Employment in Portugal’s Tourism Sector: Structural Transformation and Working Conditions from 2012 to 2022

Abstract

This study analyses employment in the tourism sector and seeks to identify trends in the evolution of worker profiles and transformation of the structure and working conditions in Portugal’s tourism sector from 2012 to 2022. It aims to understand how profiles, qualifications, and working conditions relate to the spatial distribution among Portuguese tourist regions and the typology and scale of tourism businesses, contributing valuable insights to defining sector strategies. It applies a quantitative approach based on statistical data from the Portuguese Ministry of Labour, Solidarity, and Social Security, disaggregated by the three segments of economic activity in the tourism sector: accommodation and food services, recreational and cultural activities, and transport and logistics. Descriptive statistics, MANOVA, and ANOVA tests analyse differences based on establishment size, tourist regions, and activity segments. The results show significant employment growth, regional variations, high turnover, and an increase in fixed-term contracts. Weak but significant correlations link education, nationality, gender, and working hours, indicating potential inequalities. The study highlights gender and educational differences among workers, as well as disparities related to employment status and nationality.

1. Introduction

In recent years, the tourism sector has become established as one of the most dynamic parts of the global economy. It covers various segments of economic activity that have experienced continual growth and significant diversification, emerging as one of the fastest-growing and most developed sectors worldwide. Currently, it is recognised in the literature as a strategic sector with a substantial economic, social, cultural, and environmental impact [1,2,3,4,5,6]. It is a key element for economic development, particularly in some regional economies, due to its ability to attract investment.

According to data from the World Travel & Tourism Council (WTTC) [7], in 2024 the travel and tourism sector contributed US$10.9 trillion to the global Gross Domestic Product (GDP), accounting for 10% of the global economy. The same source states that the sector is also a substantial source of employment, supporting 357 million jobs worldwide, which is roughly 1 in 10 of all jobs. Domestic visitors spent a total of US$1.9 billion, a 5.4% increase on the previous year, while international visitor expenditure grew by 11%, reaching US$1.9 trillion.

In Portugal, the sector has also experienced strong growth, and in 2024, all previous peaks in terms of economic contribution, employment, and visitor spending were surpassed [8]. According to data from the WTTC study [8], the performance of the tourism sector in Portugal reached its highest levels ever. The sector contributed €60.6 billion to the Portuguese economy, accounting for 21.3% of the national GDP, and supported 1.2 million jobs, approximately 23% of total employment in the country. For 2025, the study estimates that the tourism sector in Portugal will continue to grow, projecting a contribution of €62.7 billion to the national economy, representing 21.5% of GDP. Regarding employment, the study predicts that the sector will support 1.2 million jobs nationwide, or nearly one in four jobs in Portugal, an increase of 200,000 jobs compared to the previous peak.

Tourism has the potential to drive growth and generate employment opportunities due to its labour-intensive nature and multiplier effect, making it a vital sector for job creation [1]. According to the International Labour Organization [9], each job directly created in the tourism sector results in approximately 1.5 additional jobs in related industries. In this context, it is crucial to examine issues related to employment in the sector, particularly trends in employment and working conditions.

Since employment and human talent are vital resources for the tourism sector, this topic has attracted increasing academic interest, although the existing literature on it remains limited. According to the literature, while tourism is one of the most crucial sectors for economic growth and employment worldwide, it faces significant workforce challenges, including low wages, precarious employment, gender inequality, work–life balance issues, poor working conditions, and talent acquisition and retention difficulties [10,11].

Currently, strategic human resource management in the tourism and hospitality sector highlights the challenges employers face in attracting and retaining talent, which is seen as a key driver behind any company [12,13,14,15]. Even before the COVID-19 pandemic, the tourism industry was already struggling to attract and retain workers, mainly due to societal perceptions of the industry itself, including poor working conditions, low wages, seasonality, and unclear career paths [16].

New policies are essential to unlock the tourism sector’s potential to generate more and better jobs while reducing the risk of a growing skills mismatch. It is vital to develop and implement sustainable tourism promotion strategies that create employment and support local culture and products [17].

In this context, and considering the existing gap in the literature, the primary objective of this study is to analyse the evolution of employment in the tourism sector in Portugal between 2012 and 2022. It aims to identify trends in the development of worker profiles, the labour system, and working conditions in companies that respond to the current challenges facing tourism in Portugal. Specifically, the aim is to understand the sector’s dynamics, employment structure, sociodemographic characteristics of workers, working conditions, and professional qualifications across all companies with tourism as their main economic activity (Portuguese Classification of Economic Activities, CAE-REV 3), with an analysis from a regional perspective, considering the different regions of Portugal (NUT II).

To achieve the proposed objectives, a quantitative approach was adopted based on secondary data collected by the Strategy and Planning Office (GEP) of the Ministry of Labour, Solidarity, and Social Security regarding the tourism sector in Portugal between 2012 and 2022. The data include information on all companies in the sector, disaggregated by type of tourism activity (CAE), specifically the following: (i) Accommodation and Food Services; (ii) Recreational and Cultural Activities; and (iii) Transport and Logistics, as well as the geographic location of tourism companies (by NUT II—North, Algarve, Centre, Lisbon, Setúbal Peninsula, Alentejo, West and Tagus Valley, Azores, and Madeira). Descriptive analyses identified patterns in company and employee profiles. To validate the results, Multivariate Analysis of Variance (MANOVA) analyses were performed, considering year, region, or sector of activity as the dependent variables. To study the differences between the groups, ANOVA tests were carried out, considering the size of the tourist establishments, the tourist region in Portugal and the sector of economic activity.

The study is organised as follows. Section 1 introduces the context, relevance, and objectives. Section 2 reviews the relevant literature, and Section 3 explains the research design, including the context of the materials and methods, as well as the research procedures. Section 4 presents the empirical results, divided into descriptive and inferential analyses, including measures of central tendency, deviations from the mean values, Multivariate Analysis of Variance (MANOVA), distribution of firm size and number of employees by region and sector of activity (ANOVA). Finally, Section 5 concludes with a summary of the contributions of this work, a discussion of the results and conclusions within the theoretical framework, highlighting the limitations and directions for future research.

2. Theoretical Framework

This article seeks to identify trends in the evolution of worker and company profiles responding to the current challenges faced by tourism in Portugal. This issue is critical because problems in labour practices within the sector have worsened recently due to the rise in precarious employment and high staff turnover. A labour system characterised by insecurity, seasonality, and fragmentation [1] is also evident. Working conditions in tourism have been diverging from the ideals of decent work or sustainable employment, which would be their natural and desirable development considering the integration of long-term economic, social, and environmental development goals [1,18,19]. Problems such as low wages, contractual insecurity, discrimination, and serious violations of labour rights are regarded as systemic rather than exceptional, affecting individuals, families, and communities [20,21]. Reflection is both essential and urgent because the economic viability of tourism cannot be separated from social and labour justice, where compliance with human rights and ethical values must be central to the organisation of work [19,22]. Addressing these issues shows governments, employers, and consumers that competitiveness in tourism depends not only on sector growth and technological progress but also on the quality of employment promoted [3]. Several problems are linked to working conditions in the tourism industry. A comparative analysis of the principles inherent in decent work and sustainable employment, alongside the issues identified in tourism working conditions, clarifies this matter and the goal of this research (

Table 1. Theoretical perspectives and working conditions in tourism.

			Working Conditions in Tourism
Authors	Theoretical Perspective	Key Concepts	Strengths of Work in Tourism	Weaknesses of Work in Tourism
[1,2,3,4,7]	Decent Work	Legal and labour rights, equity, security, freedom and adequate remuneration	Absorption of labour, especially young and low-skilled workers. Internationalization and professional mobility. Companies with good CSR and ethical labour practices. Inclusion of vulnerable groups in regulated contexts.	Gender discrimination and occupational segmentation. Low wages and lack of career progression. Failure to meet legal work requirements. High turnover and contractual precariousness.
[5,6,8,9,10]	Sustainable employment	Social sustainability, talent retention, work ethics, social and corporate responsibility, work–life balance and companies that promote social justice

Source: authors’ preparation based on Baum and Hai (2019) [10].

).

Reinforcing this perspective, recent studies have highlighted the importance of the theory of basic human values for new generations working in tourism, arguing that younger workers are unwilling to forgo a work structure based on the following dimensions: instrumental, cognitive, prestige, and affective [23,24]. Understanding the work expectations of this group allows us to identify intergenerational profiles, highlighting the sector’s evolution and providing clues for crafting organisational policies that address the specific needs of this new workforce. The significance of analysing the impact of the principles of decent work and sustainable employment in tourism is connected to the nature of work in this sector, characterised by its labour-intensive nature and significant multiplier effect, creating direct and indirect jobs across multiple related sectors, which is crucial for social inclusion and economic growth [1]. However, the labour instability seen in tourism, worsened by increasing migratory movements for various reasons, has led to a rise in informal contracts of different types, increasing vulnerability to fraud and non-compliance with labour rights. This has resulted in decreased competitiveness in the sector, which has struggled to attract the qualified professionals it needs. Talent is being diverted from socially less valued labour markets, which suffer from poor working conditions, low wages, and a lack of effective human resources policies [18]. Another focus in this field of research is gender inequality. While not exclusive to tourism, it is a prominent issue in this sector, causing additional difficulties in balancing personal and professional life [21]. Portugal also faces this problem, especially regarding the negative pay gap against women, which is even more severe due to the increasing number of women in the labour market.

Quality in tourism and its resulting competitiveness largely depend on the standard of its workforce [5,9]. Since the sector relies heavily on face-to-face interactions between employees and customers, motivation and commitment to the organisation (employee engagement) are vital for the sustainable development of tourism companies [20]. Fostering this work attitude among professionals can boost customer satisfaction and loyalty, which are essential in such a highly competitive global sector. For these reasons, talent management within human resources has become a survival strategy for the sector, and the scientific community has already established a direct link between higher productivity and highly specialised workers [6,7,20].

This brings us back to the core issue of emotional engagement with work, a connection that younger generations are increasingly reluctant to abandon. Engagement involves a two-way relationship between employer and employee, requiring effort, commitment, and shared values [20]. Sustainability in tourism is achieved when various segments are connected, clarified, and combined into a long-term economic, social, and environmental development plan for the region. The tourism workforce includes all workers directly or indirectly involved in business and tourism management activities and is a vital part of this value chain. Salary policies, reward and benefits schemes, recognition of engagement, motivation, and quality of work, as well as training and career development opportunities, are essential for attracting and retaining talent in tourism, ensuring its sustainable competitiveness over time [3,19,22,25,26].

3. Materials and Methods

3.1. Preliminary Framing

This study investigates the structure and working conditions of Portugal’s tourism industry from 2012 to 2022. Statistical data were obtained from the Portuguese Ministry of Labour, Solidarity and Social Security (MTSSS). As part of Portugal’s integration into the European Union, formalised in 1986, the Portuguese government established a legal framework for personnel registers. This was outlined in Decree-Law 479/76 of 16 July and became mandatory for companies with more than 10 employees. These registers have been compulsory since 1985 and serve as the basis for labour regulation policies, as well as the rights and obligations of companies and workers. The Strategy and Planning Office (GEP) of the Portuguese Ministry of Labour, Solidarity, and Social Security collected the analysed data for companies whose main economic activity is tourism (Portuguese Classification of Economic Activities, CAE-REV 3). The data include information on the type of tourism activity (CAE), specifically accommodation and food services, recreational and cultural activities, and transport and logistics, covering the period from 2012 to 2022.

The data are collected in a way that protects personal data. Procedures are in place to ensure the confidentiality and anonymity of company data, including strict information security measures such as encrypted systems for storing and transmitting data, access protocols limited to authorised personnel, and anonymising sensitive information before any disclosure or analysis. These procedures ensure that data collected by the Strategy and Planning Office (GEP) of the Portuguese Ministry of Labour, Solidarity, and Social Security is processed in line with data protection laws, thereby safeguarding the privacy of the companies involved.

3.2. Variables

The main variables are the employment structure and working conditions in the tourism sector from 2012 to 2022, as well as their variation across NUT II regions (NUT II—North, Algarve, Centre, Lisbon, Setúbal Peninsula, Alentejo, West and Tagus Valley, and Azores and Madeira). Specific variables were grouped into broader categories, or analytical dimensions, to organise the information in a clearer and more interpretable manner. Five analytical dimensions were defined:

• Employment structure includes variables that characterise the size and overall composition of employment in the sector (total number of workers, sub-segments of tourism activity such as accommodation and food services; culture and recreation; transport and logistics), and the number of workers per establishment.
• Sociodemographic profile of workers groups the personal and social characteristics of employees, such as gender, age, and education.
• Working conditions focus on the contractual and functional features of employment (type of contract, work regime, seniority, and hours worked).
• Professional qualifications relate to the role and complexity of tasks performed (professional category and qualification level).
• Sector dynamics and evolution encompass the most significant indicators during a given period, including growth in total employment, changes in contract types, and shifts in workers’ age profile and qualifications.

3.3. Sample

Several tables characterise the sample.

Table 2. Sample of tourism establishments in Portugal, by number and size (2012–2022).

Dimension (N = 968)	Variables Codification	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
1–9 people	1	39,721	39,518	40,373	41,128	42,004	42,778	43,571	42,074	42,273	40,652	42,402
10–49 people	2	3983	3999	4330	4810	5225	5901	6536	7036	5965	6285	7485
50–249 people	3	378	375	390	432	509	552	622	654	486	528	695
250–499 people	4	14	13	11	13	17	19	21	20	13	13	19
500–999 people	5	3	4	4	5	4	6	4	4	4	6	5
1000 or more	6	1	1	1	1	1	1	1	1	1	1	1
Total		44,100	43,910	45,109	46,389	47,760	49,257	50,755	49,789	48,742	47,485	50,607

Source: authors’ preparation based on GEP/MTSSS tables.

provides an overview of the tourism sector over the years, based on the number of tourism establishments classified by size.

Table 3. Sample of tourism workers profile in Portugal (2012–2022).

	Variables Codification	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Gender (N = 594)
Men	1	108,390	109,190	114,539	122,644	133,092	144,218	154,447	159,638	140,014	143,027	167,239
Women	2	118,936	118,291	123,667	133,107	142,076	153,620	163,646	168,791	145,932	149,064	172,811
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Age (N = 1970)
Under 25 years old	1	25,039	24,901	27,529	31,570	35,795	41,940	46,048	48,441	34,355	38,608	48,739
25 to 34 years old	2	60,961	59,967	62,169	66,578	71,531	77,636	83,558	87,511	73,405	73,999	90,359
35 to 44 years old	3	61,060	60,919	62,888	66,647	70,162	73,713	77,393	78,549	69,800	68,277	76,721
45 to 54 years old	4	49,904	50,196	51,558	54,053	57,868	61,075	64,401	65,630	61,597	62,482	69,984
55 to 64 years old	5	27,091	28,334	30,315	32,674	34,879	37,880	40,392	41,718	40,476	41,877	46,470
65 or over	6	2922	2884	3458	3919	4574	5156	5791	6464	6271	6750	7709
Unknown	7	349	280	289	310	359	438	510	116	42	98	68
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Employment (N = 1128)
Employer	1	22,773	22,254	22,656	22,974	23,490	24,162	24,977	24,748	24,901	24,150	25,696
Unpaid family worker	2	366	301	316	308	312	300	286	264	207	185	216
Employee	3	203,537	204,235	214,184	231,477	250,435	272,511	291,979	302,399	259,909	266,908	313,258
Active member of a production cooperative	4	397	355	351	369	360	367	344	342	294	244	262
Another situation -	5	253	336	699	623	571	498	507	676	635	604	618
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Education Level (N = 1112)
Up to 3rd cycle of elementary school	1	21,674	21,839	22,425	23,902	25,394	26,383	27,106	26,670	20,426	20,861	24,352
Secondary school	2	8746	9716	10,948	12,576	14,422	16,407	17,984	19,275	15,445	16,780	21,264
Higher education	3	2358	2771	2921	3342	3724	4269	4608	4606	3989	4114	4762
Unknown	4	280	262	284	451	378	401	459	502	359	398	513
Total		33,058	34,588	36,578	40,271	43,918	47,460	50,157	51,053	40,219	42,153	50,891
Qualification level (N = 2376)
Senior managers	1	23,939	23,874	24,508	25,504	26,730	27,670	29,475	30,332	30,758	30,307	33,124
Middle managers	2	10,614	10,206	9989	10,531	10,852	12,052	12,789	12,461	12,050	11,944	13,280
Foremen, masters and team leaders	3	8948	9002	8945	9639	10,540	11,203	11,669	13,601	12,761	13,282	15,535
Highly qualified professionals	4	18,339	18,269	18,947	19,742	21,468	24,171	27,019	27,333	26,047	25,884	29,922
Qualified professionals	5	85,157	84,637	87,438	94,227	101,610	114,852	125,157	125,965	106,978	109,377	124,360
Semi-skilled professionals	6	46,784	48,088	51,524	55,816	60,564	60,234	62,331	63,615	52,016	52,952	63,186
Unqualified professionals	7	18,778	18,903	20,193	21,767	23,608	25,485	27,303	32,647	29,981	31,206	40,183
Trainees, practitioners and translator apprentices	8	14,760	14,480	16,657	18,517	19,788	21,772	22,343	22,471	15,352	17,136	20,459
Unknown	9	7	22	5	8	8	399	7	4	3	3	1
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050

Source: authors’ preparation based on GEP/MTSSS tables.

presents the sociodemographic data of tourism workers in Portugal during the period studied.

Table 4. Sample of the labour situation in the tourism sector in Portugal (2012–2022).

	Variables Codification	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Number of workers (N = 968)
Accommodation and Food Services	1	180,820	180,793	190,525	205,563	221,409	241,687	259,496	268,075	231,474	239,925	281,343
Recreational and Cultural Activities	2	21,062	21,003	20,988	22,216	23,787	25,627	28,058	29,289	25,877	25,633	28,945
Transport and logistics	3	25,444	25,685	26,693	27,972	29,972	30,524	30,539	31,065	28,595	26,533	29,762
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Seniority (N = 1919)
Less than 1 year	1	49,612	56,869	67,604	81,506	90,447	105,140	112,666	118,386	64,450	84,957	133,790
1 to 4 years	2	77,165	68,810	68,969	72,313	82,535	93,313	106,875	113,420	120,372	103,325	97,248
5 to 9 years	3	43,333	42,615	41,304	40,563	38,988	35,859	34,309	34,201	38,548	42,528	48,121
10 to 14 years	4	27,765	28,278	27,609	26,158	24,837	24,414	23,680	22,237	21,928	20,807	19,412
15 to 19 years	5	11,646	12,464	13,723	15,549	17,993	18,317	18,839	17,984	17,064	15,938	15,947
20 and + years	6	17,768	18,408	18,611	19,304	20,000	20,376	21,308	21,752	23,146	24,051	25,057
Unknown	7	37	37	386	358	368	419	416	449	438	485	475
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Profession (N = 2845)
Managers	1	30,308	29,661	29,720	30,469	30,580	31,907	32,562	32,592	32,199	31,369	34,682
Specialists	2	4217	4436	4570	4901	5173	5360	6086	6476	6273	6325	7089
Technicians	3	9121	9111	9381	9790	10,859	11,645	12,522	12,979	12,151	12,344	14,253
Administrative staff	4	24,389	24,777	25,371	27,573	30,000	32,278	33,987	35,223	30,467	29,861	35,032
Personal service	5	101,999	101,744	107,932	116,989	126,443	138,569	148,692	153,429	130,970	136,061	158,168
Farmers	6	1183	1212	1280	1349	1522	1588	1624	1601	1434	1523	1642
Skilled industrial	7	4913	4886	5022	5266	5750	6058	6439	6464	6369	6554	7213
Machine operators	8	11,113	11,225	11,375	11,926	12,280	12,329	12,580	12,593	10,615	10,040	11,194
Unskilled workers	9	40,033	40,317	42,965	47,216	52,167	57,873	63,504	66,945	55,367	57,934	70,656
Unknown	10	50	112	590	272	394	231	97	127	101	80	121
Total		227,326	227,481	238,206	255,751	275,168	297,838	318,093	328,429	285,946	292,091	340,050
Employment contract (N = 881)
Permanent	1	133,717	129,212	128,708	131,883	137,870	142,167	147,720	152,313	155,314	161,871	177,281
Fixed-term	2	67,867	73,107	83,359	97,213	110,038	127,596	141,359	147,637	102,749	103,343	134,142
Other situation	3	1953	1916	2117	2381	2527	2748	2900	2449	1846	1694	1835
Total		203,537	204,235	214,184	231,477	250,435	272,511	291,979	302,399	259,909	266,908	313,258
Working hours (N = 595)
Full-time	1	186,585	185,751	193,636	207,967	225,047	244,517	261,180	270,033	233,077	238,204	279,131
Part-time	2	16,952	18,484	20,548	23,510	25,388	27,994	30,799	32,366	26,832	28,704	34,127
Total		203,537	204,235	214,184	231,477	250,435	272,511	291,979	302,399	259,909	266,908	313,258

Source: authors’ preparation based on GEP/MTSSS tables.

describes the working conditions in this sector. The coding of the variables across the tables is also indicated to assist in reading the results.

Table 2 details the distribution of tourism establishments in Portugal between 2012 and 2022, according to their size, defined by the number of employees. During this period, there was an overall growth in the number of establishments, from 44,100 in 2012 to 50,607 in 2022, an increase of approximately 14.7%.

Smaller establishments (1–9 people) constitute the majority throughout the period, although their relative share tends to decline slightly. In 2012, they accounted for approximately 90.1% of the total, decreasing to about 83.8% in 2022, despite an absolute increase of 681 units.

Conversely, medium-sized establishments (10–49 people) show significant growth, rising from 3983 units in 2012 to 7485 in 2022 (an increase of 88%). This growth may indicate a trend towards professionalisation and sector consolidation.

Establishments with 50–249 employees also show notable growth, increasing from 378 to 695 units over the period, an increase of 83.8%. Larger establishments (250 or more staff) remain marginal, with no major structural changes over the years, maintaining a stable presence and limited representation in total figures. In summary, the data reveal an expanding tourism sector in Portugal, with particular emphasis on the growth of medium-sized establishments, suggesting a trend towards business strengthening and a potential increase in capacity to meet tourism market demands.

Table 3 reveals that between 2012 and 2022, the number of workers in Portugal’s tourism sector grew by 49.5%, increasing from 227,326 to 340,050 individuals. This growth was more pronounced among women, who consistently make up the majority of the workforce, reaching 50.8% of the total in 2022.

The age distribution shows a gradual rejuvenation until 2019, followed by a decline in 2020. In 2022, notable growth is seen in the 25–34 age group, which accounts for approximately 26.6% of the total.

Regarding employment status, most workers are salaried (92.1% in 2022), maintaining stability throughout the period. The number of employers showed moderate growth (13.3%), while unpaid and volunteer workers remained a marginal presence.

In terms of education, there was a steady increase in qualification levels, with a 101.9% rise in the number of workers with higher education between 2012 and 2022. Concurrently, qualified and highly qualified professionals represented over 45% of the workforce in 2022, indicating a rising trend towards specialisation and higher qualification within the sector.

Table 4 summarises a series of data characterising working conditions in tourism between 2012 and 2022. During this period, the number of workers in the tourism sector in Portugal increased from 227,326 to 340,050, representing a 49.5% rise. The highest concentration of employment is observed in accommodation and food services, which in 2022 accounted for approximately 82.7% of the total. This is followed by recreational and cultural activities, as well as transport and logistics, with significantly lower shares.

The duration of employment reveals a high turnover. Specifically, in 2022, 39.3% of workers had been in their jobs for less than a year, a figure that nearly tripled compared to 2012. Conversely, relationships lasting more than 20 years remain minimal (7.4%).

Regarding professional classification, workers in personal services, security, and sales form the majority (46.5% in 2022), followed by unskilled workers (20.8%). Despite this, there is a steady growth in the number of technicians, specialists, and managers, indicating a slow upskilling of the sector.

The contractual framework is dominated by indefinite-term contracts, although fixed-term contracts have increased substantially (from 67,867 in 2012 to 134,142 in 2022, +97.7%), reinforcing the precariousness of work. In terms of working hours, employment remains primarily full-time (89.1% in 2022), with stability maintained throughout the decade.

These data point to a quantitative expansion of tourism employment, characterised by a high turnover, rising precariousness, and a predominance of operational roles, alongside signs of increasing professionalisation.

3.4. Data Analysis

Descriptive analyses were conducted by calculating means and deviations from these means to identify patterns and trends in data variability and consistency related to company and employee profiles [1,23,25]. Multivariate Analysis of Variance (MANOVA) was used to validate the descriptive statistics results, determining if the dependent variables vary with the year, region, or sector of activity and whether these differences are statistically significant [20]. Additionally, ANOVA tests were employed to examine statistically significant differences between groups based on the number of workers, the size of the tourism establishments, the region (NUT II), and sector of activity [12,13,14].

Pearson correlation tests (r) were also performed to measure the degree of linear association between two continuous or ordinal variables within the accommodation and food sector, following these criteria: a correlation coefficient ranging from −1 (perfect negative correlation) to +1 (perfect positive correlation), with statistical significance at the levels of 0.01 (99% confidence) and 0.05 (95% confidence).

4. Results and Discussion

4.1. Profile Identification Based on Data Variability

Descriptive analysis of the variables (

Table 5. Descriptive analysis: mean (μ), variance (s²), maximum and minimum value.

Variables	N	Min (x)	Max (x)	Mean (µ)	Standard Deviation (σ)	Sample Variance (s2)
Profession	2845	1	10	5.33	2.793	7.8
Qualifications level	2376	1	8	4.63	2.288	5.237
Worker age	1970	1	7	3.83	1.923	3.698
Seniority	1919	1	7	3.75	1.877	3.522
Employment	1128	1	5	2.90	1.497	2.242
Educational qualifications	1112	1	4	2.40	1.083	1.172
Year	968	2012	2022	2017	3.158	9.976
Regions of Portugal (NUT II)	968	1	9	4.96	2.588	6.7
Sector of Activity (CAE)	968	1	3	1.98	0.815	0.664
Number of workers	968	1	41,105	1084.84	3569.378	12,740,461.083
Establishments dimension	968	1	6	2.26	1.134	1.286
Employment contract	881	1	3	1.99	0.815	0.664
Nationality	670	1	3	1.67	0.670	0.449
Working hours	594	1	2	1.5	0.50	0.25

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

) allows us to identify four typical profiles based on data dispersion and mean values. These profiles group the variables differently from in Section 3.2 (Variables), particularly in how they cluster into dimensions.

The year variable (2012–2022), with a mean in 2017 [mean (μ) = 2017] and low data dispersion [standard deviation (σ) = 3.158], indicates a balanced distribution of the different profiles over time, supporting the representativeness of the sample and its suitability for longitudinal comparisons or trend analysis. The main groupings are:

• Profile 1—Common Profile: composed of variables with low variability, indicating homogeneous behaviours and characteristics such as: (i) employment contract stability [mean (μ) < 2; low standard deviation (σ) = 0.815]; (ii) workers mainly of Portuguese nationality; (iii) a nearly binary split between full-time and part-time work, with a fairly even distribution; and (iv) concentration of work in one or two tourism activity sectors (CAE).
• Profile 2—Regional and Functional Dispersion: This profile combines variables such as profession and region (NUT II). These variables show wide data dispersion [standard deviation (σ) > 2.2.5], with a broad geographic spread across all regions of Portugal and significant variability at the profession level. It appears that workers are distributed across multiple fields of activity, levels of specialisation, regions, and functional areas, without concentration in a single work area or region, which may relate to the geographic distribution of sectors with varying economic and hiring capacities.
• Profile 3—Transitional Professional Trajectory: This group includes workers at different stages of their careers, reflecting mobility and progression within the labour market. The variables analysed are worker age [mean (μ) = 3.83], seniority [mean (μ) = 3.75], and qualification level [mean (μ) = 4.63], with moderate variability suggesting heterogeneity among both younger and more experienced workers.
• Profile 4—Contrasting Business Situations: based on variables related to employers, establishment size, and number of workers, where variability indicates two distinct economic scenarios: (i) SMEs (Small and Medium-Sized Enterprises), with fewer staff, a smaller structure, and lower labour absorption capacity; and (ii) large companies, with stronger structures and larger workforces.

4.2. Multivariate Analysis of Variance (MANOVA)

A Multivariate Analysis of Variance (MANOVA) examined the main effects and interactions of the factors of Year, Region, and Sector on the dependent variables (profession, qualification level, workers’ age, seniority, employment, educational qualifications, number of workers, establishment size, employment contract, nationality, and working hours). To evaluate the statistical significance of the effects, the multivariate tests of Pillai’s Trace, Wilks’ Lambda, Hotelling’s Trace, and Roy’s Largest Root were calculated. Only the Roy Largest Root test produced results close to the significance level (p = 0.021) (

Table 6. Results of multivariate tests for main effects and interactions.

Effect	Pillai’s Trace	Wilks’ Lambda	Hotelling’s Trace	Roy’s Largest Root	Sig. (p)	η2 Partial	Significant?
Intercept	0.962	0.038	25.435	25.435	<0.001	0.962	✔ Yes
Year	0.010	0.990	0.010	0.007	1.000	0.005	✘ No
Region	0.122	0.882	0.129	0.075	≥0.876	0.015	✘ No
Sector	0.118	0.884	0.129	0.108	<0.001	0.059–0.098	✔ Yes
Year × Region	0.043	0.957	0.044	0.020	1.000	0.004–0.020	✘ No
Year × Sector	0.032	0.968	0.033	0.019	1.000	0.008–0.019	✘ No
Region × Sector	0.143	0.864	0.149	0.058	1.000	0.014–0.055	✔ Yes
Year × Reg × Sector	0.104	0.900	0.107	0.049	1.000	0.010–0.047	✘ No

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

).

The results showed a statistically significant multivariate effect for the Sector factor, as indicated by all tests (p < 0.001). The effect size (partial η²) ranged from 0.059 to 0.098, indicating a small to moderate impact on the dependent variables. The intercept was also statistically significant (p < 0.001), with a very large effect (partial η² = 0.962), which was expected, as it reflects the overall mean of the groups. The classification of the effect size (partial η²) followed these guidelines: p < 0.01 = very small; p between 0.01 and 0.06 = small; p between 0.06 and 0.14 = moderate; p > 0.14 = large (Table 6).

In contrast, no statistically significant effects were observed for the Year and Region factors, nor for their interactions with Sector. The only partial exception was the “Region × Sector” interaction, which showed significance in the Roy Largest Root Mean (p = 0.021), although the other tests did not confirm this result. Even in this case, the effect size was modest (partial η² = 0.055). This analysis indicates that, among the factors examined, only Sector had a significant influence on the dependent variables, with the other variables and interactions being insignificant within the multivariate context.

The results of the adjusted model, which explains the previous data, are presented below. The adjusted model revealed significant effects only for the variable “Number of workers,” with a p-value < 0.001 and a considerable effect size (partial η² = 0.210), suggesting that the factors under review explain a substantial portion of the variability observed in this variable. The other variables did not show statistical significance in the overall model (p > 0.05), presenting small effect sizes (η² < 0.06), which indicates that, overall, the model has limited explanatory power for these dimensions (

Table 7. Summary of Significant Effects (p < 0.05).

Factor/Interaction	Dependent Variable	p-Value	Partial Eta Squared (η2)	Interpretation
Adjusted Model	Number of people	<0.001	0.210	Strong overall effect
Region	Number of people	<0.001	0.063	Regional differences in employment
Region	Size of tourism establishments	0.016	0.035	Regional differences in the size of establishments
Sector	Number of people	<0.001	0.084	Sectoral differences in employment
Sector	Size of tourism establishments	0.017	0.016	Small sectoral effect
Region × Setor	Number of people	0.032	0.052	Significant interaction: sector varies by region

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

).

Table 7 summarises the most significant effects among the variables, noting that:

• The Year variable had no statistically significant effects on any of the dependent variables (all p-values > 0.05), with very low partial eta-squared values (η² ≤ 0.01). These results indicate that, during the analysed period, there were no notable changes in labour variables as a function of year.
• The Region factor demonstrated a statistically significant effect on the number of employed persons (p < 0.001; η² = 0.063) and on the size of tourism establishments (p = 0.016; η² = 0.035), suggesting some geographic variation in employment structures and establishment characteristics. The other variables showed no significant differences between regions.
• The Sector item showed notable differences in two variables: (i) Number of workers (p < 0.001; η² = 0.084), with a moderate effect; (ii) Size of tourism establishments (p = 0.017; η² = 0.016), with a small effect. These findings indicate that various tourism sub-sectors (e.g., lodging, restaurants, travel agencies, and transport) vary in terms of the number of workers and the size of their establishments.
• The interaction between ‘Region × Sector of Activity’ was statistically significant only for the number of workers (p = 0.032; η² = 0.052), indicating that employment distribution across tourism sectors differs by region.
• None of the other interactions (Year × Region, Year × Sector, or Year × Region × Sector) showed statistical significance (p > 0.05), with marginal or null effects (η² ≤ 0.01), indicating considerable stability over time and the absence of complex interactions between the factors analysed.
• Regarding effect sizes (partial η²), the highest values were observed for: (i) the intercept across all variables, with η² > 0.65 (p < 0.001), indicating the extent of variability explained by the overall mean; (ii) the variable “Number of workers,” which is significantly explained by the factors and interactions (η² ranging from 0.05 to 0.21). The other variables showed low or negligible effect sizes (η² < 0.05), emphasising that the impact of the studied factors is limited to a few specific dimensions.

Analysis of the between-subject effects indicates that only a few variables—specifically, the number of employees and the size of tourism establishments—are significantly influenced by Region and Sector. Year and most of its interactions did not show statistically significant effects. Region and sector have notable effects on just a few variables: (a) number of workers (Region: partial η² = 0.063; Sector: partial η² = 0.084); and (b) size of tourism establishments (Region: partial η² = 0.035; Sector: partial η² = 0.016). Most other variables (age of workers, type of contract, working hours, gender, among others) did not demonstrate significant or relevant effects (partial η² < 0.01 or p > 0.05). Regarding the coefficient of determination of the individual models (R²), it is clear that the R² values adjusted for several variables are low or negative, indicating a poor fit of the models, with examples such as: R² adjusted for Profession = –0.104; R² adjusted for Type of Employment = –0.094; R² adjusted for Employment Contract = –0.110.

4.3. Distribution of Company Size and Number of Employees by Region and Sector of Activity (ANOVA)

The previous findings showed that it was not possible to establish a relationship between region, sector of activity, and sociodemographic variables (qualification level, worker age, employment, educational qualifications, nationality, gender) or working conditions (profession, seniority, number of employees, establishment size, employment contract, and working hours). The only statistically significant associations are between regions of Portugal and the tourism sector, specifically concerning the number of employees and establishment size. These findings justify conducting ANOVA tests, focusing on the relationships between these variables.

Comparative Analysis of the Number of Workers by Region of Portugal and Sector of Activity

Data on the number of workers distributed across nine regions of Portugal were analysed: North, Centre, Lisbon, Setúbal Peninsula, Alentejo, Algarve, West and Tagus Valley, Azores, and Madeira, with a total of 2845 observations. A similar analysis was conducted to examine the relationship between the number of workers and the economic activity segments of the tourism sector (accommodation and food services, recreational and cultural activities, and transport and logistics).

Homogeneity of Variance Test

Levene’s test was used to verify the assumption of homogeneity of variances between regional groups regarding the number of workers. Regardless of the method employed (mean, median, adjusted median, or trimmed mean), the test showed significance values below 0.001. This indicates that the variances in the number of workers differ significantly between regions, suggesting that the data dispersion around the regression line is not uniform (heteroscedasticity). This factor should be kept in mind in subsequent analyses, as it can impact the validity of standard parametric tests. The same analysis, performed for the activity sector variable, confirmed the assumption that the error variance is constant, which is crucial for the validity of statistical tests such as regression analysis (homoscedasticity). The comparison results between groups were highly significant across all approaches (based on mean, median, trimmed mean, and median with adjusted degrees of freedom), with p-values < 0.001.

Analysis of Variance (ANOVA)

The model generated by the ANOVA revealed a statistically significant difference in the average number of workers between the regions analysed [F (8,2836) = 22.640, p < 0.001]. The sum of squares between the groups was significant (2,175,129.136), indicating that a substantial part of the total variation in the variable can be attributed to regional differences. Furthermore, the tests for the linear term also showed high significance (p < 0.001), confirming a systematic trend across the regions. For the activity sector, the ANOVA also indicated a significant effect on the number of workers [F (2,2842) = 141.981, p < 0.001]. The sum of squares between the groups (SQ = 3,291,480,881.06) accounts for a considerable portion of the total variance observed. The linear and quadratic contrasts were also highly significant, suggesting a non-linear trend across the evaluated sectors. This result confirms that the average number of workers varies significantly across the three sectors analysed.

Effect Size

The effect size, indicated by the estimates of Eta-squared (η² = 0.060) and Epsilon-squared (ε² = 0.057), shows that approximately 5.7% to 6% of the variability in the number of workers can be explained by regionality. Although this effect is statistically significant, it is moderate, suggesting that factors other than regionality also influence the number of workers. Omega-squared (ω²), which is regarded as a more reliable estimate, supports this conclusion, with fixed values close to ω² = 0.057 and random values near ω² = 0.008. For the relationship between sector and the number of workers employed, the coefficients of determination remain consistent and indicate moderate to large effects, namely: Eta-squared (η²) = 0.091; Epsilon-squared (ε²) = 0.090; Fixed Omega Squared (ω²) = 0.090; and Random Omega Squared (ω²) = 0.047.

The η² = 0.091 estimate indicates that 9.1% of the variability in the number of workers can be explained by the activity sector, representing a significant effect. This is important because organisational variables often have low coefficients of determination. The more conservative value of random Omega (ω²) suggests that, in generalisable contexts, the effect size may be smaller but still significant.

Multiple Comparisons and Homogeneous Subsets

Tukey HSD post hoc analyses revealed significant differences between several regions, particularly in the North, Setúbal, Alentejo, Azores, and Madeira, which had statistically different means compared to each other and to other regions such as Lisbon and the Algarve. For example, the North region had a significantly higher mean than the Central region (mean difference = 1353.42, p < 0.001) and the Algarve (mean difference = 852.27, p = 0.046).

Three distinct groups are observed in the distribution of the number of workers across regions: (i) a group with the North and Lisbon regions, characterised by the highest means; (ii) an intermediate group formed by the Algarve, Central region, and other regions; and (iii) a group with lower means, comprising the Azores and Alentejo. The marginal significance (p = 0.052) for group three indicates some proximity between the intermediate and highest groups.

Multiple comparisons within the activity sector, carried out using the Tukey HSD and DMS methods, identified statistically significant differences between Accommodation and Food Services and the other sectors, as follows:

• Accommodation and Food Services versus Recreational and Cultural Activities shows a mean difference in the number of workers of 2296.4 (p < 0.001).
• Accommodation and Food Services versus Transport and Logistics shows a mean difference in the number of workers of 2240.8 (p < 0.001).
• Recreational and Cultural Activities versus Transport and Logistics reveal a mean difference of 55.6 workers (p > 0.9, non-significant p-value).

These results demonstrate that the Accommodation and Food Services sector employs, on average, a significantly higher number of workers than the other two activity sectors, which, statistically, do not differ significantly.

Comparative analyses were then conducted between the size of the tourism establishment, the region, and the activity sector. ANOVA tests compared the size of tourism establishments across nine regions of Portugal: North, Centre, Lisbon, Setúbal Peninsula, Alentejo, Algarve, West and Tagus Valley, Azores, and Madeira, with a total of 968 observations. A similar procedure was used to analyse the relationship between the size of tourism establishments and the sector of activity (accommodation and food services, recreational and cultural activities, and transport and logistics).

Homogeneity of Variance Test

Levene’s homogeneity tests indicated that the comparisons between groups (for Region and Sector) all produced statistically significant results (p < 0.001), confirming that the variances differ significantly. This was consistent across all Levene’s test criteria (based on the mean, median, and trimmed mean), making the conclusion more robust. The same applies to the study of how the size of tourism establishments affects the activity sector. Significant differences in variances between groups were found, with all p-values < 0.05, indicating that the variances of the three tourism activity groups are not homogeneous. Given these results, we chose to use a Welch ANOVA, which is more suitable when variances differ between groups (heteroscedasticity), as the assumptions of traditional ANOVA were not met—that is, the assumption of equal variances was violated.

Welch’s ANOVA also identified a significant difference in group means (p < 0.001 for regional comparisons; p < 0.006 for activity sector), consistent with the findings from standard ANOVA. These results suggest that the size of tourism establishments does not depend on either the region or the sector of activity.

The findings reveal notable regional differences in worker numbers, with more economically developed regions, such as Lisbon and the North, showing higher labour concentrations. The variation in the number of workers across regions likely reflects structural and economic differences, as well as the concentration of companies within the hotel and restaurant sectors.

Multiple comparisons within the activity sector also revealed statistically significant differences between accommodation and food services and the other sectors, with the latter employing the highest number of workers. These data confirm the results of descriptive statistics, which suggest the existence of contrasting business situations (Profile 4), characterised by the relationship between the size of the tourism establishment and the number of employees. For these reasons, and although the ANOVA tests did not produce significant results regarding the distribution of variables by activity sector, a Pearson correlation (P) was used between hotel and catering companies and the other variables (profession, number of employees, qualification level, age, seniority, type of contract, nationality, type of employment, size of tourism establishments, working hours, gender, and educational qualifications). The accommodation and food services sector employs, on average, a significantly larger number of workers than the other two sectors, which statistically do not differ significantly.

4.4. Relationship Between the Accommodation and Food Services Sector and Working Conditions

A Pearson correlation (r) was conducted between the various variables and the accommodation and food services sector. The correlation matrix is shown in

Table 8. Most significant variables in the accommodation and food services sector.

	Sector = 1 (FILTER)	Hours	Gender	Nationality	Profession	Number	Qualification	Age	Seniority	Contract	Employment	Dimension	Education
Sector = 1 (FILTER)	1	0.007	0.007	0.035	0.019	0.301 **	0.003	−0.007	−0.004	−0.001	0.004	−0.087 **	−0.007
Working hours		1	1.000 **	0.092 *	0.001	−0.035	−0.001	0.009	−0.016	−0.002	−0.019	−0.016	0.122 **
Gender			1	0.092 *	0.001	−0.035	−0.001	0.009	−0.016	−0.002	−0.019	−0.016	0.122 **
Nationality				1	0.023	−0.010	0.002	−0.037	−0.002	−0.044	0.016	−0.001	0.001
Profession					1	−0.008	−0.001	0.001	−0.011	0.009	0.033	−0.025	−0.014
Number of workers						1	−0.008	0.022	−0.027	0.008	−0.056	−0.034	−0.042
Qualification Level							1	−0.009	0.002	0.008	−0.017	0.020	−0.037
Workers’ age								1	−0.022	0.000	−0.011	0.032	−0.011
Seniority									1	0.029	−0.026	−0.019	−0.013
Employment Contract										1	0.003	−0.023	−0.005
Type of employment											1	0.004	−0.033
Dimension of tourist establishments												1	0.009
Educational qualifications													1

Source: authors’ preparation based on SPSS output (version 28.0.0.0). Note: * = p < 0.05; ** = p < 0.01.

. Associations between variables were recognised if their strength, direction, and significance (r) met the criteria of p < 0.05 * (significant at the 5% level) and p < 0.01 ** (significant at the 1% level).

The significant (statistically relevant) correlations standing out in the work structure and profile of workers in this sector are:

Number of Workers vs. Sector (r = 0.301 **), which indicates a strong, positive, and significant correlation, suggesting that establishments with more workers tend to be more represented in the accommodation and food services sector;

Dimension vs. Sector (r = 0.087 **), presents a negative, weak, but significant association, inferring that larger establishments are less likely to be part of this sector;

Working Hours vs. Education (r = 0. 122 **), reveals a weak, positive, and significant correlation and may indicate that higher levels of education are associated with a greater workload. This result should be interpreted with caution.

Gender vs. Education (r = 0.122 **), similar to the previous association, also shows significant connections between these two variables, although the correlation is weak, so any inference that one gender has more education than the other should be made cautiously.

Working Hours vs. Nationality (r = 0.092 *), the correlation, although weak, is significant and possibly indicates that workers’ nationality may influence their working hours.

These findings reveal some significant associations, notably between the number of workers and the accommodation and food services sector, suggesting that this is where the majority of people working in tourism find employment. However, the largest companies are not part of this sector.

The correlations between education, nationality, gender, and working hours are modest but statistically significant, indicating potential inequalities or structural patterns worth further investigation. The lack of strong overall correlations suggests that the variables studied are relatively independent within this tourism sector.

To understand the relationship between the categorical variables of education, nationality, gender, and working hours, cross-tabulation explored patterns of association, dependence, or independence between these variables, as identified in the previous analysis.

Cross-tabulation of educational qualifications and working hours

Analysis of the relationship between educational qualifications and working hours (full-time or part-time), followed by application of the chi-square test, aimed to determine whether there is a statistically significant association between these two categorical variables. The distribution of educational qualification categories by working hours shows distinct patterns (Table 9):

• Workers with neither high school education nor higher qualifications have a higher proportion of full-time work (62.3%) compared to part-time work (37.7%);
• Conversely, workers with only a high school education have a higher proportion of part-time work (62.3%) than full-time.
• Individuals with higher education follow a similar pattern to the first group, with 62.3% working full-time and 37.7% working part-time;
• Lastly, the group with an unknown education level presents the highest proportion of part-time employment (66.7%).

Table 9. Cross tabulation between educational qualifications × working hours.

			Working Hours		Total
			Full-Time	Part-Time
Educational qualifications	Up to the 3rd cycle of elementary school	N	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Secondary school	N	60	99	159
		% in Educational qualifications	37.7%	62.3%	100.0%
	Higher education	N	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Unknown	N	39	78	117
		% in Educational qualifications	33.3%	66.7%	100.0%
Total		N	297	297	594
		% in Educational qualifications	50.0%	50.0%	100.0%

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

Table 9. Cross tabulation between educational qualifications × working hours.

			Working Hours		Total
			Full-Time	Part-Time
Educational qualifications	Up to the 3rd cycle of elementary school	N	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Secondary school	N	60	99	159
		% in Educational qualifications	37.7%	62.3%	100.0%
	Higher education	N	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Unknown	N	39	78	117
		% in Educational qualifications	33.3%	66.7%	100.0%
Total		N	297	297	594
		% in Educational qualifications	50.0%	50.0%	100.0%

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

Chi-square tests confirm a statistically significant association between educational qualifications and work schedule (χ² = 41.698; df = 3; p < 0.001). This pattern may reflect labour market dynamics or the profile of the sectors in which these groups operate. For instance, a high school diploma could be connected to roles with greater flexibility or instability, whereas higher education is usually required for more stable, full-time contractual jobs.

Cross-tabulation between educational qualifications and gender

The objective of this analysis is to determine whether there is a statistically significant relationship between individuals’ levels of academic qualifications and gender (male or female). The total sample comprised 594 participants, equally divided between men (n = 297) and women (n = 297).

A clear pattern emerged between the groups, with the majority of individuals with primary or higher education being men (62.3%), and the majority of individuals with secondary education (62.3%) or unknown qualifications (66.7%) being women. It has already been confirmed that workers with this level of education tend to hold part-time jobs, leading to the conclusion that, in the accommodation and hospitality sector, women face greater job instability and generally have lower levels of education (

Table 10. Cross tabulation between educational qualifications × gender.

			Gender		Total
			Man	Woman
Educational qualifications	Up to the 3rd cycle of elementary school	N	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Secondary school	N	60	99	159
		% in Educational qualifications	37.7%	62.3%	100.0%
	Higher education	Contagem	99	60	159
		% in Educational qualifications	62.3%	37.7%	100.0%
	Unknown	N	39	78	117
		% in Educational qualifications	33.3%	66.7%	100.0%
Total		N	297	297	594
		% in Educational qualifications	50.0%	50.0%	100.0%

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

).

Similarly, the chi-square test indicates a statistically significant association between gender and level of academic qualification (χ² = 41.698; df = 3; p < 0.001). The assumption that no cell in the table had an expected frequency below 5 (N < 5) was met, with an expected count of 58,50. Additionally, the significant linear-by-linear association (p = 0.003) suggests a systematic trend, with data potentially polarising between genders at certain education levels.

The findings highlight notable differences between men and women regarding reported education levels:

• Men are overrepresented at the extremes of the educational spectrum (elementary and higher education), which could reflect two distinct profiles: a group with less education, potentially working in unskilled roles, and a more qualified group with greater access to management and leadership positions;
• Women tend to be concentrated mainly in secondary education and among the group with unknown qualifications. The predominance in secondary education may indicate interrupted educational pathways or a greater focus on sectors requiring this level of education. The high percentage of women in the “unknown” group might suggest gaps in administrative records or barriers to formal recognition of qualifications.

These results could have significant implications for shaping public education and gender equality policies, such as:

• Enhancing skills recognition and validation programmes for populations with unrecorded qualifications, especially women;
• Promoting women’s access to higher education;
• Tailoring training programmes to gender-specific needs to promote equitable participation in the labour market.

These data underscore the importance of considering the intersection of gender and education in strategies aimed at reducing social inequalities and fostering equity in education and employment within the accommodation and food services sector.

Cross-tabulation between nationality and working hours

This analysis explores the relationship between workers’ nationality and their working hours. There is a noticeable variation in working hours based on nationality (Table 11):

• Among Portuguese workers, full-time employment is most common (57.3%);
• Among foreign workers, part-time employment is more common (57.5%);
• Among workers of unknown nationality, the distribution of working hours is almost evenly split (full-time = 50.7%; part-time = 49.3%).

Table 11. Cross tabulation between nationality × working hours.

			Working Hours		Total
			Full-Time	Part-Time
Nationality	Nationals	N	150	112	262
		% in Nationality	57.3%	42.7%	100.0%
	Foreigners	N	111	150	261
		% in Nationality	42.5%	57.5%	100.0%
	Unknown	N	36	35	71
		% in Nationality	50.7%	49.3%	100.0%
Total		N	297	297	594
		% in Nationality	50.0%	50.0%	100.0%

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

Table 11. Cross tabulation between nationality × working hours.

			Working Hours		Total
			Full-Time	Part-Time
Nationality	Nationals	N	150	112	262
		% in Nationality	57.3%	42.7%	100.0%
	Foreigners	N	111	150	261
		% in Nationality	42.5%	57.5%	100.0%
	Unknown	N	36	35	71
		% in Nationality	50.7%	49.3%	100.0%
Total		N	297	297	594
		% in Nationality	50.0%	50.0%	100.0%

Source: authors’ preparation based on SPSS output (version 28.0.0.0).

Chi-square tests meet the assumptions that all cells have expected counts greater than 5, with the minimum expected count at 35.50, above the statistically acceptable threshold. Values of p = 0.003 indicate a statistically significant association between nationality and working hours. The linear-by-linear association (p = 0.025) also demonstrates a systematic trend between the variables, showing a steady increase in part-time work as one moves from national to foreign workers.

The data suggest that holding a foreign nationality is linked to a higher likelihood of working part-time. This may be related to:

• Language or legal barriers that hinder foreigners’ access to full-time contracts;
• Labour market segmentation, where foreigners are often directed into less stable positions;
• The flexibility required by these workers, often in situations of precariousness or high turnover;
• The possible presence of foreign students or temporary migrants, whose work schedules are limited.

Conversely, the prevalence of full-time employment among national workers may reflect greater structural integration within the labour market, as well as better access to contractual rights and job stability. It is essential to develop inclusive policies to ensure equal access for foreigners to full-time contracts, combat market segmentation, invest in socioeconomic integration programmes (such as language training and skills validation), strengthen enforcement of labour legislation in part-time situations, and ensure that this is not a regime imposed due to a lack of alternatives. Additionally, it is important to consider these differences in human resource planning and employment equity monitoring.

5. Conclusions

Tourism is a strategic sector that creates jobs and promotes economic and social development. Human resources are crucial in supporting the competitiveness of tourism firms and driving the industry’s growth. Portuguese tourism-labour research shows that, despite the sector’s economic weight, work remains highly seasonal and precarious (including in platform-mediated accommodation), with persistent gender inequalities and work–life imbalance; these dynamics depress wages, heighten insecurity (especially post-COVID in hotspots like the Algarve), and erode the “decent work” dimension of social sustainability—hence recent calls for stronger retention practices and policy steering via subnational sustainability indicators to align tourism growth with sustainable development outcomes [17,19,24,25].

Given the significance of this topic, this research aims to examine further employment in the tourism sector, its development from 2012 to 2022, and explore the connection between profiles, qualifications, and working conditions, as well as inequalities across economic sectors, different company types, and regions in Portugal.

The empirical results of the study enable us to draw several relevant conclusions. Firstly, based on the data examined, between 2012 and 2022, the number of tourism establishments in Portugal increased by approximately 14.7%, rising from 44,100 establishments in 2012 to 50,607 in 2022. The business sector is predominantly characterised by smaller companies, accounting for 90.1% of firms in 2012 and 83.8% in 2022, with an increase in medium-sized companies within the tourism sector, which grew by 88%, indicating consolidation and expansion of the sector, similar to findings in existing literature [6]. The number of workers in Portugal’s tourism sector grew by 49.5%, the majority of whom were female. The evolution of the age distribution indicates a rejuvenation of the sector’s workforce. In terms of education, the findings point towards increased qualification and professionalisation, with the number of workers with higher education qualifications doubling between 2012 and 2022.

Regarding the duration of employment contracts, similar to the findings in the literature, the length of these contracts reflects a high turnover, with most contracts lasting less than a year. This trend grew between 2012 and 2022, with the number of workers employed for less than a year nearly tripling, while contracts lasting more than 20 years remained low. Furthermore, although indefinite-term contracts dominate the contractual framework, fixed-term contracts have increased significantly, reinforcing the level of job insecurity.

The results of the descriptive analysis identified four worker profiles. Profile one is considered a common profile, characterised by employment contract stability and predominantly Portuguese nationality, although it should be noted that the nationality of part of the sample is unknown. Profile two, regional and functional dispersion, combines variables such as profession and region (NUT II), and shows extensive geographic spread across all regions of Portugal, along with significant variability at the profession level. This indicates that workers are spread across multiple areas of activity, levels of specialisation, regions, and functional sectors, without any concentration in a single work area or region. Profile three, known as the transitional professional trajectory, reflects diversity between younger and more experienced workers in terms of careers. Finally, profile four, contrasting business situations, highlights two distinct economic scenarios: on one hand, SMEs with smaller workforces, limited structures, and lower labour absorption capacity; and on the other, large companies with more extensive structures and larger workforces.

Analysis of the effects and interactions of the Year, Region, and Sector factors on the dependent variables, such as employee profile, profession, contract, and working hours, reveals differentiated effects. The Sector factor shows results with statistically significant effects on the dependent variables. The Region and Sector factors demonstrated notable effects on only a few variables, such as the number of workers and the size of tourism establishments.

The data related to the sector indicate that the accommodation and food services sector employs, on average, a significantly larger number of workers than the other two sectors, where there is no statistically significant difference. The correlations between education, nationality, gender, and working hours, although moderate, are statistically significant, highlighting inequalities or structural patterns that warrant further investigation.

Analysis of the relationship between educational qualifications and the type of workday—full or part-time—reveals varying results across profiles. Workers without a high school diploma or higher education qualification are more likely to have full-time work. Conversely, workers with only a high school diploma are more likely to work part-time. Workers with a college degree show a higher tendency for full-time employment, and the group with unknown education levels has the highest proportion of part-time work.

Finally, the analysis of educational qualifications and gender revealed a pattern across groups, where most individuals with a basic diploma or higher education degree are male, and most with a high school diploma or unknown qualifications are female. Concerning nationality, a higher prevalence of full-time employment was observed among national workers, which may reflect greater structural integration into the labour market.

This study, therefore, contributes to the debate on employment in the tourism sector. It concludes that implementing policies to unlock the sector’s potential is essential. Such policies should promote the creation of more and better jobs, with greater stability, equality, improved working conditions, and a strong capacity to attract and retain talent, thereby supporting regions’ sustainable development. This study also contributes to the discussion on the importance of tourism in regional development, and fundamentally, its great importance in reducing regional inequalities. The empirical results support a critical view of the literature, contributing information to the sector and to all conclusions that identify the need for better jobs and equality, but could further specify practical recommendations for policymakers, industry, and education in the tourism and hospitality sector.

This study’s limitations include the use of only a quantitative method. Although it features a regionally disaggregated analysis, variables related to territorial indicators and regional tourism data were not included.

Future research could adopt a mixed-methods approach, combining quantitative techniques with qualitative insights by including the perceptions of workers, business owners, or local decision-makers. It would also be beneficial to incorporate territorial variables, allowing the analysis of how working conditions influence tourism indicators at a regional level, thus enabling a more comprehensive discussion of the sector and the factors that affect it.