects and provide higher levels of wellbeing: mitigation policies and adaptation policies. The first aim to encourage agents to reduce the intensity of the externality, in this case by reducing GHG emissions; the latter’s aim is to reduce the negative effects of the externality.

Mitigation policies

In essence, the first prescription of a mitigation economic policy is based on seminal work by Pigou (1920). Given that economic agents generating GHG emissions do not take into account the negative impacts of their activity on others, a market failure arises. The mitigation policy prescribed in that case is a tax on GHG emissions levied at source, equal to the total marginal damage not considered by the emitter. The aim is to align economic agents’ incentives to further reduce GHG emissions. This measure can also be considered an incentive for innovation in less carbon-intensive technologies.

The second mitigation policy prescription is inspired by the work by Coase (1960) on the allocation of intellectual property rights. Instead of implementing corrective taxes, this measure sets a certain amount of carbon emissions linked to a given amount of tradable carbon permits. Purchasing a permit in the market grants the right to emit one carbon unit. One example of this measure is the current EU Emissions Trading System (EU ETS), in force since 2005.6

Regulations are also a mitigation measure, such as restrictions on fossil energy production. Regulations requiring certain energy efficiency criteria in new construction are another example.

Adaptation policies

Adaptation measures tackling climate change do not mitigate the climate externality itself but rather its effects. As this is a form of public or private investment, adaptation absorbs resources that could be used elsewhere, whereby an optimal policy strategy will take this trade-off into account. Economies’ adaptive capacity hinges on factors such as the quality of institutions or the degree of financial development, which are linked to economic growth. As such, policies stimulating economic growth or the provision of public goods (e.g. in health and in the management of drinking water supply) will tend to increase countries’ adaptive capacity and mitigate the effects of the climate externality. The literature recognises that negative impacts are greater in poorer countries (Tol, 2024). These economies are located in warmer areas of the world and their sectors of activity tend to be more exposed to the effects of rising temperatures, such as agriculture, and such economies also tend to be less diversified.

An example of public adaptation policies is investment in coastal protection to counter the effect of rising average sea water levels. A large share of these investments is aimed at protecting existing physical capital and therefore do not correspond to productive investments per se. Finally, it should be noted that there is some substitutability between adaptation and mitigation policies. This suggests that, given that there are two instruments, mitigation incentives are reduced, so large adaptation investment needs may be a sign of ineffective mitigation policies in the fight against climate change.

Some unresolved challenges

The assessment of the economic impacts of different climate scenarios, such as the exercise presented in this Special issue, is at a more advanced stage than the one where the scenarios assume a deterioration in natural capital, understood as natural resources as a whole (soil, water, minerals, air, living organisms). In particular, a thorough, comprehensive assessment of the economic impact of biodiversity loss and ecosystem degradation is yet to be undertaken. One reason for that is the lack of geo-referenced data on economic activity units, which are not available in many jurisdictions, that would allow for a proper assessment of economic activities’ exposure to this type of risks.

There is also a problem in quantifying the economic value of biodiversity and the flow of services associated with ecosystems.7 The functioning of ecosystems is complex, and quantifying their economic value is difficult, as there is no general way of pricing natural capital (IPBES, 2019). What is the price on saving an animal species from extinction? What is the price on preserving a primeval forest? Often, there are no markets for biodiversity and ecosystem services, their economic assessment only being done indirectly and imprecisely.

For this purpose, the development of risk scenarios associated with nature degradation has been studied by the Taskforce on Nature-related Risks of the Network for Greening the Financial System (NGFS). These methodological developments require the use of bio-physical models and models integrating the economy and ecosystem services into a unified structure. This requires a collective effort by researchers in various fields, such as biology, geology and economics. The current generation of integrated assessment models is an example of this collective effort.

The following analysis does not yet address other factors that may be relevant to analyse wellbeing in different climate scenarios. These include, for instance, migration conflicts across regions or the possible discomfort of living in a different climate than that typically experienced in a given location. Including such factors in the analysis would introduce additional complexity, given how hard it is to quantify their effects. Similarly, the effects of climate change on the population’s health and mortality are not covered. These tend to be studied through cost-benefit analyses. There are also hard-to-measure subjective factors, such as distributional effects of climate policies across generations. The study of the relevance of these factors will be left for future research.

IPCC global scenarios

Since the 1990s, the IPCC assessment reports have presented a number of long-term climate projections, underpinned by narratives for various climate scenarios. The latest generation of these scenarios has reconciled the narratives between the “Representative Concentration Pathways” (RCPs) and the “Shared Socioeconomic Pathways” (SSPs), which are included in the Sixth IPCC Assessment Report (IPCC 2021b).

The RCPs correspond to different GHG emission pathways and are organised in terms of their concentration in the atmosphere by 2100: examples are the RCP2.6, RCP4.5 and RCP8.5 scenarios (van Vuuren et al., 2011). They are named after the radiative forcing value caused by the concentration of GHGs in the atmosphere, i.e. the difference between the average incoming and outgoing energy at the top of the atmosphere making the planet progressively hotter. As such, RCP2.6 assumes a radiative forcing value of 2.6 watt/m2 by 2100, relative to the period 1850–1900, understood by the IPCC as a proxy to pre-industrial levels. Of these three scenarios, this is the most aligned with the goals set in the Paris Agreement to limit the temperature increase. RCP4.5 assumes a radiative forcing value of up to 4.5 watt/m2 by 2100. The worst emissions scenario is RCP8.5. As mentioned above, this forcing value implies, over a longer horizon, an increase in the overall Earth’s temperature compared to the pre-industrial period, referred to as a temperature anomaly.

The SSPs correspond to narratives on how the future may be, using internally consistent assumptions about the joint evolution of socio-economic factors such as demographics, technology, institutional organisation and lifestyles. As in the RCPs, SSPs comprise pathways compatible with a more sustainable world (SSP1, “Taking the Green Road”) and a world characterised by rapid growth in the use of fossil fuels (SSP5, “Taking the Highway”), by way of an intermediate scenario where observed historical trends remain (SSP2, “Middle of the Road”).

The latest IPCC report proposes combinations of SSP narratives and GHG emissions pathways implicit in the RCPs. The SSP1-2.6, SSP2-4.5 and SSP5-8.5 scenarios are produced in such a way. Chart 1 presents an illustration of the three scenarios. The SSP1-2.6 scenario is characterised by a temperature path which, by 2100, is already decreasing, albeit very slightly. The SSP2-4.5 scenario broadly stabilises the temperature anomaly below 3 °C in 2100. The “worst-case” scenario, corresponding to SSP5-8.5, projects a temperature anomaly slightly below 5 °C in 2100, but still rising at a strong pace.

Each projection is associated with a hard-to-quantify likelihood. There is some consensus that the SSP5-8.5 scenario can be seen as an extreme case, while the SSP1-2.6 scenario can be seen as ambitious, but not impossible, because it presumes strict public attitudes and policies. Of all three, SSP2-4.5 seems the most likely, given the commitments already made and the policies currently implemented by the largest emitters.

Some estimates8 suggest that, if countries comply with their Nationally Determined Contributions (NDCs) (instruments set out in the Paris Agreement where countries set carbon emission targets and strategies to achieve them), the temperature anomaly will be around 2.5 °C by 2100. If, in addition, other firm targets already taken by countries are also met, the anomaly will amount to 2.1 °C by the end of the century. These two estimates are lower than those in SSP2-4.5. In the optimistic assumption that all the targets already announced by countries are scrupulously met, it is possible to limit global warming to 1.8 °C by 2100, a similar figure to that advanced in the SSP1-2.6 scenario.

It is important to interpret these projections with great caution, given that uncertainty is significant. For instance, in the SSP1-2.6 scenario, the confidence interval for temperature in 2100 is 1.3-2.4 °C, with 2.5-3.5 °C for the SSP2-4.5 scenario and 3.3-5.7 °C for the SSP5-8.5 scenario.

  1. Global Earth surface temperature anomaly 1950–2100 relative to the period 1850-1900 for different scenarios

Source: RNA 2100. | Note: The uncertainty arises from the temperature being projected using many simulations within each model and from using different models for these projections.

  1. Climate projections for mainland Portugal in the 21st century

One of the core elements for assessing the vulnerability of the Portuguese territory to global warming is the projection of climate variables such as average, maximum and minimum temperature as well as precipitation. It is also important to analyse how such developments contribute to the more frequent materialisation of extreme events, such as droughts or floods. This box illustrates some of the results of the RNA 2100 project’s work packages dedicated to this issue.9 Climate scenario analysis takes as a benchmark the climate observed over a 30-year period, from 1971 to 2000.

Daily average temperature projections in mainland Portugal suggest an increase of 0-2 °C in the various scenarios for the period 2011–40 compared to the reference period (Chart B1.1). These differences in the projected daily average temperature widen significantly as the 21st century advances. The contrast between the coast and inland is noticeable in the RCP4.5 scenario in mid-century (2041–70), with a smaller temperature increase in the coastal region. In the extreme scenario RCP8.5 – very unlikely with currently available information – temperature may rise by 4–5 °C in 2071–2100.

  1. Daily average temperature projections in mainland Portugal

Sources: RNA 2100, Lima et al. (2023a, b) and Soares et al. (2023). | Note: Deviations from the period 1971-2000.

Precipitation projections for mainland Portugal suggest a decline in the course of the 21st century (Chart B1.2) for the RCP4.5 and RCP8.5 scenarios. In RCP4.5, there could be decreases of 0-20% compared to the reference period. The RCP8.5 scenario projects country-wide decreases of 10-40%.

Another relevant aspect for future climate classification is the number of days per year in which the maximum daily temperature exceeds 25 °C, known as summer days. The historical period shows a growing north-south gradient for this type of days in mainland Portugal, ranging from around 40 to 100 days per year in northern Portugal, and from 100 to 140 days per year in southern Portugal (Chart B1.3 – Panel A). The projected developments keep this gradient. In this metric, the increase occurs in all regions and for all three scenarios (Chart B1.3 – Panel B). In the RCP4.5 scenario, the increase will be gradual over the course of the century, with no significant increase in the first decades, to around 30 to 50 days per year in 2100. In the RCP2.6 scenario, summer days are projected to increase by 20-30 days by 2100, particularly in the coast. In the worst emission/concentration scenario, mainland Portugal will have around 70 additional summer days.

  1. Projections for total cumulative precipitation in mainland Portugal

Sources: RNA 2100, Lima et al. (2023a, b) and Soares et al. (2023). | Note: Deviations from the period 1971-2000.

  1. Panel A – Average number of days per year in mainland Portugal where the maximum daily temperature exceeds 25 °C (classified as summer days). Panel B – Change in the average number of summer days compared to the period 1971–2000

Panel A

Panel B

Sources: RNA 2100, Lima et al. (2023a, b) and Soares et al. (2023).

One of the consequences of global warming is the distortion of precipitation patterns. This effect is particularly noticeable in Chart B1.4. The maximum five-day cumulative precipitation makes it possible to measure the time concentration of precipitation. For the same area, the higher this value, the higher the likelihood of flash floods, landslides and floods. Despite the geographical pattern of five-day cumulative precipitation developments remaining over the horizon, the trend is on the upside across the country. These results, together with those presented above, show that in certain scenarios and locations precipitation will decrease, but there will be a greater number of heavy precipitation episodes. This simple observation may have major consequences in terms of water management policy in Portugal, as it implies a less steady water flow in Portuguese rivers.

  1. Panel A – Maximum five-day cumulative precipitation. Panel B – Changes in maximum five-day cumulative precipitation compared to the period 1971–2000

Panel A

Panel B

Sources: RNA 2100, Lima et al. (2023a, b) and Soares et al. (2023).

Climate projections also point to a small increase in the frequency of moderate drought. The largest increase will be in the RCP4.5 scenario, with approximately two events per decade in some regions of the country, towards the end of the century (Chart B1.5 – Panel A). However, the average duration of each event is expected to increase (Chart B1.5 – Panel B), especially in the RCP8.5 scenario, and could reach 12 additional months per event. These results are linked to projections of annual precipitation, given that the Standard Precipitation Index (SPI) is based solely on precipitation. In regions where projections point to a substantial reduction in annual precipitation, the frequency of moderate drought increases, while in regions where there is a reduction in the number of moderate drought episodes per decade, close-to-zero changes in annual precipitation are expected.

Finally, climate projections point to an increase in the number of days of extreme fire risk, particularly in the worst concentration scenarios and from the mid-century onwards (Chart B1.6). Inland regions will be the most affected.

  1. Panel A – Changes in the average number of moderate drought episodes per decade. Panel B – Changes in the average duration of moderate drought episodes in mainland Portugal

Panel A

Panel B

Sources: RNA 2100, Lima et al. (2023a, b) and Soares et al. (2023). | Notes: Figures calculated for the SPI index over a 12-month cumulative period, taking 1971-2000 as the reference period. The SPI index measures the severity of droughts based on the probability of precipitation in a given period.

  1. Anomaly in the number of extreme fire risk days in mainland Portugal

Sources: RNA 2100 and Bento et al. (2023). | Note: 1971–2000 is used as a reference period.

 

Macroeconomic impact estimates

The macroeconomic impact of global warming is a difficult concept, because it depends on the baseline scenario. In the literature on the topic, this impact is measured in relation to the gross domestic product (GDP) that would be observed if the temperature anomaly were zero. In any economy, climate change will lead agents to make an effort to adapt. Literature estimates already take into account such endogenous changes; the effect of public adaptation policies will also be discussed below.

General equilibrium integrated assessment models

The following analysis is based on a general equilibrium model comprising several interrelated modules: the carbon cycle, climate, economic damages and the economy (Adão et al., 2024). Given that it is a consistent description of how the natural climate system works and how the economy evolves, this is one of the preferred instruments used in cost-benefit analyses of climate policies.

The carbon cycle describes how CO2 emissions determine the amount of carbon in the atmosphere, while the climate module illustrates how the amount of carbon affects temperature. Changes in temperature directly influence input productivity, generating economic damages. As the performance of economic activities requires the use of fossil-fuel-based or renewable energy, and the use of fossil fuels generates carbon emissions, mitigation policies are needed to increase the relative importance of green energy.

The exercise combines SSPs with RCPs as presented in the IPCC’s Sixth Assessment Report. The model includes global instruments for taxes on fossil fuel emissions, which are only possible in light of SSP narratives, in such a way as to create paths in line with RCPs.

Results for the world

Estimates obtained from general equilibrium climate assessment models imply generally increasing economic damages worldwide in tandem with the temperature anomaly (Chart 2), and there seems to be an element of non-linearity. The estimates obtained from this work are somewhat more pessimistic than the average of available studies. In the range covered by the three scenarios considered, the results point to a 0.9 percentage point loss in global GDP for each additional Celsius degree in the overall Earth’s temperature.

Table 1 presents the results for four variables of interest: net emissions, temperature anomaly, gain in consumption against the SSP1-2.6 scenario and economic damage, for the various scenarios and at two points of time. The results refer to the world and a small open economy, calibrated for Portugal with economic and natural science data.10 In the exercise, the only policy variable is the level of the carbon tax: the higher the taxation on emissions, the more intense the replacement of fossil fuels, the lower the carbon emissions and the lower the temperature anomaly. In the model used in this work it is possible to calculate the optimal level for carbon taxation that maximises the wellbeing of agents. Using the level of taxation for each carbon unit issued, it is possible to mimic the three scenarios in terms of global emissions and temperature anomaly in 2050 and 2100. The worst emission scenario (SSP5-8.5) is linked to a lower tax level of around 15% of the optimal value. Similarly, the SSP1-2.6 scenario is underpinned by high taxation, 15 times the optimal value. The SSP2-4.5 scenario can be obtained with intermediate taxation, of around two-thirds of the optimal value. The optimal tax is the same in all regions of the world, including Portugal. It should be noted that the model does not include several negative aspects of climate change, which explains why the optimal value found using the model differs from the taxation level in the lower carbon scenario.

  1. Economic damage as a percentage of global GDP using integrated assessment models with general equilibrium

Sources: Tol (2024), RNA 2100, and Adão et al. (2024). | Notes: Economic damage as a percentage of global GDP relative to the scenario with no global land temperature anomaly, in 2100. The results were obtained by various authors using integrated assessment models with general equilibrium. In cases where different damages were reported for the same anomaly, the average of those damages was used. The blue dots refer to the results from this Special Issue.

The metric used to measure individuals’ consumption gains corresponds to the change in consumption, measured in terms of utility, that economic agents living in an economy characterised by the implementation of mitigation policies under the SSP1-2.6 scenario would have to experience to remain indifferent to living in the scenario under assessment.11 Granted, this metric does not address all relevant effects affecting the wellbeing of economic agents, some of which mentioned above, which would alter the magnitude of these consumption gains.

In addition, no significant investment is considered to have been made in public adaptation policies, which makes it possible to assess the effectiveness of mitigation policies as a separate instrument. An analysis of the optimal adaptation policy at local level is presented below.

The results at global level depicted at the top of Table 1 show that, in the SSP1-2.6 scenario, the global temperature anomaly stands at 1.3-1.4 °C, with net emissions decreasing from 35 GtCO2 in 2020 to 5 GtCO2 in 2050, turning negative towards the end of the century. In the intermediate scenario, however, net emissions fall by 10 GtCO2 by the end of the century compared to 2020. Yet, in the SSP5-8.5 scenario, they increase considerably, meaning that the temperature anomaly by 2100 is high (2.9 °C). This leads to economic damages that may be as high as 3.1% of GDP by 2100.

  1. Main results without adaptation policy for the world and Portugal

World

 

SSP1-2.6

SSP2-4.5

SSP2-8.5

 

2020

2050

2100

2050

2100

2050

2100

Net emissions (GtCO2 per year)

35

5

-2

27

25

42

111

Temperature anomaly (°C)

1.0

1.3

1.4

1.5

2.2

1.6

2.9

Consumption gain (%)

0.3

0.3

0.2

0.4

Economic damage (%)

0.6

0.9

0.9

1.2

1.8

1.4

3.1

Portugal

 

SSP1-2.6

SSP2-4.5

SSP2-8.5

 

2020

2050

2100

2050

2100

2050

2100

Net emissions (MtCO2 per year)

36

0

-8

18

10

25

41

Temperature anomaly (°C)

1.2

1.6

1.7

1.8

2.6

1.9

3.5

Consumption gain (%)

-0.4

-0.4

-0.6

-0.9

Economic damage (%)

1.0

1.0

1.0

1.3

1.9

1.5

3.3

Souce: RNA 2100. | Note: The consumption gain refers to the indicated year and is calculated relative to consumption in the SSP1-2.6 scenario for the same year. The economic damage is calculated relative to a scenario with no temperature anomaly.

Overall, the results show that the average economic agent would be slightly better, in terms of consumption, at the end of the century in scenarios with less strict mitigation policies. The model suggests that individuals experience a consumption loss of 0.3% in the higher taxation scenario compared to the intermediate taxation scenario (both in 2050 and 2100). This may create some resistance to the implementation of very high corrective taxes, as is the case in this scenario.12

The model projects that consumption gains in 2020-2100 will be quantitively similar in the SSP2-4.5 and SSP5-8.5 scenarios compared to the SSP1-2.6 scenario. The fact that carbon taxation is below the optimal value in these two scenarios leads the economy to keep using fossil fuels more than what would be desirable, in particular in the SSP5-8.5 scenario.

Let us present two final notes on these results. In the model, revenues from carbon taxes are considered to be returned to economic agents proportionally to income, for the sake of simplification. However, the literature recognises that carbon taxation produces different outcomes depending on how that revenue is used to finance economic activities.13 On the other hand, it should again be noted that the consumption measures used do not include non-market effects, including biodiversity loss and other consequences of global warming.

The Portuguese case

Although policies to reduce carbon emissions should be implemented globally, the effects of climate change are heterogeneous and felt locally, which warrants an analysis of impacts on the Portuguese economy. Climate scientists recognise that the materialisation of the effects of climate change depends, among other things, on latitude and proximity to the sea. Portugal thus has significant differences compared with the world economy, as a result of its geographical positioning above the equatorial line and by the sea.

Climate scientists in the RNA 2100 consortium estimate that the temperature anomaly in Portugal is 20% higher than in the rest of the world. In the SSP1-2.6 scenario, the temperature anomaly will stabilise at 1.7 °C by the end of the century (Chart 3). In Portugal, carbon neutrality can be achieved by 2050, i.e. net zero emissions, aided by two factors. First, 39% of the Portuguese territory is covered by forest, more than the rest of the world (31%). In Portugal, forests are responsible for the sequestration of about 1/6 of emissions. Second, this reduction in emissions is underpinned by a population decrease in Portugal in all three scenarios, of around 30% in 2050, in line with demographic developments estimated by Statistics Portugal. By contrast, it is assumed that in the rest of the world the population does not vary, as discussed in Hassler et al. (2021).

  1. Temperature anomaly in Portugal | Degrees Celsius (°C)

Source: RNA 2100. | Note: Deviations compared to the pre-industrial period.

The scientific literature also acknowledges that the global externality has a greater macroeconomic impact on Portugal than on the rest of the world. This means that for any change in the carbon stock in the atmosphere, the economic damage is greater for Portugal than for the rest of the world. This results from the fact that temperature is increasing more in Portugal. In this way, a higher global carbon tax provides higher wellbeing levels for Portugal than for the rest of the world. In the SSP1-2.6 scenario, the economic damage amounts to 1% of GDP over the projection horizon, the same as that currently observed (Chart 4).

Climate scientists in the RNA 2100 consortium estimate that the SSP2-4.5 scenario is currently the most likely of the three considered. Carbon taxation underlying this scenario is around two-thirds of the optimal value and produces a temperature anomaly in Portugal of 1.8 ºC and 2.6 °C in 2050 and 2100 respectively. Net emissions in Portugal decrease from 36 MtCO2 in 2020 to 10 MtCO2 in 2100, without ever being cancelled off. In this scenario, the impacts on the Portuguese economy are considerable. Compared to 2020, they amount to an additional loss of 0.3% of GDP in 2050, and an additional loss of 0.9% of GDP in 2100. Utility gains in equivalent consumption are negative compared to the very high carbon tax scenario (SSP1-2.6) (Chart 5). This is because the climate externality penalises Portugal more harshly than the rest of the world.

  1. Economic damages in Portugal | As a percentage of annual GDP

Source: RNA 2100. | Note: The economic damage is measured relative to what would be observed without temperature anomaly.

  1. Consumption gain in Portugal relative to the high carbon tax scenario | As a percentage

Source: RNA 2100.

Finally, underlying the low carbon taxation scenario in Portugal (SSP5-8.5) is a tax of around 15% of the optimal value. This scenario is particularly severe and produces a temperature anomaly of 3.5 °C in 2100, with an upward net carbon emission path, standing at 41 MtCO2 in 2100. Economic damages increase by 2.3% of GDP over the whole simulation horizon. In terms of consumption, the scenario is increasingly unfavourable compared to the other two as the end of the century approaches.

Portugal under adaptation policies

The results described so far do not take into consideration public adaptation policies. Table 2 breaks down the main outcomes when optimal public adaptation policies are included, in conjunction with global mitigation policies. Spending on adaptation aims to mitigate economic damages associated with the effects of climate change. According to the literature, this mechanism is modelled as an effect limiting economic damages depending on the intensity of public investment in adaptation. The effect grows, but is concave, in that intensity.

As mentioned earlier, investment in adaptation has two effects. First, it reduces the impact of the climate externality on the economy. Second, it absorbs resources that could be allocated for other uses.14

  1. Results with optimal adaptation policy for Portugal

Portugal

 

SSP1-2.6

SSP2-4.5

SSP2-8.5

 

2020

2050

2100

2050

2100

2050

2100

Net emissions (MtCO2 per year)

36

0

-8

18

10

25

42

Temperature anomaly (°C)

1.2

1.6

1.7

1.8

2.6

1.9

3.5

Consumption gain (%)

-0.3

0.0

-0.5

0.2

Economic damage (%)

1.0

1.0

0.9

1.1

1.3

1.2

1.9

Adaptation costs (%)

0.1

0.0

0.2

0.4

0.2

0.7

Difference to no-adaptation scenario

 

SSP1-2.6

SSP2-4.5

SSP2-8.5

 

2020

2050

2100

2050

2100

2050

2100

Net emissions (MtCO2 per year)

0

0.0

0.0

0.1

0.2

0.1

1.0

Consumption gain (pp)

0.1

0.4

0.1

1.1

Economic damage (pp)

0

-0.1

0.0

-0.2

-0.6

-0.3

-1.4

Source: RNA 2100.

In the SSP1-2.6 scenario, adaptation reduces economic damages by a very small share of GDP. Adaptation costs (as a percentage of annual GDP) are negligible in this case. In the intermediate scenario (SSP2-4.5), adaptation reduces the economic damages more significantly, by 0.2% and 0.6% of GDP in 2050 and 2100 respectively.

Finally, in the worst emissions scenario (SSP5-8.5), the adaptation reduces economic damages by 1.4% of GDP in 2100 and leads to an equivalent consumption gain of 1.1% compared to the no-adaptation scenario. Adaptation costs in 2100 are estimated at 0.7% of GDP, significantly below the reduction in economic damages.

It should be noted that adaptation alone does not entail a reduction in emissions. In fact, the opposite may be true. By reducing the damage caused by the climate externality, but not the externality itself, adaptation makes it possible for the economy to operate at a higher level and thus with higher emissions. This effect is visible in the bottom panel of Table 2, where net emissions compared to the no-adaptation scenario are marginally positive, chiefly in the SSP2-4.5 and SSP5-8.5 scenarios.

Adaptation measures are sufficient to considerably reduce equivalent consumption losses in the SSP2-4.5 and SSP5-8.5 scenarios against the high carbon tax scenario (SSP1-2.6) compared to the no-adaptation scenario. As expected, adaptation reduces economic damage in all scenarios and horizons, with greater reduction in scenarios with larger losses. The effect expands over longer horizons because the differences in the carbon stock associated with the various scenarios are larger in the long run.

These results suggest that adaptation is an effective tool to offset possible shortfalls in mitigation efforts, which are only effective if there are global policies in place. Considering these results as a whole, the higher the mitigation effort – i.e. the higher the global carbon tax or equivalent mitigation measures – the less local adaptation is needed, at least in the Portuguese case. This suggests an important role for adaptation in improving wellbeing at local level in the face of global carbon taxation policies. Without mitigating measures, which would correspond to an even worse scenario than SSP5-8.5, the costs would be very significant, not matchable by adaptation efforts.

Conclusions

RNA 2100 is a joint effort by several institutional partners dedicated to studying the vulnerability of the Portuguese territory to climate change in the 21st century. The project’s climate projections are based on three of the IPCC scenarios, subsequently converted to the Portuguese case. This Special issue uses these scenarios as a basis to make a macroeconomic assessment for Portugal. Each of the three IPCC scenarios is implemented through global taxation on carbon emissions. This simplified mitigation policy is characterised by three very different carbon tax levels: high for the SSP1-2.6 scenario, which is a proxy for that expected in the Paris Agreement; intermediate for SSP2-4.5, currently considered as feasible; low for SSP5-8.5, considered as an extreme case.

In the more benign scenario of near compliance with the Paris Agreement (SSP1-2.6), additional economic damages are negligible and entail high global carbon taxation, several times above the optimal level obtained with the model. In this scenario, there is also no need for substantial efforts to adapt via public policies. For Portugal, this is the most favourable scenario in terms of wellbeing strictly linked to consumption; presumably, it will also be the case for the other effects of global warming not included in this analysis, as these additional effects are negligible.

In the intermediate scenario, seen by several experts as feasible with the current mitigation policies (SSP2-4.5), the effects of global warming on the economy are significant. The overall carbon taxation implemented in this scenario is below the optimal level obtained through the model. From a perspective where wellbeing is strictly driven by the utility of consumption, and taking into account the world as a whole, agents prefer this scenario at any given point in time to the low emissions scenario. While it is true that many of the negative effects of global warming are not taken into account in our analysis, this observation still poses challenges for the overall implementation of a strict mitigation policy, where not accompanied by adaptation measures. Given Portugal’s geographical exposure, this scenario implies a worse situation compared to the rest of the world, both in terms of GDP and wellbeing strictly linked to consumption. These qualitative features are also observed in the most extreme scenario (SSP5-8.5), but are reinforced in quantitative terms.

The analysis also shows that public adaptation policies help mitigate the local economic impacts of climate change. The model suggests that such policies almost reverse the effects of the carbon externality, at an additional cost. For Portugal, however, the hierarchy of preference between scenarios remains, with blurred differences: the most beneficial scenario for Portugal in terms of wellbeing from consumption is SSP1-2.6, in line with the Paris Agreement.

Among the issues that warrant an in-depth analysis is the development of technologies that can bring about considerable mitigation, including increased efficiency in renewable energy production and storage, and carbon absorption in the atmosphere through natural (e.g. forests) or artificial methods. Other issues not analysed here, partly due to their complexity, are the taxation of carbon content on international trade flows and the climate change-driven mobility of individuals.

On a final note, high uncertainty surrounds this exercise. For that reason, the results presented here should be read as imprecise estimates of an uncertain future – which can be chosen by everyone today, countering a problem that affects us all.

References

Adão, B., Antunes, A., Gouveia, M., Lourenço, N., and Valle e Azevedo, J. (2022), “Climate Change and the Economy: An Introduction”, Occasional Papers, No 1, Banco de Portugal.

Adão, B., Antunes, A., and Lourenço, N. (2024), “Assessing the Macroeconomic Effects of IPCC Scenarios: Mitigation, Adaptation, and Carbon Sinks”, Working Papers, No 13, Banco de Portugal.

Aligishiev, Z., Bellon, M., and Massetti, E. (2022), “Macro-Fiscal Implications of Adaptation to Climate Change”, Staff Climate Note 2022/002, International Monetary Fund.

Bento, V. A., Lima, D. C. A., Santos, L. C., Lima, M. M., Russo, A., Nunes, S. A., DaCamara, C. C., Trigo, R. M., and Soares, P. M. M. (2023), “The future of extreme meteorological fire danger under climate change scenarios for Iberia”, Weather and Climate Extremes, Vol. 42 (100623). https://doi.org/10.1016/j.wace.2023.100623

Coase, R. H. (1960), “The Problem of Social Cost”, Journal of Law and Economics, 3, 1-44.

Hasna, Z., Lourenço, N., and Santos, C. (2022), “On the aggregate and distributional effects of carbon taxation in Portugal”, Banco de Portugal Economic Studies, Vol. VIII, No 3, Banco de Portugal.

Hassler, J., Krusell, P., and Olovsson, C. (2021), “Suboptimal climate policy”, Journal of the European Economic Association, 19(6), pp. 2895-2928.

IPBES (2019), Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, E. S. Brondizio, J. Settele, S. Díaz, and H. T. Ngo (eds.). IPBES Secretariat, Bonn, Germany. https://doi.org/10.5281/zenodo.3831673

IPCC (2021a), “Summary for Policymakers”, In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Masson-Delmotte, V. et al. (eds).

IPCC (2021b), Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Masson-Delmotte, V. et al. (eds).

Lima, D. C. A., Lemos, G., Bento, V. A., Nogueira, M., and Soares, P. M. M. (2023a), “A multi-variable constrained ensemble of regional climate projections under multi-scenarios for Portugal – Part I: An overview of impacts on means and extremes”, Climate Services, Vol. 30, 100351. https://doi.org/10.1016/j.cliser.2023.100351

Lima, D. C. A., Bento, V. A., Lemos, G., Nogueira, M., and Soares, P. M. M. (2023b), “A multi-variable constrained ensemble of regional climate projections under multi-scenarios for Portugal – Part II: Sectoral climate indices”, Climate Services, Vol. 30, 100351. https://doi.org/10.1016/j.cliser.2023.100351

Pigou, A. C. (1920), The Economics of Welfare, London: Macmillan.

Soares, P. M. M., Careto, J. A. M., Russo, A., and Lima, D. C. A. (2023), “The future of Iberian droughts: a deeper analysis based on multi-scenario and a multi-model ensemble approach”, Natural Hazards, Vol. 117, pp. 2001-2028. https://doi.org/10.1007/s11069-023-05938-7

Tol, R. S. J. (2024), “A meta-analysis of the total economic impact of climate change”, Energy Policy, Vol. 185, 113922.

van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., Hurtt, G. C., Kram, T., Krey, V., Lamarque, J.-F., et al. (2011), “The representative concentration pathways: an overview”, Climatic Change, 109(5): 5–31.

Policy insights

The implementation of the Recovery and Resilience Plan in Portugal1

The Recovery and Resilience Facility (RRF) is the cornerstone of NextGenerationEU (NGEU), designed to strengthen the European economy in the aftermath of the COVID-19 pandemic. Through this instrument, the European Commission raises funds in capital markets by issuing bonds on behalf of the European Union (EU), which are then distributed to Member States in the form of grants and loans. The latter should implement reforms and investments by the end of 2026 that make their economies and societies more sustainable, resilient and prepared for the green and digital transitions.

To benefit from the support, EU governments have submitted national recovery and resilience plans (RRPs), setting milestones and targets to be achieved, on which the disbursement of the funds depends. These plans were later revised, in particular following the REPowerEU initiative that was set up to address the socio-economic difficulties and disruption of the energy market caused by Russia’s unjustified invasion of Ukraine. This Policy insights focuses on the implementation of Portugal’s RRP, highlighting the amounts involved, the characterisation of projects and beneficiary entities, as well as the progress achieved so far. Reforms of the RRP are not included in this analysis and should be the subject of future studies.

pin

Portugal is the fifth country in the euro area to receive more RRF funds as a percentage of GDP.

The RRF allocations to the 20 countries of the euro area amount to €529 billion, equivalent to 4.4% of 2019 GDP (Chart 1). Of this amount, 55% are grants and the rest are loans. The RRF grants were distributed across EU countries based on criteria such as population, GDP per capita and the unemployment rate, to ensure an allocation of funds benefiting those countries whose recovery appears more complex given the economic and social shock resulting from the pandemic crisis. The Member States could apply for RRF loans until August 2023, and approximately 75% of the total available was requested.

Portugal is the fifth euro area country that will receive more RRF funds: 10.4% of 2019 GDP, corresponding to €22.2 billion (€16.3 billion in grants and €5.9 billion in loans).

The RRPs are structured around three dimensions: green transition, digital transition and resilience. At least 37% of the funds should be allocated to green measures and 20% to digital initiatives. Based on a less detailed classification, the funds in Portugal are distributed among these three dimensions as follows: 20% for green transition, 12% for digital transition and 68% for resilience. However, as the resilience dimension includes investments with green and digital elements, the Portuguese RRP meets the European requirements, by allocating 41% and 21% of the funds to climate and digital objectives (Chart 2).

  1. Total RRF grants and loans in the euro area countries | Percentage of 2019 GDP

  1. Funds allocated to the green and digital transitions in the euro area countries | Percentage of the total allocation per country

Source: European Commission.

pin

Portugal has received four tranches so far, corresponding to 38% of the allocation, and has achieved 23% of the milestones and targets, with compliance becoming more demanding in the last payment requests.

Each Member State’s RRPs include specific and verifiable milestones and targets to ensure effectiveness and accountability in the use of funds. Milestones are qualitative steps or specific events that indicate progress in project implementation, while targets represent quantitative objectives to be achieved. The European Commission monitors the fulfilment of these milestones and targets, making the disbursement of funding tranches conditional on their achievement.

Portugal has set 202 milestones and 261 targets broken down into 123 investments and 44 reforms. So far, 75 milestones and 30 targets have been fulfilled, equivalent to 23% of the total (Chart 3 – Panel A). Most milestones focused on the adoption of legislation and the conclusion of contracts and agreements. The targets reached cover several areas, with particular emphasis on affordable housing and investment and innovation. In terms of funds received, Portugal has already received the first four tranches, totalling €7 billion, corresponding to 38% of the planned allocation (Chart 3 – Panel B). Currently, the implementation percentages, measured in terms of fulfilled milestones and targets or funds received, are below the euro area average.

In early July this year, Portugal submitted its fifth payment request to the European Commission, which, according to the schedule, should have been made in the first quarter. The fulfilment of the milestones and targets associated with this request brings the fulfilment rate to 32%, which could improve Portugal’s relative position. Moreover, if the request for the sixth payment is submitted and approved in 2024, the delay in meeting targets and milestones compared to the original plan would be recovered (Chart 4 – Panel A). As the payment requests progress, the number of targets becomes more significant, especially as of the ninth tranche, making fulfilment more demanding (Chart 4 – Panel B).

  1. RRF funds received and fulfilment of milestones and targets in the euro area | Percentage of the total

Panel A – Milestones and targets

Panel B – Funds received

Source: European Commission. | Note: Status as of September 24, 2024. In the charts, the values for Portugal associated with the triangles correspond to the percentages after the approval of the 5th payment request.

  1. Milestones and targets in Portugal | Number

Panel A – Per year

Panel B – Pey payment request

Sources: Recuperar Portugal and European Commission. | Notes: In Panel A, two of the 75 milestones met and one of the 30 targets achieved only reached that status in August 2024. In Panel B, the x-axis represents the payment requests and the corresponding amount.

pin

Of the RRP funds, 87% have been approved and 24% have been paid out to beneficiaries so far.

The data provided by Recuperar Portugal allows tracking the progress of RRP funds approvals and payments. The funds are distributed across 21 components, aligned with the three structuring dimensions mentioned before. Currently, 87% of the allocations to beneficiaries have been approved, and 24% have already been disbursed (Chart 5). The percentage of funds paid out is highest in the digital transition (30%) and lowest in the green transition (20%).

The components with the highest allocation are C5 – Investment and Innovation, C2 – Housing and C6 – Qualifications and Skills, which account for almost half of the total allocation. These components also stand out for their high share of loans. Most components have an approval rate above 75%, with the lowest rates seen in C21 – REpowerEU and C13 – Energy Efficiency in Buildings. For C21 – REpowerEU, the lower rate is due to its later inclusion in the RRP. The share of payments is highest in components C20 – Digital School, C10 – Sea and C5 – Investment and Innovation.

  1. Financial implementation of the RRP by structuring dimension and component | In millions of euros

Sources: Recuperar Portugal and calculations by Banco de Portugal. | Notes: Status as of September 18, 2024. Components are ordered by total allocation in millions of euros. The values presented by component at the end of each bar represent the total allocation in millions of euros and the percentage of grants. The letter at the end of each component indicates the structuring dimension: 'R' – Resilience, 'TC' – Green Transition; 'TD' – Digital Transition. In C11 –Decarbonising industry - the value presented in the chart corresponds to the approved amount, which exceeds the initially planned allocation.

In National Accounts, the RRF funds are recorded as they are spent by the general government. Amounts not executed directly by general government entities appear as current or capital transfers to other sectors of the economy. Of the total allocation, 10% was recorded in National Accounts as executed by the end of 2023. In terms of composition, 71% was recorded as capital expenditure (29% public investment and 41% capital transfers to other sectors of the economy) and 29% as current expenditure. Execution and distribution by expenditure category over the remaining horizon are still uncertain.

The figures recorded in National Accounts may differ from the payment information provided by Recuperar Portugal due to differences in coverage (particularly as regards local government) and in the timing of the recording, as well as to the inclusion of loans classified as financial transactions. By the end of the first half of 2024, payments totalled €4.6 billion, compared to €3.1 billion in the National Accounts.

pin

More than half of the funds approved are allocated to the general government, but payments are progressing faster in the private sector.

The same information provided by Recuperar Portugal makes it possible to classify entities with approved projects by institutional sector. Of the approved amounts for direct and final beneficiaries,2 62% are allocated to the general government (GG), 5% to state-owned enterprises outside the GG perimeter and 33% to private sector entities, firms or households.

The distribution of approved amounts by entity varies across institutional sectors. The mean, median and interquartile range of approved amounts are higher in state-owned enterprises, followed by the GG and, lastly, the private sector (Chart 6 – Panel A). The higher median in state-owned enterprises indicates the prevalence of larger projects in this sector. However, it is in the GG that the entities with the largest authorised allocations are to be found: Instituto da Habitação e da Reabilitação Urbana (€756 million)3, Metropolitano de Lisboa (€748 million), Infraestruturas de Portugal (€512 million), Metro do Porto (€418 million) and Serviços Partilhados do Ministério da Saúde (€30.1 million). For state-owned enterprises outside the GG perimeter, the highest approved amounts correspond to the Banco Português do Fomento (€268 million), Águas do Algarve (€69 million) and Empresa de Eletricidade da Madeira (€97 million).

The percentage of approved amounts that has been disbursed is, on average, higher in the private sector (Chart 6 – Panel B). In the large projects mentioned above, the percentage of payments is low: 19% in the five entities with the highest approved values.

  1. Distribution of approved funds and of the percentage disbursed

Panel A – Approved funds

Panel B – Percentage disbursed

Sources: Recuperar Portugal and calculations by Banco de Portugal. | Notes: Status as of September 18, 2024. Only entities with approved projects exceeding 1 million euros are considered. The boxes show the values between the 25th and 75th percentiles (interquartile range), the horizontal lines inside them represent the median of the distribution, and the "x" marker identifies the mean. The outer horizontal lines correspond to the maximum and minimum of the distribution, excluding outliers. Outliers are observations that fall below the first quartile minus 1.5 times the interquartile range or above the third quartile plus 1.5 times the interquartile range.

pin

In the private sector, approved funds were channelled to older, more productive and more export-intensive firms.

Focusing on the private sector, the number of entities with approved projects is similar across micro, small, medium and large-sized enterprises (Chart 7 – Panel A). The average approved amount per entity is higher in large enterprises. Nevertheless, almost half of the funds approved are allocated to micro and small-sized enterprises. In terms of the classification of economic activities, diversity is quite high (Chart 7 – Panel B). Three activities with greater weight stand out: (i) manufacturing (32%), such as the manufacture of chemicals, man-made fibres, pulp and paper; (ii) professional, scientific and technical activities (24%), in particular, scientific research and development; (iii) and financial and insurance activities (13%). These activities accumulate 70% of the approved funds, more than twice their share in the total economy (30%). In turn, activities related to wholesale and retail trade, despite being the most prominent in the economy, are small in terms of approved amounts.

A comparison with all private non-financial firms in the Portuguese economy shows that firms benefiting from the RRP have an average and median age above that observed in the total economy, regardless of size (Chart 8 – Panel A). On average, beneficiary firms are 11 years older. These firms also tend to be more productive (Chart 8 – Panel B). However, in large firms, the average is much lower due to a minority of non-beneficiaries with very high productivity. Regarding external openness, beneficiary firms stand out for their higher export intensity, particularly strong in microfirms and SMEs, where the weight of exports reaches, on average, 32% of turnover, in contrast to the 7% observed in the total for these firms (Chart 8 – Panel C).

  1. Breakdown of approved amounts in the private sector | In number, millions of euros, and percentage

Panel A – By size

Panel B – By economic activity

Sources: Recuperar Portugal, Sistema de Partilha de Informação de Referência (SPAI) and calculations by Banco de Portugal. | Notes: Status as of September 18, 2024. Only entities with approved projects exceeding 1 million euros are considered.

  1. Age, apparent labour productivity, and export intensity in private sector firms benefiting from the RRP and in the total economy

Panel A – Firm age | Years

Panel B – Labour productivity | 1000 euros per employee

Panel C – Export intensity | Percentage

Sources: Recuperar Portugal, IES and calculations by Banco de Portugal. | Notes: Status as of September 18, 2024. The information from the IES corresponds to the year 2022 and does not include firms created afterwards. The financial and real estate sectors have been excluded from the analysis. Labor productivity is measured by the ratio of total sales over the number of employees, and export intensity is measured by the weight of exports in total sales. Only entities with approved projects exceeding 1 million euros are considered. The boxes show the values between the 25th and 75th percentiles (interquartile range), the horizontal lines inside them represent the median of the distribution, and the "x" marker identifies the mean. The outer horizontal lines correspond to the maximum and minimum of the distribution, excluding outliers. Outliers are observations that are below the first quartile minus 1.5 times the interquartile range or above the third quartile plus 1.5 times the interquartile range.

pin

The pace of RRP implementation varies across countries, with no clear link to the quality of the general government, and challenges in implementing the plans are expected.

As mentioned above, the disbursement of funds to each Member State is contingent on meeting the established milestones and targets. Chart 9 shows that the percentage of milestones and targets achieved does not seem to correlate with the relative importance of RRF funds in each country. Moreover, the comparison between the European Quality of Governance Index and the fulfilment rate of milestones and targets also indicates the absence of a relationship between the two variables, i.e. countries with a similar quality of governance have, so far, different fulfilment rates.

In the case of Portugal, the first 2024 report of the Comissão Nacional de Acompanhamento do PRR (CNA-PRR, Portuguese RRP monitoring committee) was recently released. The report identifies several constraints, including delays in the analysis of applications and difficulties in the functioning of the platforms for submitting costs. Among major investments, such as metropolitan expansion and modernisation projects, the report points to significant delays due to difficulties in public procurement, shortages of skilled labour, and problems in issuing opinions and environmental authorisations. These delays may jeopardise the deadlines for completion, initially set for 2026. The report attaches a critical or worrying assessment to almost 40% of investments (Chart 10). The Government has recently adopted legislation aimed at speeding up the implementation of the RRP by enhancing the monitoring of milestones and targets and the transparency of procedures.

  1. Quality of government and percentage of RRF funds received | Index and percentage

  1. Qualitative assessment of investment execution by CNA-PRR | Percentage

Sources: European Commission and calculations by Banco de Portugal. | Note: The European Quality of Governance Index is based on a survey to citizens, where respondents are asked about their perceptions and experiences with government participation and accountability, government effectiveness and the quality of public services provided, the rule of law, as well as the control of corruption in the public sector. The size of the circles corresponds to the RRF allocation for each Member State as a percentage of their 2019 GDP. The percentage of milestones and targets met reflects the situation as of September 24, 2024.

Source: CNA-PRR Report 1 of 2024. | Note: The qualitative assessment is based on the execution of investments up to June 2024.

The European Commission also announced guidelines that increase Member States’ flexibility to adjust RRP-funded projects, allowing them to add, delete or modify investments and reforms to reduce the associated costs and to improve the response to unforeseen challenges, such as delays in public procurement and skilled labour shortage. Moreover, the Member States can transfer RRP funds to a financial scheme that then uses the amount to boost investments by private entities. Member States generally have two options: to transfer the funds to the Member State compartment under the InvestEU (as a budgetary guarantee) or use another structure, for instance at national level.

pin

Continuous monitoring and adaptation of implementation strategies are essential to ensure an efficient execution of the plan.

The findings highlight that although Portugal has made significant progress in implementing the RRP, it is crucial to speed up project implementation to achieve the objectives set. The fulfilment of the milestones and targets, which will progressively become more demanding in the upcoming phases, will require further attention. The European Commission’s recent decision to provide greater flexibility in implementing RRP-funded projects offers a valuable opportunity to face unforeseen challenges and ensure investment materialisation. In this context, continuous monitoring and adaptation of implementation strategies are key.

As implementation progresses, the increased amount of available information will allow for more comprehensive and in-depth analyses. These should include a more detailed characterisation of the beneficiaries of the funds and an assessment of the economic impacts, considering both the investments made and the reforms carried out.


The Portuguese economy:

Climate scenarios for the Portuguese economy

The implementation of the Recovery and Resilience Plan in Portugal