Background and context
The global health crisis has presented the Gulf states with unparalleled public health and economic stability challenges, prompting swift and comprehensive policy actions. In contrast to the extensive focus on COVID-19 responses in the Global North, less attention has been directed toward the Gulf states (Aspalter, Reference Aspalter2023; Jawad and Gal, Reference Jawad, Gal and Greve2019). This study fills that gap with a comprehensive description of the social protection policies enacted after the onset of the pandemic. Moreover, this study contributes to our broader understanding of regional social policy. The welfare states in the region have traditionally been conceptualised as ‘exclusion-based’ (Craven, Reference Craven2015; Aspalter, Reference Aspalter2017), and our analysis highlights whether the expansion of social protection policies during the pandemic extends this trend.
In response to the global health crisis, Gulf states took various early interventions throughout the period of 2020 and 2022. These measures had a significant impact on various sectors, distinctively affecting vulnerable individuals, resulting in both protection from the virus and heightened social isolation (Alkhamees et al., Reference Alkhamees, Alrashed, Alzunaydi, Almohimeed and Aljohani2020; Alshehri et al., Reference Alshehri, Alatawi, Alghamdi, Alhifany and Alharbi2020; Algahtani et al., Reference Algahtani, Hassan, Alsaif and Zrieq2021; Alyami et al., Reference Alyami, Naser, Orabi, Alwafi and Alyami2020; Alenazi et al., Reference Alenazi, Bin Dhim, Alenazi, Tamim, Almagrabi, Aljohani, H Basyouni, Almubark, Althumiri and Alqahtani2020; Balkhi et al., Reference Balkhi, Nasir, Zehra and Riaz2020; Alamri et al., Reference Alamri, Algarni, Shehata, Al Bshabshe, Alshehri, ALAsiri, Hussain, Alalmay, Alshehri, Alqarni and Saleh2020; Iqbal and Tayyab, Reference Iqbal and Tayyab2020; Sonbol et al., Reference Sonbol, Alahdal, Alanazi, Alsamhary and Ameen2021; El Keshky et al., Reference El Keshky, Alsabban and Basyouni2021; Alyoubi et al., Reference Alyoubi, Halstead, Zambelli and Dimitriou2021).
The significance of examining the social protection measures introduced by Gulf states becomes evident in their impact on non-citizens such as migrant workers. The types and prevalence of measures adopted across countries reveal how these welfare regimes responded to the crisis and to what extent they addressed the needs of citizens and non-citizens. While the pandemic opened a temporary policy window for expanding social protection (Bengtsson et al., Reference Bengtsson, Høgedahl and Svalund2024), this research shows that the exclusionary structure of these welfare regimes largely persisted. Gulf states primarily offered short-term relief without addressing the structural inequities within these systems, particularly concerning the inclusion of non-citizen groups like migrant workers.
This study addresses the gap in research on the Gulf states’ responses to the COVID-19 pandemic. By examining social protection measures implemented in response to the crisis, it highlights how traditionally exclusionary welfare systems in the Gulf adapted under global pressures. This research also adds to the discourse on the adaptability and resilience of welfare states during unprecedented challenges, providing valuable insights for policymakers and researchers focused on this region.
During the pandemic, social protection initiatives emerged as critical support mechanisms across the Gulf states. Much like during the 2008 economic crisis (Grosh et al., Reference Grosh, Fruttero and Oliveri2013), these measures became central to managing the social and economic impacts of the pandemic. Social protection systems, designed to buffer against immediate rises in poverty and joblessness, include components such as social insurance, direct assistance, and labour market interventions (Devereux, Reference Devereux2021; Günther and Lahoti, Reference Günther and Lahoti2021). These interventions were instrumental in removing financial obstacles to essential health services and testing, allowing individuals to comply with quarantine without risking income loss (Bilo et al., Reference Bilo, Dytz, Hammad, Sato and Soares2023; Moussié et al., Reference Moussié, Goulder, Jha, Kip, Moloney, Pasqual, Raghavan, Rossmann, Wong, Addati and Behrendt2021). They also enabled households, particularly those relying on informal income sources, to meet basic needs amid declining economic activity and rising unemployment (Moussié et al., Reference Moussié, Goulder, Jha, Kip, Moloney, Pasqual, Raghavan, Rossmann, Wong, Addati and Behrendt2021).
Gulf states enacted a multitude of new and expanded social protection policies that constituted a divergence from the region’s historical trend. We see this divergence as a temporary policy window opened by the tumult of the pandemic. What remains to be seen is whether these policy changes will be durable beyond the punctuated equilibrium.
Theoretical framework
This study examines social protection responses through two theoretical lenses. The first theoretical lens, welfare regime theory, suggests that the Gulf states have had ‘exclusion-based’ welfare regimes that offer generous social protection for citizens while excluding non-nationals (Craven, Reference Craven2015; Aspalter, Reference Aspalter2023). These systems typically emphasise a Bismarckian social insurance system plus public payouts to citizens and rely on non governmental organisations (NGOs) for social service provision (Abrahmason, Reference Abrahamson and Aspalter2017; Aspalter, Reference Aspalter2023; Gal, Reference Gal and Aspalter2017; Jawad and Gal, Reference Jawad, Gal and Greve2019). This theoretical lens allows us to determine whether the social protection systems remained exclusion-based or underwent fundamental changes.
However, the onset of the pandemic initiated a period of dramatic policy change, raising important questions about how and why Gulf states enacted and expanded their social protection policies. To address these questions, the second theoretical lens, punctuated equilibrium theory (PET), frames these policy changes as occurring during a temporary policy window opened by the external shock of the pandemic (Baumgartner and Jones, Reference Baumgartner and Jones2009; Eissler et al., Reference Eissler, Russell and Jones2016). Eissler et al. (Reference Eissler, Russell and Jones2016) explain that PET posits that policy change is typically slow, incremental, and characterised by long periods of stability. Amid these trends, there are policy windows when sudden and significant shifts in policy can occur. Policy windows open in response to shocks or events that capture the attention of policymakers and the public, creating a sense of urgency for policy change. Punctuated equilibriums have led to significant policy change through mechanisms such as modified policymaking agendas and the influence of mass media and public pressure on policymakers (Baumgartner and Jones, Reference Baumgartner and Jones2009; Eissler et al., Reference Eissler, Russell and Jones2016).
Present study
This study examines the nature, extent, and implications of policy changes implemented during this crisis. Our research objectives are multifaceted: we seek to identify and analyse the social protection policies expanded or enacted in the Gulf states, examining their content and characteristics in depth. A key focus of our investigation is to assess whether these policies were extended to non-citizens. Furthermore, we aim to examine how these policy responses align with or deviate from the traditional exclusion-based welfare regimes of the Gulf states. This approach allows us to not only document the immediate policy responses, but also contextualise them within broader theoretical frameworks and historical patterns of social protection in the region.
Applying the frameworks of the welfare regime theory and punctuated equilibrium theory offers a unique perspective on policy changes in the context of the Gulf states. The study employs an innovative AI-assisted data extraction methodology, demonstrating a novel approach to regional policy analysis. Furthermore, it offers insights into the potential for welfare regime transformation in response to global crises. The research highlights the implications of these policy changes for migrant workers, a critical yet often overlooked population in Gulf welfare systems. These contributions enhance our understanding of crisis response in the Gulf states and inform broader discussions on the adaptability and resilience of welfare systems globally.
Methods
Overview and justification
Our study employs a mixed-methods approach combining database analysis, systematic review techniques, and AI-assisted data extraction to thoroughly investigate social protection measures during the global health crisis in the gulf region. Our methodological choices were guided by the unique challenges of studying rapidly evolving policy responses across multiple countries.
We selected the timeframe of March 2020 to March 2023 for our analysis to encompass both the immediate responses to the pandemic and potential longer-term policy adaptations. March 2020 marks the onset of widespread lockdowns and policy responses in the Gulf region. Extending the analysis to March 2023 enables us to capture not only the immediate crisis response but also the medium-term policy adaptations and the early stages of the post-pandemic policy landscape. This period offers a comprehensive view of how social protection measures were implemented, adjusted, and in many cases, phased out as the pandemic progressed.
We selected databases with comprehensive coverage of COVID-19 responses and a focus on social protection policies. Our keyword search strategy was informed based on a preliminary review of relevant literature and policy documents, to ensure we captured all pertinent policy measures. The decision to use AI-assisted data extraction was driven by the large volume of data across multiple languages and the need for rapid analysis of continuously updated information. Our analysis is centred on the six Gulf Cooperation Council (GCC) countries which share characteristics such as high-income, oil-exporting nations with similar welfare state structures. This approach enables meaningful comparison while also highlighting significant differences in policy responses.
Data sources and collection
Our study employed three complementary data collection and extraction techniques: database analysis, systematic review, and AI-assisted data extraction.
Our analysis drew on a range of data sources to capture a comprehensive view of social protection responses across the Gulf Cooperation Council (GCC) countries. Databases such as the Oxford COVID-19 Government Response Tracker (Hale et al., Reference Hale, Angrist, Goldszmidt, Kira, Petherick, Phillips, Webster, Cameron-Blake, Hallas, Majumdar and Tatlow2021), social protection and jobs (SPJ) policy inventory (Gentilini et al., Reference Gentilini, Almenfi and Dale2020), the United Nations Economic and Social Commission for Western Asia’s COVID-19 data tracker (ESCWA, 2021), and the International Labour Organisation database (ILO, 2020) were instrumental.
In addition, data from official government websites and publications for each GCC member country were reviewed. To deepen the analysis, we conducted a systematic literature review following PRISMA guidelines (Moher et al., Reference Moher, Liberati, Tetzlaff and Altman2009), focusing on terms linked to social protection responses within the GCC context. This review encompassed major academic databases, including Web of Science, Scopus, and PubMed, along with specialised regional databases like Al Manhal. Both peer-reviewed studies and grey literature in English and Arabic were examined to ensure a broad and inclusive understanding of relevant policies and interventions.
Our inclusion criteria for the systematic review were: (1) studies addressing social protection or labour market policies in response to COVID-19; (2) studies conducted in one or more of the six GCC countries; (3) studies published between March 2020 and March 2023; and (4) studies published in English or Arabic. We used a sample search string, modified for each database: (‘social protection’ or ‘social assistance’ or ‘cash transfer’ or ‘in-kind transfer’ or ‘labour market policy’ ORor ‘employment policy’ or ‘social insurance’) and (‘COVID-19’ or ‘coronavirus’ or ‘pandemic’) and (‘Gulf Cooperation Council’ or ‘GCC’ or ‘Saudi Arabia’ or ‘United Arab Emirates’ ‘or ‘UAE’ or ‘Kuwait’ or ‘Qatar’ or ‘Bahrain’ or ‘Oman’). Titles and abstracts were independently screened by two researchers, with a third resolving disagreements.
To efficiently process the large, multilingual data volume and quickly analyse evolving information, we employed an AI-assisted data extraction methodology using natural language processing (NLP) techniques in both Arabic and English. This approach was necessary due to rapid policy changes during the pandemic and the need to analyse diverse sources across multiple languages.
For NLP techniques, we implemented the spaCy library (version 3.1.0), fine-tuning spaCy’s pre-trained English and Arabic language models on a manually annotated subset of 500 policy descriptions from our dataset. For named entity recognition (NER), we used custom entity labels including ‘PolicyMeasure’, ‘TargetPopulation’, and ‘ImplementationDate’.
For machine learning classification and following Rider et al., (2021), we applied Naive Bayes and support vector machine (SVM) classifiers using scikit-learn (version 0.24.2). The models were trained on a labeled subset of 1000 policy descriptions using an eighty to twenty train-test split and five-fold cross-validation. Hyperparameters were optimised using grid search, with the final SVM model using a radial basis function kernel and C=10. The models achieved an F1 score of zero point eight five on the test set.
Our data preprocessing steps included removing HTML tags to extract clean text content from web-scraped data. This step ensures that only the relevant textual information is analysed, without interference from formatting elements. We used the BeautifulSoup library for this task, as it efficiently parses HTML and XML documents.
Additionally, we implemented our custom rule-based system for keyword-based classification using regular expressions and domain-specific dictionaries. This approach complemented the machine learning models by capturing specific policy-related terms and phrases that might be missed by more general classification algorithms.
To ensure transparency and replicability, we have made our code available in Annex A. This includes our data preprocessing scripts, NLP pipelines, and machine learning model implementations. Researchers interested in replicating or building upon our study can refer to this code for detailed implementation specifics.
Data integration and standardisation
The final results presented in this study represent a synthesis of data obtained through AI-assisted extraction, systematic review, and database analysis. We used a triangulation approach to cross-verify findings across different sources. We prioritised information from official government sources and peer-reviewed literature to tackle discrepancies. The AI-assisted extraction generated the initial dataset, and was then verified and supplemented by the systematic review findings. Additional context and data gaps were addressed through database analysis.
To maintain consistency across varied data sources, we applied a structured hierarchical categorisation framework. Initially, information collected from different sources was grouped into broad thematic categories, such as 'social protection policies’ and ‘labour market policies’, and was then further divided into specific subcategories within each theme. Likewise, the data was systematically organised by individual Gulf countries to allow for country-specific analysis. This framework facilitated a descriptive evaluation of policy shifts in the Gulf region, classifying initiatives as social protection, social insurance, or labour market actions. These classifications were then examined along dimensions like horizontal expansion, changes in implementation, vertical scaling, and ‘piggybacking’ (Bastagli, Reference Bastagli2014; Cherrier, Reference Cherrier2014; Gentilini et al., Reference Gentilini, Almenfi and Dale2020), providing a comparative perspective on policy adaptations across different Gulf states. We integrated these data using Python’s pandas library to create a unified dataset that harmonised information from multiple sources. Where inconsistencies or overlaps were identified between datasets, we relied on and prioritised official government sources and further cross-referenced these sources with academic literature for verification. Discrepancies that could not be resolved were documented in our database and excluded from aggregate analyses to maintain data integrity.
Sensitivity analysis
Due to the potential differences in data availability and quality across the six Gulf countries, we conducted a sensitivity analysis to assess how our results might change if countries with less comprehensive data were excluded. Table 1 of our results shows that Saudi Arabia and the UAE had the highest reported social protection measures (thirty two and twenty six, respectively), while Qatar reported the lowest (eleven). Accordingly, we conducted our primary analysis with all six countries, and then repeated the analysis excluding Qatar to determine if this significantly altered our findings.
Table 1. Social protection measures to global health crisis, by country, 2020–2021
Our sensitivity analysis focused on two key metrics: the distribution of social protection measures across categories (social assistance, social insurance, and labour market policies) and the proportion of measures targeting different groups (nationals, migrant workers, and universal coverage). When including all countries, the distribution of measures showed 47 per cent allocated to social assistance, 33 per cent focused on labour market policies, and 19 per cent directed toward social insurance programmes. Excluding Qatar, these percentages shifted slightly to 48 per cent, 32 per cent, and 20 per cent, respectively. For target groups, the analysis with all countries showed a distribution of 39 per cent targeting nationals, 35 per cent universal, and 25 per cent migrant workers. Excluding Qatar, these proportions changed to 41 per cent, 34 per cent, and 24 per cent, respectively.
The sensitivity analysis reveals that excluding Qatar, the country with the least reported measures, does not substantially affect our main findings. The distribution of social protection measures across policy categories remains largely consistent, with only slight percentage shifts. Similarly, the targeting of measures shows only slight variations, with a small increase in the proportion of measures targeting nationals when Qatar is excluded. This suggests that our results are robust, even when accounting for the exclusion of the country with the least reported data. However, we acknowledge that data limitations could still impact our findings, particularly for more granular analyses or comparisons of specific policy instruments across countries.
Data analysis
This study utilised IBM SPSS to analyse social protection measures across the Gulf states thoroughly. Key statistical tools included z-scores and the Diversity Index to provide a meaningful interpretation of the data. The z-score, a metric for assessing how a particular data point compares to the dataset’s average, was central to our analysis (Chubb and Simpson, Reference Chubb and Simpson2012). It is determined by subtracting the dataset’s mean from the value of the data point and then dividing by the standard deviation. The formula used to calculate the z-score is:
$$Z = {{X - {\rm{\mu}}} \over {\rm{\sigma}} }$$
where:
X, represents the individual data value
μ is the dataset’s mean, and
σ is the dataset’s standard deviation
Z-scores reveal how many standard deviations a particular measure is from the mean, thus indicating the relative position of each data point within the overall distribution. By standardising the data, we can compare values across different categories, facilitating a clearer interpretation of the social protection measures implemented by each GCC state. For instance, a positive z-score indicates that a measure is above average, while a negative z-score shows it is below average (Hogg and Tanis, Reference Hogg and Tanis2001). This analysis is useful for identifying outliers and understanding variations in the types of social protection instruments utilised across countries.
In addition to z-scores, we utilised the Diversity Index (Simpson, Reference Simpson1949) to assess various social protection measures implemented across the GCC states. The Diversity Index is a statistical measure that quantifies the distribution of different categories within a dataset, providing insight into the evenness of the representation of those categories (Simpson, Reference Simpson1949). The index ranges from zero to one, where zero indicates no diversity (all measures in one category) and 1 represents maximum diversity (measures evenly distributed across all categories).
In this study, we calculated the Diversity Index to evaluate the balance of social protection measures among different instruments – social assistance, social insurance, and labour market policies. A higher Diversity Index suggests a more comprehensive approach to social protection, indicating that a country is employing a broader range of measures to address various social needs (Stirling, Reference Stirling2007). Conversely, a lower index reflects a more focused approach, potentially prioritising specific types of support at the expense of others. This analysis helps us understand the relative complexity and adaptability of the welfare regimes in the Gulf states during the pandemic.
Findings
Social protection measures in the GCC
In response to the global health crisis, Gulf states substantially expanded their social protection systems, though the number of measures each country implemented varied widely. Qatar implemented the fewest measures (eleven) while Saudi Arabia implemented the most (thirty two), with a mean of eighteen point five measures per country and a standard deviation of eighteen point one seven (see Table 1). This variation reflects significant differences in the scale of policy responses among GCC states. Saudi Arabia and the UAE, implementing thirty two and twenty six measures, respectively, stand out as uniquely proactive in their policy approaches, significantly exceeding the regional average.
When examining the distribution of measures across the three categories, distinct patterns and priorities emerge among the Gulf states. Social assistance measures demonstrate the highest mean (eight point six seven) and standard deviation (four point one three) among the categories, indicating both its prominence and the variability of its implementation across countries. The UAE, with sixteen social assistance measures (z-score: two point three nine), prioritises direct support mechanisms within its social protection strategy, exceeding the regional average by more than two standard deviations.
In contrast, social Insurance measures show the lowest mean (three point five zero) and standard deviation (one point four six), suggesting a generally lower and more consistent emphasis on this category across the region. Notably, Saudi Arabia implemented six social insurance measures (z-score: one point four nine), reflecting a more comprehensive approach to social protection encompassing longer-term social security mechanisms.
Labour market measures reveal an interesting case of high variability (mean: six point three three, standard deviation: four point one three), with Saudi Arabia standing out again as a significant outlier with fifteen measures (z-score: one point five five). This underscores Saudi Arabia’s multifaceted approach to social protection spans immediate and labour market-specific interventions.
Calculating z-scores provides valuable insights into how each country’s measures in each category deviate from the regional mean. The UAE’s extreme positive deviation in social assistance (two point three nine) contrasts with its near-average scores in other categories, highlighting a specialised focus on social protection strategy. Saudi Arabia’s consistently positive z-scores across all categories (ranging from zero point nine zero to one point five five) underscores its comprehensive approach, exceeding regional averages in each area. Conversely, Qatar and Oman’s consistently negative z-scores across categories suggest a lower level of policy implementation compared to their regional counterparts.
To assess the balance of measures across categories within each country, we calculated a Diversity Index, ranging from zero (no diversity) to one (maximum diversity). This index, ranging from zero (no diversity) to one (maximum diversity), provides insight into the evenness of distribution across the three categories of social protection measures. The observed range of Diversity Index values (zero point five five four for the UAE to zero point six four three for Saudi Arabia, with a mean of zero point six zero nine) suggests that most GCC countries maintain some level of balance in their approach to social protection. Saudi Arabia’s high Diversity Index (zero point six four three) corroborates the observation of its comprehensive strategy, while the UAE’s lower score (zero point five five four) reflects its concentrated focus on social assistance.
The total number of instruments implemented across all categories is 111, with individual instrument types ranging from three to twenty seven in frequency. The mean number of instruments per category is eleven poitnt one, with a substantial standard deviation of eight point two three, indicating significant variability in the implementation of different types of instruments.
Utility waivers and financial obligation support emerge as the most frequently implemented instrument, with twenty seven instances (24 per cent of the total), yielding a z-score of zero point nine three (see Table 2). The use of this instrument is nearly two standard deviations above the mean, indicating a strong emphasis on providing immediate financial relief to individuals and households.
Table 2. Social protection measures (number, % of all measures), 2020–2021
Social Assistance Measures in Gulf Countries.
Employment regulations follow closely, with twenty-three implementations (21 per cent) and a z-score of one pont four four. This high frequency underscores the significant focus on adapting work environments and employment conditions in response to the crisis.
At the other end of the spectrum, social security contributions show the lowest frequency, with only three implementations (3 per cent) and a z-score of minus zero point nine eight. This suggests a relatively lower emphasis on long-term social security measures in the immediate response.
The z-scores for in-kind transfers (15 per cent) and cash transfers (10 per cent) are zero point four seven and minus zero point one three, respectively, which means that these direct support mechanisms are used above or close to the average amount. The remaining categories, including health insurance, paid sick leave, wage subsidies, activation measures, and unemployment benefits, all fall below the mean in terms of frequency, with z-scores ranging from minus zero point five zero to minus zero point eight six. This distribution suggests a relatively lower emphasis on these specific instruments compared to the more frequently implemented measures.
To assess the diversity of the implemented instruments, Diversity Index for this distribution is zero point eight six three, indicating a high level of diversity in the types of instruments implemented. This high value suggests a multi-faceted approach to social protection, with a relatively even distribution of efforts across various instrument types.
This study reveals a strategic focus on immediate financial support and labour market adaptations, as evidenced by the high frequency of utility waivers, financial obligation support, and employment regulations. The lower emphasis on long-term measures such as social security contributions may reflect the urgency of the situation and the need for immediate relief. The high Diversity Index indicates a comprehensive approach to social protection, utilising a wide range of instruments to address various aspects of social and economic challenges.
Table 3 provides valuable insights into the distribution and emphasis of social assistance measures across the Gulf countries. The analysis reveals significant variations in approach and focus, reflecting diverse policy priorities and socioeconomic contexts among these nations.
Table 3. Social assistance measures by country, 2020–2021
The total number of social assistance measures implemented varies considerably across countries, ranging from six in Oman to sixteen in the UAE. This disparity suggests differing levels of governmental intervention in social welfare, potentially influenced by factors such as economic resources, population size, and social policy frameworks. The UAE’s higher number of measures may indicate a more comprehensive approach to social assistance, possibly linked to its efforts at diversification and its vision for social development.
Waivers in the form of utility and financial support emerge as the most prevalent form of assistance, with a mean of four point eight zero measures across countries. This emphasis aligns with research suggesting that GCC countries often prioritie subsidies and financial support as key mechanisms for wealth distribution and social stability. Notably, the UAE stands out with eleven measures in this category (z-score: one point five seven), significantly above the regional average. This focus may reflect the UAE’s strategy to address cost of living concerns in its rapidly developing urban centres.
Cash transfer programmes have the lowest mean (one point eight zero) and the most consistent application across countries, with three nations implementing three measures each. This consistency suggests a regional recognition of the importance of direct financial support, albeit at a moderate level. The absence of cash transfer measures in Oman and the UAE (z-score: minus one point one zero for both) is noteworthy and may indicate a preference for other forms of assistance or concerns about the long-term sustainability of direct cash programmes.
In-kind transfers demonstrate considerable variability (std dev: two point one two), with Kuwait and the UAE leading (five measures each, z-score: one point one three). This approach may reflect these countries’ efforts to provide tangible, immediate assistance to vulnerable populations, potentially influenced by their relatively higher gross domestic product (GDP) per capita and ability to finance such programmes.
To further analyse the distribution of social assistance measures, we calculated the Diversity Index, based on Simpson’s Diversity Index (Simpson, 1949). This index, commonly used in ecological studies and increasingly applied in the social sciences (Stirling, Reference Stirling2007), provides insight into how evenly measures are distributed across categories. It ranges from zero to one, where zero indicates no diversity (all measures in one category) and one represents maximum diversity (measures equally distributed across all categories).
Saudi Arabia exhibits the highest diversity (zero point six six zero), suggesting a more balanced approach across different types of assistance. This aligns with the country’s recent efforts to reform and diversify its social protection systems as part of its Vision 2030 plan. The higher diversity score indicates that Saudi Arabia is employing a varied toolkit of social assistance measures, potentially enabling a more comprehensive coverage of diverse social needs.
In contrast, the UAE’s lower Diversity Index (zero point four three nine) reflects its strong emphasis on utility waivers and financial support, which may come at the expense of a more balanced assistance portfolio. This concentration suggests a strategic focus on particular types of social assistance that the UAE government deems most effective or necessary within its socioeconomic context.
The variation in Diversity Index scores across the Gulf countries (ranging from zero point four three nine to zero point six six zero) highlights the different strategies employed in social assistance policies. A higher diversity in measures could indicate an attempt to address a wider range of social issues or to cater to more diverse population needs. Conversely, a lower diversity score might suggest a more targeted approach, focusing resources on specific types of assistance deemed most crucial or effective in a particular national context.
Social insurance
In examining the distribution of social insurance measures across six Gulf countries, we observe variations in both the number and types of instruments implemented. The dataset covers four categories of social insurance: health insurance, paid sick leave, social security contributions, and unemployment benefits.
The total number of measures implemented ranges from two (Bahrain and Qatar) to six (Saudi Arabia), with a mean of three point five measures per country and a standard deviation of one point six four. This indicates considerable variability in the overall implementation of social insurance measures among GCC states.
To facilitate a comparison, we calculated z-scores for each measure type within each country. Health insurance and paid sick leave measures show identical distributions, with a mean of one point one seven and a standard deviation of zero point seven five for each. Saudi Arabia and the UAE lead in both categories with two measures each (z-score: zero point five eight), while Kuwait stands out for not implementing any health insurance measures (z-score: minus one point one five) and Oman for not implementing any paid sick leave measures (z-score: mius one point one five) (see Table 4).
Table 4. Social insurance measures, by instrument, by country, 2020–2021
Social security contributions demonstrate the lowest mean implementation (zero point five zero) with a standard deviation of zero point five five. Kuwait, Oman, and the UAE have implemented one measure each in this category (z-score: zero point three four), while the other countries have not implemented any, resulting in negative z-scores (mius zero point six seven).
Unemployment benefits show slightly higher implementation with a mean of zero point six seven and the highest standard deviation of zero point eight two. Saudi Arabia leads with two measures (z-score: one point six three), while Bahrain, Qatar, and the UAE have not implemented any unemployment benefit measures (z-score: minus zero point eight two).
With regards to the Simpson Diversity Index, Kuwait, Oman, and Saudi Arabia show the highest Diversity Index (zero point six six seven), indicating a balanced approach across the implemented measures. The UAE follows closely with a Diversity Index of zero point six four zero. Bahrain and Qatar have the lowest Diversity Index (zero point five zero zero), reflecting their implementation of measures in only two out of the four categories.
This analysis reveals varying approaches to social insurance among GCC countries. Saudi Arabia and the UAE demonstrate a more comprehensive approach, implementing more measures across categories. In contrast, Bahrain and Qatar show a more focused strategy, concentrating on health insurance and paid sick leave. The higher Diversity Indices of Kuwait, Oman, and Saudi Arabia suggest a more balanced distribution of measures across the four categories, potentially indicating a more holistic approach to social insurance.
These differences may reflect varying policy priorities, economic conditions, and social needs across the GCC states. Future research could explore the factors driving these variations and assess the effectiveness of these different approaches in addressing social welfare challenges in the region.
Labour market measures
We observe significant variations in both the number and types of labour market measures implemented. The total number of measures implemented ranges from four (Kuwait and Oman) to fifteen (Saudi Arabia), with a mean of six point three three measures per country and a standard deviation of four point two three (see Table 5). This indicates substantial variability in the overall implementation of labour market measures among GCC states.
Table 5. Labour market measures, by country, 2020–2021
Activation measures show the lowest mean implementation (zero point six seven with a standard deviation of one point two one. Saudi Arabia stands out significantly in this category with three measures (z-score: one point four one), while most other countries have not implemented any activation measures (z-score: minus zero point six zero).
Employment regulations demonstrate the highest mean implementation (three point eight three) with a standard deviation of one point three three. Saudi Arabia leads again with six measures (z-score: zero point six eight), while Kuwait has the lowest implementation with two measures (z-score: minus one point four one).
Wage subsidies show considerable variation with a mean of one point eight three and the highest standard deviation of two point one four. Saudi Arabia significantly leads this category with six measures (z-score: two point six one), while Oman and the UAE have not implemented any wage subsidy measures (z-score: mius one point zero four).
Saudi Arabia shows the highest diversity index (zero point six four zero), indicating a relatively balanced approach across the implemented measures. Kuwait follows with a diversity index of zero point five zero zero. Bahrain (zero point four eight zero), Qatar (zerp point three two zero), and the UAE (zero point two three zero) show lower diversity, reflecting a more focused approach on specific measures. Oman has the lowest Diversity Index (zero point zero zero zero), as all its measures are concentrated in one category.
This analysis reveals varying approaches to labour market measures among GCC countries. Saudi Arabia demonstrates a comprehensive approach, implementing the highest number of measures across all categories. In contrast, other countries show more focused strategies, concentrating on specific types of measures. The higher Diversity Index of Saudi Arabia suggests a more balanced distribution of measures across the three categories, potentially indicating a more holistic approach to labour market interventions. The strong emphasis on employment regulations across all countries (highest mean of three point eight three) suggests a common focus on adapting work environments and employment conditions.
Expansion of social protection
In analysing the implementation changes of social protection measures across the Gulf states, we observe diverse policy adaptations that reflect the evolving nature of social safety nets in the region. This heterogeneity in approach aligns with contemporary discourse on adaptive social protection (ASP) and shock-responsive social protection (SRSP) frameworks (Davies et al., Reference Davies, Béné, Arnall, Tanner, Newsham and Coirolo2013; O’Brien et al., Reference O'Brien, Scott, Smith, Barca, Kardan, Holmes, Watson and Congrave2018).
The data reveals significant variations in both the quantity and typology of implementation changes across the six GCC countries. Saudi Arabia emerges as the vanguard, with thirty two total changes and a Diversity Index of zero point seven nine nine, suggesting a comprehensive approach to social protection system reform.
The prevalence of ‘implementation changes only’ across all countries (mean of seven point three three, σ = two point five eight) indicates a focus on parametric reforms rather than paradigmatic shifts in social protection architecture. This trend can be interpreted through the lens of incremental institutionalism in social policy (Mahoney and Thelen, Reference Mahoney and Thelen2009), where gradual changes accumulate to produce significant transformations over time. The emphasis on modifying existing programmes rather than creating new ones suggests that GCC countries are engaging in what Barrientos (Reference Barrientos2013) terms ‘social assistance institutionalisation’, strengthening the foundations of their existing social safety nets.
Horizontal expansion, the second most common change type (μ = three point eight three, σ = two point nine three), demonstrates significant inter-country variation (see Table 6). This heterogeneity could be explained by differing capacities in implementing what Devereux and Sabates-Wheeler (Reference Devereux and Sabates-Wheeler2004) term ‘transformative social protection’. which aims to address social equity and exclusion. The variability may reflect diverse approaches to expanding social protection coverage to previously excluded populations, a key tenet of the universalism principle in social protection discourse (ILO, 2013).
Table 6. Implementation features of social assistance measures, by country, 2020–2021
The less frequent yet significant use of vertical expansion and combined approaches points to a nuanced understanding of the multi-dimensional nature of vulnerability. This multi-tiered approach aligns with the concept of ‘social protection plus’ (Roelen et al., Reference Roelen, Delap, Jones and Chettri2017), which advocates for complementary interventions that address both immediate needs and long-term resilience building.
The calculated z-scores reveal interesting patterns of specialisation and innovation in social protection reform. Saudi Arabia’s significant positive z-scores across multiple categories suggest a role as a regional policy entrepreneur in social protection, potentially influencing other GCC states through policy diffusion mechanisms, as theorised by Obinger et al. (Reference Obinger, Schmitt and Starke2013) in their work on social policy interdependence.
The Diversity Index scores, ranging from zero point six six nine (Oman) to zero point seven nine nine (Saudi Arabia), provide insight into the balance of social protection instruments within each country. This variation can be understood through the theoretical lens of social protection systems approach (SPSA) (World Bank, 2012), which emphasises the importance of coherent and coordinated interventions across various social protection instruments. The higher diversity indices of Saudi Arabia and the UAE indicate a more balanced policy mix, potentially allowing for a more comprehensive addressing of lifecycle risks and chronic poverty.
These findings have significant implications for understanding social protection system development in the GCC region. The variations in approach suggest a dynamic interplay between regional policy convergence and national innovation, reflecting what Barrientos and Hulme (Reference Barrientos and Hulme2008) term the ‘new politics of social protection’ in developing countries. This balance between regional trends and national specificity underscores the complex interplay of global social protection paradigms, regional cooperation, and domestic political economies in shaping social protection trajectories.
Social protection coverage
We observe a complex landscape of social protection targeting across the GCC states, reflecting diverse approaches to what Devereux and Sabates-Wheeler (Reference Devereux and Sabates-Wheeler2004) term ‘transformative social protection’. The variation in beneficiary focus among countries suggests differing conceptualisations of vulnerability and prioritisation of social risks.
The emphasis on universal measures (μ = six point five zero, σ = three point one five) across all countries aligns with the growing global discourse on Universal Social Protection (USP2030), as advocated by the ILO and World Bank (2015). The UAE’s strong focus on universal measures (z-score: zero point seven four) suggests a move towards what Esping-Anderson (Reference Esping-Anderson1990) might term as a more ‘universalist’ welfare regime within the context of a rentier state economy. On the other hand, the fact that formal employees got a lot of attention (μ = four point one seven, π = four point two seix), especially in Saudi Arabia (z-score: two point two five) and Oman (z-score: zero point five six), shows that what Rudra (Reference Rudra2015) calls ‘protective’ welfare states that put formal sector workers first are still around. This approach may reflect the ongoing challenges of labour market dualism in the region, as discussed by Baldwin-Edwards (Reference Baldwin-Edwards2011) in the context of GCC labour markets.
The focus on migrant workers (μ = four point six seven, σ = one point eight six) is particularly noteworthy, with the UAE (z-score: one point three zero) and Qatar (z-score: zero point seven eight) showing above-average attention to this group. This trend aligns with recent calls for extending social protection to migrant workers, as articulated in the Global Compact for Safe, Orderly, and Regular Migration (Global Compact, 2018). However, the variability in approach suggests differing levels of what Holzmann and Koettl (Reference Holzmann and Koettl2015) term the ‘portability of social benefits’ for migrant populations.
The relatively lower focus on poor and vulnerable households (μ = one point three three, σ = one point two one) and vulnerable individuals (μ = one point eight three, σ = one point seven two) across most countries, with some exceptions such as Kuwait (z-scores: one point five two and one point five three, respectively), raises questions about the adequacy of targeted social assistance in the region. This pattern may reflect the challenges of implementing what Barrientos (Reference Barrientos2013) terms ‘social assistance institutionalisation’ in contexts where traditional family-based welfare systems have been predominant.
The Diversity Index scores, ranging from zero point six three nine (Qatar) to zero point seven six zero (Kuwait), provide insight into the balance of targeting approaches within each country (see Table 7). Gentilini (Reference Gentilini2015) terms this variation as the ‘targeting trilogy’ – the interplay between universal, categorical, and poverty-targeted approaches. Kuwait’s high diversity score suggests a more balanced approach across beneficiary categories, potentially indicative of a more comprehensive social protection system.
Table 7. Target groups of social protection measures, by country, 2020–2021
These findings have significant implications for understanding the evolution of social protection systems in the GCC region. The variations in targeting approaches suggest a dynamic interplay between the traditional social contract, based on the distribution of oil rents (Hertog, Reference Hertog2017), and emerging global social protection paradigms. This tension reflects what Jawad (Reference Jawad2015) describes as the ‘new politics of social protection’ in the Middle East, where states are grappling with the challenges of expanding and deepening social protection coverage in the context of economic diversification efforts and changing demographic realities.
The analysis of social protection measures across the Gulf states reveals a complex landscape of policy implementation. Table 6 demonstrates significant variations in approach and focus across different beneficiary groups – Nationals, Migrant Workers, and Universal coverage. The distribution of measures exhibits considerable heterogeneity, with employment regulations and Shorter Work Time Arrangements showing the highest mean (eight point zero zero) and standard deviation (seven point nine four). This variability underscores the diverse approaches taken by GCC states in adapting their labour markets, mainly in response to the needs of both national and migrant workforces. The high z-score (three point five zero) for Migrant Workers in this category highlights a pronounced focus on workplace protections for this group, aligning with the recent global discourse on extending social protection to migrant populations (ILO, 2013). Nationals receive the highest mean number of measures (five point one zero), with particular emphasis on Cash Transfers (z-score: one point four nine), Utility Waivers, and Financial Obligation Support (z-score: two point four six). This trend aligns the traditional social contract, prioritising citizens in the distribution of state benefits. However, the data also reveals an emerging shift towards more inclusive policies, as evidenced by the significant z-scores for Universal coverage in In-kind Transfers (one point five five) and Health Insurance (zero point eight zero). This move towards universalism in certain areas reflects global trends in social protection expansion and aligns with the Sustainable Development Goals call for universal social protection systems (ILO and World Bank, 2015).
The Diversity Index calculations provide further insight into the balance of approaches across beneficiary groups for each instrument type. Utility Waivers and Financial Obligation Support demonstrate the highest diversity (zero point six two six), indicating a more balanced distribution of benefits across all groups. This approach resonates with Barrientos’ (Reference Barrientos2013) concept of ‘social assistance institutionalisation’, suggesting a maturation of social protection systems that aim to address a broader spectrum of social risks. Conversely, the lack of diversity in measures such as Cash Transfers and Health Insurance (both with a Diversity Index of zero point zero zero zero) points to a more targeted approach in these areas, potentially reflecting fiscal constraints or specific policy priorities.
Table 8 also reveals interesting patterns in Social Insurance measures. While Unemployment Benefits and Social Security Contributions are exclusively targeted at Nationals (Diversity Index: zero point zero zero zero), Paid Sick Leave shows a more balanced distribution (Diversity Index: zero point four zero eight), including coverage for Migrant Workers. This variability in approach within the Social Insurance category suggests an ongoing negotiation between traditional welfare state models and the unique demographic realities of GCC countries, characterised by large migrant worker populations.
Table 8. Target groups of social protection measures, by type of instrument, 2020–2021
Labour Market measures present a particularly intriguing picture. The high mean and standard deviation in employment regulations, coupled with the significant focus on Migrant Workers, reflect the complex dynamics of GCC labour markets. This emphasis aligns with what Hertog (Reference Hertog2014) terms the ‘segmented marketisation’ of GCC economies, where labour market reforms often aim to balance national employment priorities with the needs of a diverse expatriate workforce.
The overall pattern that emerges from this analysis is a social protection landscape in transition. GCC states appear to be navigating between the traditional model of citizen-focused welfare provision and a more inclusive approach that acknowledges the diverse composition of their populations. This balancing act is evident in the varied Diversity Index scores across instrument types and the distribution of measures among different beneficiary groups.
These findings have significant implications for understanding the evolution of social protection systems in the GCC region. The data suggests a gradual move towards what could be termed a ‘hybrid’ model of social protection – one that maintains strong support for nationals while incrementally expanding coverage to other population segments. This approach reflects the unique challenges faced by GCC states in adapting global social protection norms to their specific socio-economic contexts.
Technological innovations
Reflecting a broader international trend (Gentilini et al., Reference Gentilini, Almenfi and Dale2020), several Gulf states leveraged digital platforms in their strategic responses to the pandemic. For instance, the Saudi government launched the ‘Ajeer’ portal to enhance local hiring practices (https://www.ajeer.com.sa), thereby safeguarding the employment rights of foreign workers during this challenging period. This innovative approach facilitated greater mobility within the labour market. Additionally, the online resource ‘Awdah’ aided expatriates in planning their departures during the lockdown (https://www.my.gov.sa).
In tandem, the government also introduced the ‘Sehhaty’ application, enabling users to efficiently manage sick leave registration (https://sehaty.sa). The UAE also made significant strides, employing mobile technology for distance education across the public and private sectors, particularly aimed at individuals with disabilities and early intervention initiatives. Complementing this effort, vocational training schemes were established to assist people with disabilities in effectively marketing their handcrafted products.
Evolution and status of social protection measures
As of March 2023, the conclusion of our study period, many social protection measures implemented in response to the global health crisis had either been phased out or retracted across the Gulf states. However, the timing and extent of these retractions varied significantly by country and type of measure.
In Saudi Arabia, the wage protection programme for private sector employees, which provided up to 60 per cent of Saudi employees’ salaries, ended in September 2021. Nevertheless, some utility waivers and financial obligation support measures were extended into 2022. For example, the suspension of fines for expired residency permits remained until June 2022 to support migrant workers affected by travel restrictions.
The UAE phased out most COVID-19 related social assistance measures by mid-2022. The ‘Ten Million Meals’ campaign, which provided food support to vulnerable families, concluded in May 2021. However, certain labour market regulations, such as the virtual labour market platform to facilitate job searches, continued to operate beyond our study period.
Qatar retained several measures related to migrant worker protections through 2022, likely influenced by preparations for the International Federation of Association Football (FIFA) World Cup. For example, the requirement for employers to provide quarantine facilities for migrant workers persisted until December 2022. The free COVID-19 testing and treatment for all residents, including migrant workers, was one of the last measures to be phased out in February 2023.
Kuwait began retracting most temporary measures in late 2021, with the COVID-19 crisis fund for impacted families concluding in November 2021. However, some targeted support for vulnerable citizens, such as increased unemployment benefits, continued into 2023.
Bahrain’s approach to retracting measures was more gradual. The deferment of loan installments for affected individuals and companies was phased out in stages, with the final extension ending in June 2022. The suspension of electricity and water bills for individuals and companies most affected by the pandemic remained in effect until December 2022.
Oman’s retraction of measures was similarly staggered. The job security programme, which provided financial support to companies to retain Omani employees, concluded in September 2021. However, certain labour market flexibility measures, such as allowing private sector companies to place employees on unpaid leave, remained in effect until the end of our study period.
It is important to note that while many temporary measures have been retracted, some policy changes, particularly in labour market regulations, have resulted in more permanent adjustments in social protection systems across the Gulf states. Saudi Arabia’s expansion of unemployment insurance coverage and the UAE’s reforms to increase labour market flexibility appear to have become permanent features of their social protection landscapes. However, the long-term impact of these changes, specifically on migrant workers and vulnerable populations, remains to be fully assessed.
Discussion
This study examines social protection measures implemented as responses to the pandemic across the gulf countries. Specifically, it examines how the abrupt disruptions caused by the pandemic impacted the exclusionary welfare paradigms prevalent in these nations. Our research indicates that these countries carefully assessed the risks associated with external shock and subsequently implemented a series of flexible changes to protect their citizens and residents. These changes resulted in 111 different actions that built upon existing on welfare programmes.
The analysis reveals an expeditious and comprehensive response, primarily characterised by the initiation and implementation of temporary interventions. Much like other affluent nations employing Bismarckian welfare models, the GCC countries heavily depended on social assistance and employment regulations to mitigate the socio-economic repercussions of the pandemic.
The utilizstion of cash transfer programmes to maintain income stability was relatively limited. Gulf states instituted new emergency cash transfer schemes, expanded existing programmes to new beneficiary groups, or increased the value of benefits. Some countries vertically expanded existing programmes by enhancing payments to current beneficiaries, while others opted for horizontal expansion by increasing the number of beneficiaries or creating temporary emergency schemes targeting vulnerable demographics. However, access to cash transfers remained restricted for a significant portion of migrant workers, as no modifications were made to formal access rules during the pandemic. Furthermore, the pandemic prompted the provision of in-kind support, primarily in the form of food baskets and hygiene supplies, focusing on essential needs for at-risk households, primarily during Ramadan.
To stabilise labour demand and enhance market efficiency (Squire and Suthiwart-Narueput, Reference Squire and Suthiwart-Narueput1997; Betcherman, Reference Betcherman2015), GCC nations prioritised adjustments to employment regulations that fostered greater job flexibility, job rotation, and shortened working hours. These adjustments included changes in remuneration practices and mandates requiring employers to provide paid leave for private sector workers, including those under self-quarantine. Furthermore, they established new working conditions that regulated wage reductions, termination processes, and hiring flexibility, all of which were previously absent. The variability of these regulations across GCC nations was notable; some states prohibited firms from dismissing workers (particularly citizens), while others allowed temporary contract terminations. Wage policies also differed, with some nations permitting reductions while others mandated increases for citizens.
We anticipate that the GCC states, like other affluent nations, will retract the temporarily introduced expanded benefits as the policy landscape shifts and the crisis begins to close. From a welfare regime theory perspective, it appears that these countries have intensified their exclusionary welfare frameworks while providing limited coverage to non-citizens. It appears that the welfare regimes in the Gulf states did not transition to a different model during the pandemic, although the treatment of non-citizens, particularly migrant labourers, necessitates further exploration.
Despite their uneven implementation across the region and potential inadequacy, these initiatives mark a significant shift in the exclusionary regimes’ welfare philosophy during this policy period. While these modifications do not warrant a complete reclassification of the welfare regimes due to their temporary nature, they may lay the groundwork for more inclusive social policies in the future.
The findings of our study reveal interesting patterns in social protection responses across the Gulf countries during the COVID-19 pandemic. Notably, these nations predominantly relied on utility and financial obligation support, with less emphasis on unemployment insurance and social security expansions compared to other countries globally. This pattern warrants further examination within the unique socio-economic context of the Gulf region.
The limited expansion of unemployment insurance and social security in GCC countries during the pandemic can be attributed to several factors. Firstly, most GCC countries already had relatively generous social welfare systems for their citizens before the pandemic. For instance, in Saudi Arabia, the unemployment insurance system (SANED) provided up to 60 per cent of the average wage for up to twelve months, even before the COVID-19 crisis (General organization for social insurance, 2023). Similarly, the United Arab Emirates had a comprehensive social security system that covered various aspects of citizens’ welfare (UAE Ministry of Community Development, 2019). This existing safety net may have reduced the perceived need for substantial expansions during the pandemic. However, it is crucial to note that these generous systems primarily benefit citizens, who constitute a minority in most GCC countries. The majority of the workforce in these nations consists of expatriate workers who often have limited access to social security benefits (Hertog, Reference Hertog2014). This disparity may explain why utility and financial obligation support became the predominant form of assistance during the pandemic – it was a measure that could reach both citizens and non-citizens alike.
The heavy reliance on utility waivers and financial obligation support across GCC countries can be interpreted as a pragmatic approach to providing immediate relief to a broad segment of the population. These measures were likely easier to implement quickly and could benefit both citizens and residents without fundamentally altering existing social security structures. For example, in the UAE, the Dubai Electricity and Water Authority (DEWA) implemented a 10 per cent reduction in water and electricity bills for a three-month period, benefiting all consumers (Abbas, Reference Abbas2020).
Comparative analysis reveals some variations in approach among GCC countries. On the one hand, Saudi Arabia implemented a wider range of measures, including wage subsidies and expanded unemployment benefits, possibly reflecting its larger population and more diverse economy. On the other hand, the UAE focused more heavily on financial sector support and SME assistance, aligning with its role as a regional business hub (UAE Ministry of Economy, 2020).
These findings carry several policy implications. Firstly, the pandemic has underscored the need for developing more inclusive social protection systems that cover all residents, not just citizens. This is particularly crucial in GCC countries with large expatriate populations. Secondly, the emphasis on utility and financial obligation support demonstrates the value of flexible, quickly implementable measures in crisis situations. Future policy planning should consider maintaining this flexibility while also strengthening underlying social security systems. Thirdly, while immediate relief measures were necessary, there is a need to balance these with investments in long-term social protection infrastructure that can better withstand future shocks. Lastly, the differential impact of social protection measures on citizens versus non-citizens necessitates policies that address existing disparities in social safety nets.
Future research should explore the long-term impacts of these policy choices, particularly on labour market dynamics and social cohesion in GCC countries. Additionally, comparative studies with other regions could provide valuable insights into the effectiveness of different social protection strategies in crises. Our findings contribute to the broader discourse on welfare state adaptability and resilience in times of crisis. The Gulf states’ responses to COVID-19 align with global trends observed in other high-income countries, specifically those with Bismarckian welfare systems (Greve et al., Reference Greve, Blomquist, Hvinden and Van Gerven2021). However, our study reveals unique aspects of the responses of the Gulf welfare regimes. The predominance of social assistance and labour market policies over social insurance measures in Gulf states contrasts with the responses of many European welfare states, which relied heavily on expanding existing social insurance schemes (Béland et al., Reference Béland, Dinan, Rocco and Waddan2021). This difference likely reflects the Gulf states’ traditionally limited social insurance coverage, particularly for non-citizens.
Our findings also contribute to the ongoing debate about the potential for welfare state expansion during crises (Starke et al., Reference Starke, Kaasch, Van Hooren and Van Hooren2013). While the Gulf states did expand some protections, notably for citizens, the limited and temporary nature of many measures supports the argument that crises often lead to short-term expansions rather than fundamental regime shifts (Farnsworth and Irving, Reference Farnsworth and Irving2018). The treatment of migrant workers during the pandemic adds nuance to discussions of dualised welfare systems (Emmenegger, Reference Emmenegger2012). Indeed, Gulf states made some concessions to include migrant workers in their COVID-19 responses, however, our findings suggest that the crisis did not fundamentally alter the exclusionary nature of these welfare regimes, contrary to what some scholars have predicted for other dualised systems facing similar pressures (Avlijas et al., Reference Avlijas, Hassel and Palier2021; Hassel and Palier, Reference Hassel and Palier2023, Reference Hassel and Palier2021).
Limitations and future research
This study has many limitations. Bracketed as a descriptive, exploratory study, we incite future research to consider explicitly testing theories and employing causal inferences. Our treatment of time can be contested. Since we are relatively close to the onset of the COVID-19 pandemic at the time of writing, we lack the hindsight to clearly demarcate the end of the pandemic or the subsequent opening and closing of related policy windows. To address this, future research can carefully examine the effects of time on policy expansion and retrenchment in the Gulf nations.
While our multi-method approach aimed to provide comprehensive coverage of policy responses across the GCC countries, we acknowledge that data availability and quality varied across nations. Saudi Arabia and the UAE generally had more accessible and detailed policy information compared to smaller GCC states. To mitigate potential biases, we conducted sensitivity analyses that excluded countries with less comprehensive data. However, readers should interpret cross-country comparisons with caution, particularly in areas where data gaps are identified or are more pronounced. Future research could benefit from more standardised data collection methods across the region to facilitate more precise cross-country comparisons.
Furthermore, the data collection and initial analysis for this study were completed in June 2023. Given the rapidly evolving nature of COVID-19 policy responses, readers should be aware that additional changes may have occurred between the end of our study period (March 2023) and the time of publication. Future research could build on this work by examining the longer-term impacts and potential permanence of the social protection measures implemented during the pandemic. This study invites future research to contribute to the philosophical drift among exclusionary-based welfare regimes in times of crisis and post-COVID-19. Future studies should seek to understand policies that affect non-citizens and migrant workers in the Gulf nations with greater granularity and investigate whether Gulf nations’ policy approaches to these groups evolve over time.
Conclusion
The COVID-19 pandemic has served as a pivotal moment for social protection policies in the Gulf states, challenging long-held welfare paradigms and prompting significant policy adaptations. This study critically examined the evolving welfare policies of six Gulf countries – Saudi Arabia, the UAE, Kuwait, Bahrain, Qatar, and Oman – through the lenses of welfare regime theory and punctuated equilibrium theory. Our analysis reveals that while the pandemic created a temporary policy window allowing for the expansion of social protection measures, the underlying exclusionary nature of these welfare regimes remained largely intact.
Historically characterised as exclusion-based welfare systems, Gulf states have predominantly focused on providing social protection to their citizens while largely excluding non-citizens, particularly migrant workers. The pandemic highlighted the vulnerabilities of these workers, prompting Gulf states to adopt measures extending some level of protection to them. This shift was distinctly evident in implementing utility waivers, financial obligation support, and emergency cash transfers to mitigate the health crisis’s immediate socio-economic impacts. Notably, the data indicated a substantial variation in the number and types of measures adopted across the GCC countries, with Saudi Arabia and the UAE standing out for their proactive policy approaches.
Despite these advancements, the long-term sustainability and implications of the implemented policies remain uncertain. While the pandemic opened opportunities for welfare expansion, it did not catalyse a fundamental transformation of the exclusionary welfare regimes in the Gulf. Instead, the measures introduced were often temporary and reactive, aimed primarily at alleviating the immediate impacts of the crisis rather than addressing the deeper structural issues within the welfare systems. As evidenced by the data, social assistance measures comprised the most significant proportion of the responses, with limited emphasis on social insurance policies, which continue to reflect the traditional priorities of Gulf welfare states.
Furthermore, the reliance on utility waivers and financial obligation support – measures that could be quickly implemented and broadly disseminated – underscores a pragmatic approach to crisis management. This approach facilitated rapid relief to both citizens and migrant workers but did not significantly alter the structural dynamics of the welfare system. Consequently, the exclusionary nature of these regimes continues to present difficulties for the most vulnerable groups, especially non-citizens, who are susceptible to slipping through the gaps in social protection systems.
Looking ahead, the GCC states must grapple with the question of whether the temporary measures adopted during the pandemic can lead to lasting reforms in their welfare systems. The data suggests that while some countries may be moving toward a more inclusive social protection framework, particularly regarding migrant workers, the overarching trends indicate a continued preference for policies that primarily benefit citizens. The potential for a more comprehensive and inclusive approach to social protection is contingent upon a re-evaluation of the existing welfare paradigms in the region.
In conclusion, this study contributes to the understanding of the adaptability of Gulf welfare states in times of crisis and provides insights into the future trajectory of social protection in the region. As the GCC countries navigate the aftermath of the pandemic, it will be crucial for policymakers to consider the lessons learned from this crisis, specifically the importance of inclusivity in social protection measures. By fostering a more inclusive approach that recognises the contributions and needs of all residents, including migrant workers, Gulf states can enhance the resilience and sustainability of their welfare systems in the face of future challenges. Ultimately, the long-term success of social protection reforms in the Gulf will depend on a commitment to addressing the fundamental inequities embedded in their welfare regimes, paving the way for a more equitable social safety net for all.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1474746424000630
Open access
