Confirmatory Factor Analysis of Assets That Influence Informal Garment Workers ’ Livelihood Security in Laos

Laos is a socialist transition economy that has continued to increase its household characteristics, social assets, and human assets, all of which significantly influence livelihood security. The study sample generated 333 responses with confirmatory factor analysis (CFA) conducted using the LISREL software 9.30 for Window. The statistical model showed good fit with χ2/df = 62, CFI (comparative fit index) = 0.968, GFI (goodness-of-fit index) = 0.969, and RMSEA (root-mean-square error of approximation) = 0.0358, respectively. The findings confirmed that the standardized coefficients for household characteristics (β = 0.429, p < 0.01), social assets (β = 0.505, p < 0.01), and financial assets (β = 0.423, p < 0.01) positively influenced livelihood security. The implication of the current findings regarding the model will be discussed.


Introduction
Laos has been concerned with the ongoing and gradual implementation of the new economic mechanism since 1986, when the country transitioned from a socialist economy into a more market-oriented economy [1,2].The primary aim of new economic mechanism was to reform the agriculture sector (agriculture production to self-reliance), state-owned enterprises (public-sector investment), increase industrialization (promoting foreign direct investment), and support the economic growth (by increasing manufacturing production).Along with the public enterprise mechanism, major steps towards the socialist transition economies are now scheduled for privatization.The private sectors played a significant role in the national economy, contributing 40% of gross domestic product (GDP), and having created 1.3 million jobs in Laos [3].
Laos is one of the socialist transition economies that has had to confront its informal economy, an issue that 85% in the ASEAN Economic Community (AEC) countries have also dealt with.According to the National Human Development Report [4], informal workers continued to rise from 70.1% in 2005 to 79.4% in 2010, and then to 85% in 2015.These informal workers are categorized into several segments: 49% employed in household sectors, 43% employed in contemporary work, and only 8% were in the outsourcing of manufacturing production.With regard to the skill level of laborers, these workers were 85.2% unskilled, 8.5% skilled workers, and only 6.3% professional skilled.In 2010, the industrial sectors in Laos were downsized in order to reduce the exploitation of workers in informal economies.Regarding the overall exploitation of laborers, more than 10,000 in 2010 were exploited, which then increased to 12,000 in 2015, and is expected to increase to over 15,000 workers in 2020 [5,6].
The structure of employment in Laos changed during the transition from the formal to the informal economy.According to Bourdet [7], informal workers were unprotected by existing labor law, especially regarding social security, labor protection, welfare, and health care.The underlying impetus for the change was linked to employment conditions that have not reverted back to their earlier employment in the formal sectors.These workers were highly skilled laborers, but they cannot access employment opportunities, they face risk in subcontracting, and are vulnerable to context.Dickerson and Green [8] explained that the informal economies are more available to workers, but they must battle against employment insecurity, uncertain incomes, and poor working conditions and livelihoods.A large informal economy, characterized by poor livelihoods due to its access to jobs, faced the potential for productivity shock, seasonal employment, and vulnerability contexts [9].
These findings are extended by Rahman et al. [10] who point out that informal economy workers used their assets for seeking new employment opportunities.In addition, the sustainable livelihood (SL) framework investigated the relationship between assets in suitable livelihood strategies [11][12][13][14][15].For example, their assets may be sequentially related, and the workers generated social assets (trust, support, participation, and relationship) to create their employment opportunity, which may increase their financial assets (incomes, saving, and debt).These collectively comprised their livelihood security.Indeed, Arun et al. [16] identified that assets are transformed into social asset accumulation, which motivate the human assets and may increase the financial assets.
Despite the theoretical rigor in the existing literature and the methodological robustness of the data analysis, no previous studies have confirmed that the assets positively and significantly influence livelihood security.However, Kawaguchi et al. [17] hypothesized that social trust, social interaction, social relation, and social participation are positively correlated with their incomes.This study highlighted that social asset accumulates to build the trust of workers and results in creating employment activities [18].As Giza et al. [19] found, high levels of social assets have a strong positive effect on informal workers' financial assets.Emerson [20] and Chen et al. [21] demonstrated that social assets have a significant positive effect on income returns and results in worker wellbeing.Creed and Watson [22] and Bardasi and Francesconi [23] suggest that social assets in typical employment would be associated with the benefits of wellbeing.
The objective of this study is to examine how CFA of informal worker assets influence livelihood securities.First, the study measured the household characteristics of workers in the informal garment sectors.Second, the study examined the social assets associated with contributions to livelihood security.Third, the study specified financial assets as most important for workers' livelihoods.
The remainder of this paper is structured as follows.It introduces the informal economy worker assets of Laos in Section 1.The literature and hypothesis development are in Section 2. We describe the experimental data and the methods used to fit the data in Section 3, followed by the results in Section 4. The results are discussed in Section 5. Section 6 provides a brief conclusion.

Household Characteristics
Household characteristics have most often been used as units of analysis with variables such as age, gender, marriage, education, and family size [24].Previous authors [25,26] have explained the basic household living standard chosen is socioeconomic status, which includes occupation, education, household size, and age.Further strong factors measure gender, age, average skill levels of workers, and family sizes [27].For example, higher-wealth households are indexed using incomes, saving, consumption, and employment status [28].Some authors [29] used occupation levels, education, number of workers, and family sizes.These studies illustrate the influence of socio-demographics using gender, marriage status, and young workers [30,31].

Social Assets
The existing empirical literature examining social assets [32,33] also included informal economy workers to create a measure for the trust of workers and employment relationships.Crudelia [34] identifies that social assets are based on community support and participation to create neighboring relationships.In the collective-good view of social capital, Uzzi [35] finds that family support, interaction, and relationship feed into higher levels of generalized trust among workers employed in household sectors.Brisson et al. [36] and Woolcock et al. [37] demonstrated the importance of the adoption the workers' trust domain of social assets to create employment activities.As observed by Villalonga-Olive et al. [38] and Leventhal et al. [39], analytical explanations for the accumulation of social capital are based on the measure of long-established employment wellbeing of informal economy workers.

Financial Assets
The factor structure of financial assets is divided into three main ideas: incomes (income and sources), accumulation (savings), and consumption (debt and loan) [40].Reviewing this issue, Ribeiro et al. [41] found that employment activities create financial assets (income, wage-employment, and savings).It identified the financial items as wages (income) that can be accumulated (savings) to generate benefits comprising livelihood security [42,43].Chi [44] illustrated that employment activities, like skilled labor to produce high-income earners, accumulates assets as a means of livelihood.Morales et al. [45] used employment as the basis of workers' incomes and livelihoods.These factors are used in Stikeleather [46], who found benefits from offering workers a financial reward like earning income, wages, working hours, and the ability to earn based on piece-rate.

Livelihood Security as Wellbeing
Livelihood security as a form of wellbeing was not adequately studied in the past, which was empirically defined to measure household wellbeing [21,47,48].Bardasi and Francesconi [23] in their study provided an illustration of the influence of employment wellbeing on livelihood security.Similarly, Mokhtar and Husniyah [49] determined stress, knowledge, and behavior was significantly associated with financial wellbeing.Some studies also highlight the importance of asset building, which could otherwise aid worker incomes and employment wellbeing [21,22,39].Promphakping [50] identified these factors in the choice of workers regarding how to earn a living and therefore shape their wellbeing.

Data Collection
Data were obtained from January 2017 to January 2018 among garment workers in Xaythany, Vientiane, Laos.The study recruited participants from three groups, including kinship workers, local workers, and female post-industrial workers who are currently employed in family garment entrepreneurial sectors.The criteria for selecting the participants were: (i) registration account; (ii) occupational groups; (iii) current employed in family garment entrepreneurial; and (iv) lived in Xaythany, Vientiane.All procedures were approved by the Khon Kaen University Ethics Committee for human research of social and anthropology studies based on the Declaration of Helsinki (No. HE593008).To ensure the validity of the structural schedule interview, the authors selected three experts for testing the pilot.

Measurement
The manifest variables in this study were household characteristics (age and gender), social assets (trust workers, community support, community participations, neighboring relationship, employment relationship, household interaction, household relationship), and financial assets (income sources and incomes).The latent variable of the study was livelihood security, which is designed to measure household wellbeing and employment.These items were used as the observed variable loading in the latent variable of livelihood security.All of the items were graded by five options: 1 (very poor) to 5 (very good) [51].A dummy variable was assigned the number "1" to indicate the presence of attribute and "0" the absence of attribute.From the sample, we pre-tested 30 participants for internal consistency, which was excellent for the total scale (α = 0.924), providing an acceptable correlation between items [52].

Statistical Analysis
The statistical analyses were performed in LISREL software 9.30 for Window of Scientific Software International, Inc., Lincolnwood, IL, USA [53].Before doing CFA, we conducted the Kaiser-Meyer-Olkin test with a resulting value of 0.798 > 0.50 and Bartlett's test of sphericity (p = 0.001) for sampling adequacy.In developing the initial model, Bentler [54] compared the standardized coefficients (β), along with the standardized solution, and the t > 1.96, which means * p < 0.05; t > 2.58 means ** p < 0.01.According to Pesaran and Smith [55], the criteria of R 2 scores were greater than value 0.75, which was accepted in the testing of the CFA model.The validity of the model was tested for initial fitness and is shown in Table 1.

Results
We analyzed the factors, 12 manifest variables, and 2 latent variables within a CFA conducted using the LISREL 9.30 software.Table 2 shows the descriptive statistics for the average scale of the items included in the analysis.Factor 1 included two items, factor 2 included eight items, factor 3 included two items, and the latent included two items.Overall, the findings in this study have data with a goodness of fit.The CFA showed that χ 2 /df = 62, CFI = 0.968, GFI = 0.969, and RMSEA = 0.0358, respectively.The squared correlation 88% (R 2 = 0.880) showed that there was a positive influence on livelihood security.Table 3 shows the factor correlation between assets and livelihood security.The household characteristic score was (0.297, p < 0.01), social assets (0.829, p < 0.01), and financial assets (0.198, p < 0.01).The main total standardized effect is interpreted in Table 4.We further investigated the hypothesis that assets positively influence livelihood security.The findings confirmed that the household characteristics (β = 0.429, p < 0.01), of which direct effects accounted for 0.371, positively influenced livelihood security; therefore, (H 1 ) was fully supported.Regarding social assets (β = 0.505, p < 0.01) of which direct effects account for 0.436; therefore, (H 2 ) was supported.Lastly, the financial assets (β = 0.423, p < 0.01) with a direct effect of 0.365, means that (H 3 ) was fully supported.The final CFA model is summarized in Figure 1 and Table 5 Source: For the rows, 1= household wellbeing; 2 = employment wellbeing.For the columns, 1 = household characteristics; 2 = social assets; 3 = financial assets.This table shows, which are the total standardized effect results from LISREL 9.30.All standardized effects are significant at ** p < 0.01.
We further investigated the hypothesis that assets positively influence livelihood security.The findings confirmed that the household characteristics (β = 0.429, p < 0.01), of which direct effects accounted for 0.371, positively influenced livelihood security; therefore, (H1) was fully supported.Regarding social assets (β = 0.505, p < 0.01) of which direct effects account for 0.436; therefore, (H2) was fully supported.Lastly, the financial assets (β = 0.423, p < 0.01) with a direct effect of 0.365, means that (H3) was fully supported.The final CFA model is summarized in Figure 1 and Table 5.

Discussion
The aim of this research was to examine via CFA how the factor structure of assets positively influenced livelihood security.The overall results of the model fit the data well, and all factors are significant at p < 0.01, and have a squared correlation of 88% (R 2 = 0.880).The findings supported the original three structure factors (household characteristic, social assets, and financial assets) that livelihood security and then comprised wellbeing [21,23,32,47,48,50].We confirmed the findings of previous explanatory research [56][57][58] and a summary is discussed subsequently.The results showed that the assets are mostly positively related to each other in various countries (see also References [49,56,[59][60][61][62][63]).Schinckus [57] found that higher social asset accumulation is influenced by employment status and may increase cash holding returns to the incomes comprising their livelihoods.
First, findings indicated that the social assets were significantly higher among informal economy workers that reported increased levels of livelihoods in Laos.It did not, thereby, complement similar evidence in other studies [58,[64][65][66][67] that found that social assets (trust, support, participation, interaction, and relationship) were correlated with employment activities.In a survey by Parizeau [52], it was found that informal workers in Argentina relied on social assets for holding their incomes to enhance the livelihoods.This finding recognizes that social assets are a complex of its workers and yields significant economic returns.Therefore, it could be argued that the relationship of employment social assets was not a function of trust workers; rather, it is more reflective of the context and structure of employment.
As argued by Bebbington and Perreault [68], social assets can be organized and transformed, which are important for workers producing the livelihoods.The evidence also pointed to social asset allocation as another important factor affecting the informal economy workers in Africa.Indeed, following Lyons and Snoxell [69], social assets have long-established the opportunity for building relationships of trust in places of employment, which enables them to ensure livelihoods.Social assets played a central role in the creation of informal economy workers' livelihoods.Lyon [70] criticized the assertion that social assets create trust between workers in active participation and employers, which is an increasingly important factor of employment activities.
Second, results indicated that the financial assets were the most important factors that directly affected livelihood security, supporting the findings of previous studies [40,42,46].This was consistent with Brown and Taylor [71] and Daovisan et al. [72], who indicated that financial assets have a positive impact because of the saving-to-incomes from selling labor in the form of employment activities.Dunn and Holtz-Eakin [73] found that financial assets are a relied-on resource to produce incomes from wage-and-salary exchanges into livelihood security.According to Bruno et al. [42], incomes directly support workers by giving employment wellbeing.However, as illustrated by Erenstein et al. [74], financial assets can be stored, accumulated, exchanged, and allocated to enhance the quality of life.
Despite the positive findings and insights described above, this study contains limitations.Not all assets in the study could be measured.In an informal economy, worker assets may be difficult to define and measure precisely.Writing on livelihoods in Laos, Mclean [75] suggests that the assets are multifaceted factors, which misses all-encompassing power relations.This highlights that there is a need to recognize that measuring assets is often highly contextualized because the specific indicators of informal economies will vary between contexts.In a situation of informal employment, some assets are permanently destroyed; it is a form of assets not fully encompassed by the 'capital' outlines in the sustainable livelihood model [76].
A drawback of this approach is that it did not measurably include the human assets, physical assets, and natural assets.We found a different, but interesting, result in presenting the five types of capital assets, which were social, human, financial, physical, and natural [32,77,78].However, Su et al. [79] argued that informal workers reliance on employment activities differ across their assets did not comprise the livelihoods.Barua at al. [80] noted that the loss of physical assets may increase social assets with higher financial assets.According to Daovisan et al. [72], the natural assets in a socialist country were controlled by public authority.

Conclusions
The findings of CFA confirmed that three factors (household characteristics, social assets, and financial assets) positively influenced livelihood security in Laos.These results were also linked most strongly with social assets via employment activities and may increase financial assets that comprised the livelihood security [32,34,40,44].Also, consistent with Guiso et al. [19], Emerson [20], and Chen et al. [21], the current study confirmed that social assets are an essential resource to build employment belonging with a significant positive effect on income returns to the workers' wellbeing.It is also important that the CFA model requires further investigation regarding the strength of assets in socialist transition economies.Future research should explore the assets (human, physical, and natural) as an additional factor with the potential to be used in model analysis.More research is needed, however, to be grounded within the qualitative and quantitative forms to diminish the gap between theoretical and empirical investigations.

Table 1 .
Identifying the fitness of the model.

Table 2 .
Descriptive statistics of factors included in the analysis.

Table 3 .
Factor correlation between assets and livelihood security.

Table 4 .
Total standardized effect included in the analysis.
Source: For the rows, 1= household wellbeing; 2 = employment wellbeing.For the columns, 1 = household characteristics; 2 = social assets; 3 = financial assets.This table shows, which are the total standardized effect results from LISREL 9.30.All standardized effects are significant at ** p < 0.01. .

Table 5 .
The fitness model included in the SEM results.

Table 5 .
The fitness model included in the SEM results.