Formalisation provides social security and labour law protection; but does it also improve earnings and lift workers out of low-paying work? This chapter analyses four issues. First, it inquires into how common transitions to formal employment are. Second, it analyses whether the chances of transitioning to formal jobs are the same for all types of informal workers. Third, it investigates whether formalisation is indeed accompanied by income improvements for workers, and conversely, whether transitions into informal employment are accompanied by income losses. Finally, it inquires into whether the potential benefits of formalisation accrue to all workers.
Breaking the Vicious Circles of Informal Employment and Low-Paying Work
3. Transitions to and from formal employment and income dynamics
Abstract
Since the 1990s, many countries have achieved certain progress in formalising their workforces. Yet, formalisation gains are often considered to come too slowly. Moreover, progress appears to be regularly challenged or even reversed by various global shocks, such as the COVID‑19 pandemic, wars and conflicts, international trade disruptions, innovative disruptions and the development of new forms of work (OECD, 2023[1]).
Given this, it is instructive to look not only at changes in stocks but also at flows (in other words, transitions) between formal and informal employment over both short and long periods of time, including their magnitude and the types of workers affected by these transitions. What types of workers are more likely to experience transitions into, and out of, formal employment? Importantly, does formalisation provide tangible financial benefits, in terms of better wages and profits, in addition to the longer-term (and sometimes less tangible) benefits of social protection and protection by labour laws? And do such potential benefits from formalisation accrue to all workers? Answering these questions can help shape formalisation policies, as well as broader labour market and social protection policies, with a view to better targeting them towards workers who cannot formalise, for whom formal employment is not sustainable, or who do not enjoy all the benefits of formalisation.
To answer these questions, this chapter uses panel data for four developing and emerging economies: Indonesia, Malawi, Peru and South Africa. These countries were chosen because of the availability of panel data and because they represent examples from different continents and that are at different stages of development. The longitudinal data for these countries were collected during different time periods: the Indonesian data used were collected every seven years between 2000 and 2014; Malawian data were collected every three years between 2010 and 2019; Peruvian data were collected every year between 2016 and 2020; and South African data were collected every two years between 2008 and 2016 (for further details about the data, see Annex 3.A Data description). Because data collection was carried out during different periods across these four countries, and the lengths of time between panel data collection waves differ by country, the results are not directly comparable between the countries analysed. Despite these differences in data collection, several common observations emerge.
Transitions between formal and informal employment remain relatively limited
As a start, this chapter provides an overview of transitions between three labour market states in Indonesia, Malawi, Peru and South Africa: employment in informal jobs, employment in formal jobs and non-employment (which includes both unemployment and economic inactivity) for individuals of working age (15‑65 years). Figure 3.1 shows that formal employment remains limited and is enjoyed by the smallest group of workers in Indonesia, Malawi and Peru. In South Africa, the size of the formal and informal employment groups is relatively similar. The data show that, in all four countries, the stock of formal employment has slightly increased over time, consistent with the general patterns observed in these four countries (OECD, 2023[1]).
Regarding labour market transitions between the labour market states, three observations emerge.
First, in Indonesia, Malawi, Peru and South Africa, the labour market transitions between the three employment states were observed in all three directions: not only did workers move towards formal employment (a general policy objective) but also towards informal employment, and to non-employment (Figure 3.1).
Second, despite these dynamics, the vast majority of workers were not observed transitioning between data collection time periods, whether they were in formal, informal or non-employment.1 In other words, immobility remains the norm, and the three employment states are relatively stable. Yet, in Indonesia and Malawi, over relatively long periods, formal employment is the least stable state of all: in relative terms, fewer workers remain in formal employment compared with the other states over any studied period.
Third, flows towards formal employment are also the smallest compared with flows between the other states. By far the largest volume of transitions occur between informal employment and non-employment. Of the informal workers who transitioned out of informal employment, more than two-thirds transitioned into non-employment. This was especially prevalent in Indonesia and Malawi, where 75.8% of workers in Indonesia and 61.2% of workers in Malawi who transitioned out of informal employment went into non-employment. At the same time, those who move out of unemployment or economic inactivity tend to take up primarily informal jobs. Also, workers who exit formal employment primarily move to informal employment rather than to non-employment. This suggests that informal employment is not only an important entry point into the labour market but also a labour market segment that absorbs workers who cannot find or keep formal jobs.
Figure 3.1. Transitions between three employment states, by country
Note: Data refer to seven-year time periods for Indonesia (three waves covering 2000‑14); three-year time periods for Malawi (four waves covering 2010‑19); one-year time periods for Peru (five waves covering 2016‑20); and two-year time periods for South Africa (five waves covering 2008‑16). Estimates are generated for transitions between two consecutive time periods (e.g. 2000‑07 and 2007‑14 for Indonesia); the average for all time periods by country is reported. Wave-to-wave transitions are reported in Annex 3.B.
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and the Southern Africa Labour and Development Research Unit (SALDRU) (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
In Indonesia…
More than one-half of workers did not change their employment state over any given seven-year period between 2000 and 2014. Formal employment is the least stable state: only 55.9% of formal workers (compared with 72.3% of informal workers and 59.1% of those in non-employment) remained in the same state (Figure 3.1, Panel A). At the same time, 28.7% of formal workers transitioned into informal work, and 15.3% of formal workers transitioned into non-employment; conversely, 6.8% of informal workers transitioned into formal employment, and 20.9% of informal workers transitioned into non-employment. Only 8.2% of non-employed persons transitioned into formal employment, while 32.6% of non-employed persons transitioned into informal employment.
In Malawi…
While many workers remained stationary over the three-year time periods from 2010 to 2019, there was also much activity (Figure 3.1, Panel B). Here, too, formal employment was the least stable state. While 42.0% of formal workers remained in this state, 36.2% became informal and 21.8% transitioned into non-employment; meanwhile, 57.5% of informal workers remained informal, while 18.4% formalised and 24.1% transitioned into non-employment. Among non-employed persons, 77.7% remained in the same state, while 16.9% became informal workers and 5.4% became formal workers.
In Peru…
Year-to-year transitions between 2016 and 2020 were relatively frequent: on average, around 27.0% of persons transitioned into a new employment state (Figure 3.1, Panel C); 74.6% of formal workers remained formal, 73.9% of informal workers remained informal, and 72.9% of non-employed persons remained in non-employment. The majority of persons who changed their employment state transitioned between informal employment and non-employment: 17.1% of informal workers transitioned into non-employment (compared with 9.0% who transitioned into formal work), and 22.2% of persons in non-employment transitioned into informal work (compared with 4.9% who transitioned into formal work). Among formal workers, 15.3% transitioned into informal employment, and 10.1% transitioned into non-employment.
In South Africa…
Biannual transitions between 2008 and 2016 show that informal workers were the most likely to change their employment state, with formal workers and those out of employment being much less likely to change their state between data collection periods (Figure 3.1, Panel D). On average, 73.0% of formal workers remained in formal employment, while 12.7% took up informal work and 14.3% transitioned into non-employment. Meanwhile, 82.8% of those not in employment remained in this state, while 10.0% switched to informal work and 7.2% transitioned into formal employment. Comparatively, only 39.7% of informal workers remained informal, and those who changed employment state had nearly equal chances of transitioning to formal employment (26.3%) or into non-employment (34.0%).
Transition into formal employment is most difficult for those outside of wage employment
Status in employment2 is one of the key sources of heterogeneity among workers, whether formal or informal. Self-employed workers have typically been considered more vulnerable to shocks, as they do not benefit from the standard protections and employment benefits that formal employees have. For them, informal employment is most often not a choice, but a constrained outcome; some of these workers cannot afford not to work but may be continuously looking for formal employment opportunities. In this regard, it is instructive to examine transitions not only between informal and formal employment but also between different employment statuses within informal and formal employment and between employment and non-employment.
Figure 3.2 disaggregates transitions between two consecutive periods of time for employees and for self-employed workers in Indonesia, Malawi, Peru and South Africa. At the start of any given time period, the share of those in formal wage employment is greater than the share of those in informal self-employment only in South Africa; in the three other countries, informal self-employment is the largest segment of the labour market. The figure shows that those in informal self-employment and non-employment have the lowest probability of changing status between data collection periods. By contrast, informal employees have the highest chances of transitioning to formal jobs compared with other types of workers.
Figure 3.2. Transitions between five employment states, by country
Note: In Indonesia, the formal self-employment category is not shown, as the data for these workers are not available. Data refer to seven-year time periods for Indonesia (three waves covering 2000‑14); three-year time periods for Malawi (four waves covering 2010‑19); one-year time periods for Peru (five waves covering 2016‑20); and two-year time periods for South Africa (five waves covering 2008‑16). Estimates are generated for transitions between two consecutive time periods (e.g. 2000‑07 and 2007‑14 for Indonesia); the average for all time periods by country is reported.
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
In Indonesia…
Among informal self-employed individuals, the majority remained in that employment state (66.7%) over any given seven-year interval (Figure 3.2, Panel A). When informal self-employed individuals did change employment states, they were most likely to go into non-employment (23.7%) rather than into informal wage employment (7.3%) or formal wage employment (2.3%).
Almost one-half of formal employees left the formal economy or workforce between data collection periods: only 57.5% remained in formal employment, while those who left formal employment had similar chances of transitioning to informal self-employment (11.5%), informal wage employment (15.2%) or non-employment (15.9%).
Informal employees had the highest chances of transitioning to a different employment state between data collection periods: only 35.4% retained employment as informal employees. 17.4% of informal employees transitioned to formal wage employment, 25.9% transitioned to informal self-employment, and 21.3% transitioned to non-employment.
Lastly, those in non-employment largely remained in non-employment (61.7%). When non-employed individuals transitioned to work, they were more likely to start an informal self-employment activity (18.2%) than to become an informal employee (11.7%) or a formal employee (8.4%).
The formal self-employed category could not be analysed for Indonesia, as the data for these workers are not available.
In Malawi…
The formal self-employed were most likely to transition to either informal self-employment or non-employment: while 14.7% did not transition out of formal self-employment between three-year periods, 58.9% turned to informal self-employment and 20.2% transitioned into non-employment in the next period (Figure 3.2, Panel B). There was an extremely low probability of transitioning to formal wage employment: only 0.6% of formal self-employed individuals secured such a position between data collection periods.
In contrast, informal self-employed individuals had the highest chances of all worker types to fall into non-employment: 51.3% of informal self-employed individuals became unemployed or economically inactive between data collection periods. They were more likely to fall into non-employment than to retain their informal self-employment activity: only 33.5% did not transition to a different state between data collection periods. Informal self-employed individuals were even less likely to formalise or to switch into wage employment: only 2.5% became informal employees, 12.1% became formally self-employed, and 0.5% became formal employees.
Informal employees had similar chances of transitioning to other forms of work if they did not remain in their current employment status (28.4%) or fall into non-employment (12.4%): 19.0% became formal employees, 19.1% began formal self-employment activities and 21.1% began informal self-employment activities.
Formal employees were comparatively the least likely to fall into non-employment (8.3%) and were more likely to retain their status as employees, whether formal (32.3%) or informal (34.2%). When formal employees transitioned out of wage employment, they were more likely to turn to informal self-employment (20.9%) than to formal self-employment (4.3%).
While those in non-employment were most likely to remain in non-employment (69.4%), transitioning to employment usually meant starting their own informal self-employment activities (22.6%). They had very low chances of breaking into the labour force beyond that: 4.5% started their own formal self-employment activities, while 2.8% became informal employees and 0.7% became formal employees.
In Peru…
With the exception of informal employees, the majority of workers in the Peruvian sample did not transition to a different state between data collection time periods (Figure 3.2, Panel C).
Informal employees had a 44.7% probability of remaining in informal wage employment; where they did transition to a different state, they were most likely to transition to either informal self-employment (21.2%) or non-employment (20.3%). Otherwise, 11.8% transitioned into formal paid employment, while 2.0% had started formal self-employment activities by the next data collection period.
The formal self-employed largely remained in formal self-employment (59.5%). When they did transition, they were most likely to go into informal self-employment (16.5%) or non-employment (11.2%), rather than informal wage employment (6.2%) or formal wage employment (6.6%).
The informal self-employed also largely remained in this category of activity (68.4%) between data collection time periods, and when they did change state, they were more likely to transition to non-employment (15.6%) than to formal self-employment (3.1%), informal wage employment (9.3%) or formal wage employment (3.7%).
Formal employees were the most likely to retain their activity type (74.8%) between data collection time periods, although those who did change state had higher chances of falling into non-employment (9.8%) than transitioning to other categories of activity: 6.7% became informal employees, 6.7% became informal self-employed individuals, and only 2.0% became formal self-employed individuals.
Those in non-employment also faced a high probability of remaining in this state (72.9%) between data collection time periods, and like in the other countries, were most likely to transition to informal self-employment (13.3%) when they were able. They had low chances of transitioning to informal wage employment (8.9%), formal wage employment (4.2%) and formal self-employment (0.7%).
In South Africa…
Again, informal employees had the highest probability of transitioning to a different state rather than remaining informal employees: while 27.4% remained in informal wage employment, 34.0% moved into formal wage employment and 29.1% transitioned into non-employment (Figure 3.2, Panel D). There was a relatively lower chance that they moved into formal self-employment (2.4%) or informal self-employment (7.1%).
Similarly, those in formal and informal self-employment were highly likely to change their employment states as well. Only 35.1% of formal self-employed workers remained in this category of activity, with similar chances of moving into formal wage employment (17.4%), informal self-employment (20.3%) or non-employment (20.2%). Only 7.0% became informal employees. Informal self-employed individuals faced a similar retention rate (31.0%), but were far more likely to transition into non-employment (38.8%) than into other types of employment. They had similarly low chances of transitioning to formal wage employment (12.4%), informal wage employment (13.8%) and formal self-employment (4.1%).
Formal employees and those in non-employment had the lowest probability of changing their employment state. Among formal employees, 72.9% remained in this employment state between data collection time periods. The most probable transition out of formal wage employment was to non-employment (13.9%) compared with informal wage employment (8.9%), informal self-employment (2.8%) or formal self-employment (1.5%) wage employment. Persons in non-employment had an even higher probability of remaining in unemployment or economic inactivity (82.8%), and very low chances of becoming formal employees (6.7%), informal employees (5.9%), informally self-employed (4.1%) or formally self-employed (0.5%).
Transition probabilities also differ depending on individual characteristics and location
Looking beyond status in employment, several other factors can affect transitions between employment states.
Labour markets are segmented not only by informal employment but also by gender
One such factor is gender. Many countries have gendered societal expectations around work (OECD, 2019[15]), which often limit women’s job-seeking behaviour and reduce opportunities for women to find or create their own employment.
At the start of any transition period in all four countries considered in this chapter, there is a greater share of women not in employment compared with men. When women do work, they are also less likely than men to work formally. The percentage point difference in formal employment rates between women and men ranges from 7.2 percentage points in Malawi to 8.9 percentage points in Indonesia, 11.8 percentage points in Peru and 13.9 percentage points in South Africa (Figure 3.3 to Figure 3.6).
When looking at labour market transitions, two observations emerge (Figure 3.3 to Figure 3.6).
First, there is a greater share of women, compared with men, who change their labour market state over a given period in all four countries. More men than women remain in the same labour market state.
Second, the most likely transitions for women are between informal work and non-employment (in both directions), rather than between either of these and formal jobs.
In particular, women have lower chances compared with men of moving to formal employment from any other status (whether informal employment or non-employment), in any country and over any given period. At the same time, women have greater chances compared with men of moving to non-employment from any other employment state in any country and over any given period.
When women move to formal employment, they are more likely to make this transition from informal employment rather than from non-employment. This suggests that, even if these possibilities are limited, some informal jobs can represent a stepping stone into formal employment for women. For men, this pattern is observed only in South Africa, but not in the other three countries examined in this chapter.
Finally, when women move out of informal employment, they are more likely to transition to non-employment rather than to a formal job. This pattern does not hold for men in Malawi, Peru and South Africa, who are equally likely to move from informal employment into non-employment or to a formal job.
Figure 3.3. Labour market transitions in Indonesia, by gender
Figure 3.4. Labour market transitions in Malawi, by gender
Note: Pooled labour market transitions by gender. Data refer to three-year time periods for Malawi (four waves covering 2010‑19).
Figure 3.5. Labour market transitions in Peru, by gender
Note: Pooled labour market transitions by gender. Data refer to one-year time periods for Peru (five waves covering 2016‑20).
Source: Authors’ calculations based on data from Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020).
Figure 3.6. Labour market transitions in South Africa, by gender
Note: Pooled labour market transitions by gender. Data refer to two-year time periods for South Africa (five waves covering 2008‑16).
Source: Authors’ calculations based on data from SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
Other worker characteristics – including educational attainment and place of residence – affect transitions
Besides status in employment and gender, many other factors affect labour market transitions. For example, it is well known that young workers (aged 15‑24 years), workers living in rural areas and workers in some specific sectors (such as agriculture) are particularly vulnerable to informal employment (OECD, 2023[1]). Moreover, several factors can reinforce each other, such that workers in rural areas and in agriculture would not only have a particularly high risk of working informally but also the lowest probability of moving out of informal employment and into formal employment.
Multivariate analysis can help with understanding which individual worker characteristics have an independent significant effect on the probability of workers making certain transitions. It shows that, in Indonesia, Malawi, Peru and South Africa, in addition to working as an employee, having higher educational attainment and living in an urban area substantially increase one’s chances of transitioning to formal employment (Box 3.1). Age and sector of activity can play either a favourable or an unfavourable role, depending on the country. These factors have an almost symmetric (and opposite) effect on an individual’s probability of transitioning from formal employment into informal employment. The best guarantee of remaining in formal employment is to work as an employee, have higher educational attainment and live in an urban area.
The results presented in this chapter complement evidence from other country studies, which also suggests that transitions do not affect all workers equally. For example, in Nigeria, where transitions to and from formal employment are quite frequent, the probability of moving from informal to formal employment is substantially higher for relatively well-paid wage workers compared with lower-income workers (Folawewo and Orija, 2020[16]).
Box 3.1. Who is more likely to experience transitions into and out of informal employment?
Labour market transitions to and from informal employment are possible. However, the probability of making these transitions is not the same for all workers.
Multivariate analysis of panel data from Indonesia, Malawi, Peru and South Africa shows that education is among the key determinants of transitions. Workers with secondary and tertiary education in all four countries are significantly more likely to move to formal jobs than workers with no schooling are (Table 3.1, columns 1‑4). Moreover, the effect of tertiary education on the probability of making such transitions is twice as high in magnitude compared with the effect of secondary education. In South Africa, even primary schooling makes a difference compared with no schooling at all: workers with primary schooling have a significantly higher chance of transitioning to a formal job than workers with no education. In all four included countries, employees have a greater chance of becoming formalised (with either the same or a different employer) compared with self-employed workers. Also, in all four countries, living in urban areas substantially increases the probability of making a transition to a formal job compared with living in rural areas. Women have a significantly lower chance of transitioning to formal jobs compared with men in all countries except Malawi, where gender does not seem to play a role in the likelihood of this transition. Age and sector of activity can play either a favourable or an unfavourable role, depending on the country. All of these factors have an almost symmetric (and opposite) effect on the probability of transitioning from formal jobs into informal employment. Better education remains the best insurance against informal employment, as does working as an employee or living in an urban area (Table 3.1, columns 5‑8).
Table 3.1. Estimation coefficients from logistic regressions examining determinants of transitions
Note: In columns 1‑4, the samples are restricted to workers who were informal in the initial year of the longitudinal data collection. Dependent variables: dummy equal to 1 if transition to a formal job occurred, and 0 otherwise. In columns 5‑8, the samples are restricted to those workers who were formal in the initial year of the longitudinal data collection. Dependent variables: dummy equal to 1 if transition to an informal job occurred, and 0 otherwise. Estimation method: logistic regression. All regressions include additional controls for civil status and household size. Coefficients for these variables are insignificant and not reported to save space. Reference categories are as follows: male (for female); no schooling (for primary, secondary and tertiary); rural (for urban); self-employed (for employee); and agriculture and mining (for industry and services). Standard errors are in parentheses. The symbol (***) represents statistical significance at p<0.01, (**) represents statistical significance at p<0.05 and (*) represents statistical significance at p<0.1.
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
In Ghana, South Africa, Tanzania and Uganda, there is a high degree of persistence in the lower income segment of informal employment, with low-paid self-employed workers having particularly low chances of moving not only to formal jobs but also to better-paid informal jobs. If better-paid workers are relatively more mobile, for the vast majority of workers, informal employment is predominant in lower-paid jobs, and hence represents a dead end. Formal employment is also persistent, suggesting a strong duality of the labour market (Danquah, Schotte and Sen, 2019[17]).
In the People’s Republic of China (hereafter: China), informal employment is most common among rural migrant workers. Because these workers usually have a very low level of skills and human capital, and also face the hukou system, transitions to formal employment remain infrequent (Lin, Ye and Zhang, 2020[18]).
In India, there is a strong segmentation of the labour market, manifested by a high persistence of both formal wage employment and low-income informal wage employment. If nearly one-half of all workers change their employment state over a seven-year period, most of this mobility happens either within self-employment (from formal to informal and vice versa) or within wage employment, but not between wage employment and self-employment. Moreover, there is also a substantially higher risk of informal workers moving from the upper to the lower tier of income distribution rather than to formal employment (Natarajan, Schotte and Sen, 2020[19]). The persistence of informal employment among the lowest-earning informal wage workers also suggests that those with limited human capital and low skills are the least mobile. The Indian labour market is also still characterised by a strong gender- and caste-based labour market segmentation (Michiels, Nordman and Seetahul, 2021[20]), wherein women, workers in lower castes, workers with less formal education and rural workers are less likely to formalise compared with men, those in upper castes, workers with more education and urban workers (Natarajan, Schotte and Sen, 2020[19]).
Formalisation improves incomes, although not for the lowest-paid workers
Formalisation brings with it the protection of labour laws and social protection, at least for employees. But does it help lift workers out of low-paid work? And if so, do all workers experience the benefits of formalisation in the same way?
Data for Indonesia, Malawi, Peru and South Africa allow the linking of transitions from informal to formal jobs with various measures of absolute and relative income mobility (see Box 3.2 for definitions). To understand labour income mobility, labour earnings have been analysed in real values and indexed to 2010 values. They include wages for employees and earnings for self-employed individuals.
Box 3.2. Measuring income mobility
There are different ways to measure income mobility. Income mobility can be intergenerational (between parents, children and grandchildren) or intragenerational (changes for the same individual over time). Income mobility can also be absolute or relative. “Absolute mobility” refers to changes in one’s own income compared with oneself or with the previous generation. “Relative mobility” refers to changes in income compared with others within the same generation; it reflects one’s own changes in position in terms of income distribution. One way to measure absolute mobility is to look at income gains and income losses. Following the existing literature, an income gain/loss is considered here to be “large” if it represents an income increase/decrease of 20% or more compared with a previous period. Conversely, an income gain/loss is considered to be “small” if it represents an income increase/decrease of less than 20% compared with a previous period. Relative mobility can be measured by examining positional changes in income distribution for each individual between two consecutive periods. Upward mobility takes place when an individual moves to an upper income quantile. Conversely, downward mobility takes place when an individual moves to a lower income quantile. In addition, from a policy perspective, it may be worthwhile to separately consider the persistence of staying in the bottom quantile (referred to as “sticky floors”), as well as the persistence of staying in the top quantile of income distribution (referred to as a “sticky ceiling”). The persistence of low incomes may have a considerable long-term effect on material deprivation and health, and it may affect skills development and enhance intergenerational transmission of poverty. Coupled with the persistence of high incomes, it may exacerbate inequality and be a threat to social cohesion.
Source: Adapted from OECD (2018[21]), A Broken Social Elevator? How to Promote Social Mobility, https://www.oecd.org/social/broken-elevator-how-to-promote-social-mobility-9789264301085-en.htm.
Figure 3.7. Percentage of workers who experienced income mobility when transitioning to formal employment, or remaining in formal employment, from one period to another
Note: Income data have been indexed to 2010 values (real). Data refer to seven-year time periods for Indonesia (three waves covering 2000‑14); three-year time periods for Malawi (four waves covering 2010‑19); one-year time periods for Peru (five waves covering 2016‑20); and two-year time periods for South Africa (five waves covering 2008‑16). Estimates are generated for transitions between two consecutive time periods (e.g. 2000‑07 and 2007‑14 for Indonesia); the average for the full time period by country is reported.
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
At the aggregate level, data from Indonesia, Malawi, Peru and South Africa show that workers who transitioned from informal to formal employment are likely to experience both absolute and relative upward income mobility. Figure 3.7 shows the percentage of workers who experienced an income change following either a transition to formal employment or remaining in formal employment between data collection periods. In all four countries, there is a higher percentage of workers who saw an improvement in their income after a move to formal employment compared with those who remained informal. Figure 3.7, Panel A shows that, overall, a higher percentage of workers experienced either a small or a large income gain, rather than a loss or no change in income, when moving to formal employment or remaining in it from one period to another. Figure 3.7, Panel B also shows that, in all four countries, a higher percentage of workers moved up at least one income decile, rather than moving down, after transitioning to formal employment. At the same time, the percentage of workers who exited the bottom income decile following formalisation remained extremely small in all four countries. Moreover, some workers remained in the top tier of income distribution; this percentage was greater among workers who remained in formal employment compared with those who moved from informal to formal employment.
Figure 3.8. Impact of transitions from informal to formal jobs on income mobility
Note: A large income gain is measured as an income increase of 20% or more from one period to the next. Figures shown are estimated odds ratios of experiencing an income change when a transition from an informal to a formal job occurs, all other variables being equal. They report odds ratios (rather than the estimation coefficients) of experiencing the outcome. Control variables include sex, education, age, number of household members, marital status, status in employment, sector of economic activity and place of residence (urban or rural). Income data have been indexed to 2010 values (real). See detailed results in (Aleksynska, La and Manfredi, 2023[22]). Data refer to seven-year time periods for Indonesia (1993‑2014); three-year time periods for Malawi (2010‑19); one-year time periods for Peru (2016‑20); and two-year time periods for South Africa (2008‑16).
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016). For details, see also (Aleksynska, La and Manfredi, 2023[22]).
These aggregate results are also verified by regression analysis at the individual level, which controls for the worker’s age, gender, education, marital status, number of household members, employment status, sector of economic activity and place of residence (urban or rural). They show that moving from informal to formal employment significantly increases the odds of large absolute income gains in Indonesia, Peru and South Africa (Figure 3.8, Panel A).3 Moreover, formalisation is also associated with relative income mobility: workers switching to formal jobs also have greater chances of moving up by at least one income decile in Indonesia, Malawi, Peru and South Africa (Figure 3.8, Panel B).
Similar findings are also reported in other studies: transitions from informal to formal jobs are associated with improvements in earnings in India (Natarajan, Schotte and Sen, 2020[19]); China (Lin, Ye and Zhang, 2020[18]); Costa Rica and Nicaragua (Alaniz et al., 2020[23]); and Ghana, South Africa, Tanzania and Uganda (Danquah, Schotte and Sen, 2019[17]). For more details of these studies, see Table 3.2.
Table 3.2. Findings from empirical studies linking income mobility and transitions into and out of informal employment
Note: In a series of papers discussed in this table, analysis of transitions is done not only between formal and informal jobs but also within the category of informal jobs, looking at transitions between upper-tier and lower-tier informal jobs. While definitions vary across these papers, upper-tier informal jobs are considered to offer at least some protection and generally require higher skills compared with lower-tier informal jobs.
Source: Authors’ compilation.
Despite the overall trend of upward income mobility, the benefits of formalisation do not accrue to all workers. The results of the regression analysis for the four countries examined in this chapter show that “sticky ceilings” benefit the highest-paid workers, while “sticky floors” appear to trap the lowest-paid workers.
“Sticky ceilings” are observed in Indonesia, Malawi and South Africa, wherein well-paid formal jobs are taken up by individuals who are already relatively advantaged in terms of income in the informal economy (Figure 3.8, Panel C). Workers who transition from informal to formal employment and land a well-paid job between data collection periods are often already situated at the top end of income distribution before changing employment state.
In addition, “sticky floors” trap the lowest-paid workers in Indonesia, Malawi and South Africa: formalisation does not significantly improve the incomes of those in the bottom income decile. Even more worrisome, in Indonesia and South Africa, it is unlikely that formalisation will result in exiting the bottom decile of income distribution (Figure 3.8, Panel D; odds are below one). Only in Peru do income mobility and formalisation go hand in hand: when workers in the bottom income decile formalise, they are very likely to exit the bottom decile by the next data collection period, as opposed to workers in the other countries with available data.
These findings can have several explanations. One reason for the “sticky ceilings” is that formal jobs are probably not accessible to everyone. To the extent that these results are obtained from the regression analysis, comparing individuals with similar education, richer individuals typically have more advantageous networks and other forms of social capital that provide them with an edge in the labour market (Cleaver, 2005[25]; Rungo and Pena-López, 2018[26]). Thanks to these networks, they may not only get better access to formal jobs but also learn about the benefits of formalisation from their peers and look more proactively for formal jobs. Another explanation is that formal jobs do not exist in abundance across all sectors and occupations. For example, agriculture is a typical example where informal employment is ubiquitous (OECD, 2023[1]).
In addition to this, in order for the self-employed to fully benefit from formalisation, they should first have a substantial endowment of resources, or access to savings and capital. Only under these conditions can formalising their business be attractive and potentially even more profitable.
Finally, the fact that the lowest-paid workers hardly benefit financially from formalisation may also be related to unobservable characteristics of each occupation. Some occupations, such as domestic work or waste picking, may feature low earnings irrespective of whether they are formal or not. In these occupations, formalisation may bring other benefits, such as better working conditions and social protection. Yet, the absence of financial improvement may mean that low-paid informal workers remain stuck in low-paid formal employment, while their employers may also have little financial capacity to raise their income.
Transitioning to informal jobs worsens incomes, suggesting that it is rarely a choice
In many developed countries, labour market transitions affect the likelihood of income gains to a greater extent than they affect the likelihood of income losses. This is because developed countries typically have social protection and other safety nets that can cushion the effects of job loss (OECD, 2018[21]).
In developing and emerging economies, however, where most workers are informal and rarely enjoy income protection when they lose their jobs, becoming unemployed or economically inactive would mean the loss of income. Unemployment compensation may be limited in magnitude and duration, and in some countries may not be provided to workers who lose their jobs. In these cases, taking up informal jobs can act as a means of survival for workers who cannot afford to go without income.
The analysis of panel data for Indonesia, Malawi, Peru and South Africa shows that, on the whole, workers who transition from formal to informal employment are likely to experience both absolute and relative downward income mobility. Figure 3.9, Panel A shows that absolute small and large losses are more prevalent than gains for workers transitioning to informal employment in Peru and South Africa. Limited absolute income changes are observed in Malawi. In Indonesia, workers transitioning to informal employment are likely to enjoy absolute income gains; however, the results for Indonesia are shown over a relatively long (seven-year) period, during which the country as a whole experienced economic growth (OECD, 2021[27]). Even if some workers experienced absolute income gains when moving to informal employment, the percentage of such workers is still lower than among workers moving to formal employment (shown in Figure 3.9, Panel A).
Figure 3.9. Absolute and relative income mobility for workers remaining in or transitioning into informal employment
Note: Income data has been indexed to 2010 values (real). Data refer to seven-year time periods for Indonesia (three waves covering 2000‑14); three-year time periods for Malawi (four waves covering 2010‑19); one-year time periods for Peru (five waves covering 2016‑20); and two-year time periods for South Africa (five waves covering 2008‑16). Estimates are generated for transitions between two consecutive time periods (e.g. 2000‑07 and 2007‑14 for Indonesia); the average for the full time period by country is reported.
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
Figure 3.9, Panel B also shows that, among workers who experienced informalisation, there is a greater percentage of workers moving down at least one income decile compared with those moving up at least one decile. Moreover, in all countries, some workers entered the bottom income decile following informalisation.
These results are again verified in the regression analysis at the individual level, which controls for the same worker’s characteristics as the regressions (in Box 3.1), i.e. age, gender, education, marital status, number of household members, employment status, sector of economic activity and place of residence (urban or rural). They show that workers switching to informal employment are indeed more likely to experience a large income loss, especially in Indonesia (Figure 3.10, Panel A), or move down at least one income decile in income distribution in Indonesia, Peru and South Africa (Figure 3.10, Panel B). They are unlikely to remain in the top income decile in Indonesia, Peru, and South Africa (Figure 3.10, Panel C). They also have higher chances of entering the bottom decile of income distribution in Peru and South Africa (Figure 3.10, Panel D).4
Figure 3.10. Impact of transitions from formal to informal jobs on income mobility
Note: A large income loss is measured as an income decrease of 20% or more from one period to the next. Figures shown are estimated odds ratios of experiencing an income change when a transition from a formal to informal job occurs, all other variables being equal. Control variables include sex, education, age, number of household members, marital status, status in employment, sector of economic activity and place of residence (urban or rural). Income data have been indexed to 2010 values (real). See detailed results in (Aleksynska, La and Manfredi, 2023[22]). Data refer to seven-year time periods for Indonesia (2000‑2014); three-year time periods for Malawi (2010‑19); one-year time periods for Peru (2016‑20); and two-year time periods for South Africa (2008‑16).
Source: Authors’ calculations based on data from the RAND Institute (2000[2]; 2007[3]; 2015[4]) (the Indonesian Family Life Survey 2000, 2007 and 2014); the Malawi National Statistical Office (2010[5]; 2013[6]; 2016[7]; 2019[8]) (the Malawi Integrated Household Survey 2010, 2013, 2016 and 2019); Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020); and SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016). For details, see also (Aleksynska, La and Manfredi, 2023[22]).
Part of what explains these losses is that informal jobs typically pay less than formal jobs. As seen earlier in this report, the ratio of formal earnings to informal earnings can range from 1.3 to 4.5 in Key Indicators of Informality based on Individuals and their Households (KIIbIH) countries. These losses are also likely due to the fact that, in the absence of social protection, workers who lose formal jobs take any informal job that is available, often with a poor match to their skills and below their productive capacity. They may also switch sector or occupation, downgrading to lower-paid ones.
The benefits of formalisation vary for different groups of workers and by country
Workers with different characteristics do not necessarily benefit from transitions to formal employment in the same way. Gender, age and level of education may not only affect transitions to formal employment differently, as shown earlier in this chapter, but they may also affect the distribution of income gains and losses in the process of (in)formalisation. Indeed, a more disaggregated regression analysis is possible for Indonesia, Peru and South Africa. It shows the following findings for different groups of workers.
First, women are likely to have a higher magnitude of benefits and losses in absolute income, compared with men, following a transition between formal employment and informal employment. When transitioning to informal employment, women have higher odds of experiencing large absolute income losses compared with men (in Indonesia, Peru and South Africa, as depicted in Figure 3.11 to Figure 3.13, Panel A in each figure). When transitioning to formal employment, they are also more likely than men to experience a large absolute income gain (this holds for Indonesia and Peru, although not for South Africa; Panel C in each figure). However, men have higher odds of experiencing a downward change in income quintile when transitioning from a formal to informal job in Indonesia and South Africa, although not in Peru (Panel B in each figure).
Second, different income mobility patterns were also found across age groups. The probability of absolute labour income losses and gains, as well as of relative income mobility, is more amplified for younger workers than for older workers in Peru (Figure 3.12) and in South Africa (Figure 3.13). The opposite is true in Indonesia: transitions into and out of formal employment have greater income effects for older workers than for younger workers (Figure 3.11).
Finally, with regard to educational attainment, the results differ substantially across countries. In Indonesia, workers with no education or only primary education have significantly greater chances of experiencing absolute and relative income losses when transitioning to informal employment compared with more educated workers (Figure 3.11). Workers with secondary or tertiary education have higher chances of improving their absolute income when transitioning to a formal job compared with workers with less schooling. In Peru, workers with secondary or tertiary education compared with workers with no schooling have greater chances of both income gains and losses, in both absolute and relative terms, when experiencing transitions (Figure 3.12). In South Africa, workers with more education have greater chances of experiencing income losses when transitioning to informal employment, while workers with less education have greater chances of experiencing income gains after formalising (Figure 3.13).
Figure 3.11. Distribution of benefits and losses from (in)formality transitions across different profiles of workers: Indonesia
Note: A large income gain/loss is measured as an income increase/decrease of 20% or more from one period to the next. Figures shown are estimated odds ratios of experiencing an income change when a transition from an informal to formal (or from a formal to informal) job occurs, all other variables being equal. Control variables include sex, education, age, number of household members, marital status, status in employment, sector of economic activity and place of residence (urban or rural). See detailed results in (Aleksynska, La and Manfredi, 2023[22]). Data refer to seven-year time periods (2000‑14).
Figure 3.12. Distribution of benefits and losses from (in)formality transitions across different profiles of workers: Peru
Note: A large income gain/loss is measured as an income increase/decrease of 20% or more from one period to the next. Figures shown are estimated odds ratios of experiencing an income change when a transition from an informal to formal (or from a formal to informal) job occurs, all other variables being equal. Control variables include sex, education, age, number of household members, marital status, status in employment, sector of economic activity and place of residence (urban or rural). See detailed results in (Aleksynska, La and Manfredi, 2023[22]). Data refer to one-year time periods (2016‑20).
Source: Authors’ calculations based on data from Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020).
Figure 3.13. Distribution of benefits and losses from (in)formality transitions across different profiles of workers: South Africa
Note: A large income gain/loss is measured as an income increase/decrease of 20% or more from one period to the next. Figures shown are estimated odds ratios of experiencing an income change when a transition from an informal to formal (or from a formal to informal) job occurs, all other variables being equal. Control variables include sex, education, age, number of household members, marital status, status in employment, sector of economic activity and place of residence (urban or rural). See detailed results in (Aleksynska, La and Manfredi, 2023[22]). Data refer to two-year time periods (2008‑16).
Source: Authors’ calculations based on data from SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
Policy discussion
This chapter presented, and in some cases reiterated, several findings regarding labour market transitions and subsequent income mobility transitions. These findings call for several policy recommendations.
In terms of labour market transitions…
This chapter has shown that most individuals do not change their employment state between consecutive periods of observation, whether they are formally employed, informally employed or non-employed. This suggests that, at least in the four countries examined in this chapter, there is strong labour market segmentation, with barriers existing between various employment states. Immobility rates among these three employment states are by far the highest compared with transition rates for any given state.
When workers do transition, they are most likely to move between informal jobs and non-employment, with lower chances of transitioning to formal employment. This suggests that informal employment represents an important possibility for participating in the world of work, especially when formal employment opportunities are scarce, and staying out of employment does not enable people to sustain a living.
In order of magnitude, flows towards formal jobs are smallest when compared with flows to other employment states. Formal employment is especially difficult to retain in countries with lower levels of development, and over longer periods of time. In Indonesia and Malawi, over relatively long periods, more formal workers transitioned into informal employment or non-employment than remained in formal employment. In contrast, retention of formal workers was relatively higher in Peru and South Africa, albeit over shorter periods of time. In all countries with available data, flows from all states to informal employment are most frequent. These findings explain why, on aggregate, in many developing and emerging economies the stock of informal employment decreases very slowly, if at all.
Employees, workers with more schooling and those living in urban areas have the highest chances of moving to formal employment. In some countries, even having just primary schooling, compared with no schooling at all, substantially increases the probability of transitioning to formal jobs. Workers with higher education have the greatest chances of transitioning to formal employment, and the lowest chances of transitioning back to informal employment. Self-employed informal workers, workers in rural areas and workers with no schooling have the lowest chances of formalisation.
Based on these findings, four policy recommendations can be drawn.
1. Governments should continue striving towards the creation of more formal jobs, and especially of wage employment, with better pay in all sectors and occupations. For this, a range of policy actions can be envisaged, in line with the International Labour Organization (ILO) Transition from the Informal to the Formal Economy Recommendation, 2015 (No. 204). This includes creating enabling environments for the development of formal employment through a range of co‑ordinated policies to support innovation, formal enterprise and formal job creation, access to credit, labour mobility, and increasing employability of the workforce. Governments and private actors should be encouraged to create more formal employment opportunities in rural areas and across the full range of sectors and occupations, with the possibility to enable access to these formal employment opportunities by the poorest workers (who are likely to have the lowest level of skills).
2. Acknowledging the important role played by the informal economy in providing incomes, governments and other civil actors should not stigmatise the informal economy. Formalisation of an economy may be a long process; as such, enabling workers in the informal economy to access at least some formal employment arrangements, even if they are not fully formalised, should be considered an important outcome in the overall process of formalisation. In addition, poverty eradication and inclusive growth should be seen as the ultimate objective, whereas formalisation should be seen as a means to help achieve these goals but not necessarily an end in itself.
3. Since education remains a key “insurance” mechanism against informal employment, continuous efforts should be dedicated to providing universal and affordable access to education for all children, to ensuring that children remain in school, and to ensuring equal opportunities for children of all backgrounds to access secondary and tertiary education. These efforts are also important in order to break the intergenerational cycle of informal employment (Chapter 5), and should be made in parallel with efforts to create more formal employment opportunities at all skill levels.
4. Given that some workers will never transition to formal jobs, specific targeted actions should be developed for these workers, either to help them transition to formal jobs or to sustain their incomes.
In terms of income mobility…
Transitions to formality are found to generally improve labour incomes, whereas transitions to informal employment are found to worsen them. This can be seen as an expected result, as prior research has similarly found that the labour earnings of informal workers are generally lower than the labour earnings of formal workers. Yet – and this is a new and more nuanced finding – labour income gains from formalisation are not spread evenly. In several countries, formalisation is unlikely to significantly improve the incomes of the lowest-paid workers. In contrast, formalisation helps richer workers to remain in the top income quantile. Informalisation, in turn, increases the likelihood of entering into the lowest income decile in Peru and South Africa.
These findings confirm that informal employment has two tiers, as shown in Chapter 2: a lower tier that is characterised by low earnings, and un upper tier characterised by relatively higher earnings. It is transitions from the upper tier of informal employment to formal employment that are most likely to improve incomes.
Labour income gains from formalisation (and losses from informalisation) also differ for men and women, workers with different skills, and workers of different ages, although results vary across countries. In several instances, income losses from informalisation are significantly greater than income gains from formalisation. Of course, informal earnings do play a role in reducing poverty when the alternative is no earnings at all; but these earnings, on aggregate, are clearly lower compared with earnings in formal employment.
The persistence of low incomes for some workers, despite formalisation, may have a considerable long-term effect on workers’ well-being. It may affect skills development, negatively affect the accumulation of social capital and enhance the intergenerational transmission of poverty. In addition, in the absence of substantial income improvement, formalisation may not be a sufficiently attractive option for those workers who can choose between remaining informal or becoming formal. Coupled with a persistence of high incomes for some other workers despite their employment status, these dynamics may exacerbate inequality, promote social exclusion and become a threat to social cohesion.
These findings also have several policy implications, which can be summarised as follows:
1. Transition from a formal to an informal job – at least to the informal jobs in the lower tier of informal employment – is unlikely to be a choice. As such, governments should not place too much hope in policies aimed at creating incentives to remain in formal employment. Rather, they should promote the development of adequate unemployment support and other social safety nets. The fact that income losses from informalisation are significantly greater than income gains from formalisation provides the basis for a solid economic argument in favour of developing and strengthening unemployment benefit systems as a way of avoiding workers slipping into involuntary informal employment. Active labour market policies should also be developed to reskill workers, redirect them to formal jobs and support relocation.
2. There is a need to promote greater mobility not only from informal to formal jobs but also from low-paying to higher-paying jobs. Several avenues of policy action can be pursued for this. On the one hand, there is a need to create more formal employment opportunities in the higher-paying sectors and occupations. Cultivating a fertile environment for enterprise creation and scaling up in these sectors of activity is one of the potential ways forward. Another area is improving access to credit for informal self-employed workers, including scaling up micro-credit programmes, with the objective of enabling self-employed workers to upgrade their products and services, increase revenue, and expand their production. Workers should be equipped with the right skills in order to access these jobs. Developing skills and accompanying workers in reskilling is thus another avenue, and is discussed in chapters 4 and 5.
3. Other barriers, such as the lack of adequate social capital and networks, make it difficult for low-paid workers to experience large gains from formalisation. Supporting low-paid workers to access better-paid formal jobs may therefore require other additional measures in order for these workers to build their social capital, such as creating more social mixing opportunities (including in education and in housing) in order to increase their chances of networking with the providers of formal jobs.
4. The fact that certain jobs do not lift workers out of low-paid work, whether they are informal or formal, may also be related to the fact that some countries do not have a minimum wage, or that it is not enforced. Indeed, over the time period that is covered by the data presented in this chapter, South Africa did not have a national minimum wage. Sectoral minimum wage existed, but it did not cover all sectors, and because of the differences across sectors, enforcement was weak. The absence or the weak enforcement of a minimum wage might have meant that employers had few incentives to increase wages when formalising workers. In such settings, setting minimum wages, regularly reviewing them jointly with social partners (including informal workers’ associations) so that they reflect well the minimum living standards, and enforcing them, is also one of the ways to ensure that formalisation financially benefits the poorest workers (Berg, 2015[28]). At the same time, care should be exercised not to set the minimum wage too high, so that formalisation can be affordable (OECD, 2008[29]; ILO, 2015[30]).
5. Some workers perform “essential jobs” (such as waste pickers) whose conditions are difficult to improve: it is challenging to increase their work-related earnings, and moving them to other occupations may not be socially desirable. For such workers specifically, redistributive programmes and policies must be put in place to improve their incomes and protect them in case of economic shocks.
6. Finally, the findings in this chapter suggest that, in addition to pay, there is a need to ensure decent employment conditions for all workers. In some occupations, workers may remain low paid, but there should be a general understanding that formalisation can still bring other benefits, such as social security protection, which would benefit not only these workers but also other family members.
References
[23] Alaniz, E. et al. (2020), Transforming informal work and livelihoods in Costa Rica and Nicaragua, UNU-WIDER, https://doi.org/10.35188/unu-wider/2020/857-3.
[31] Alderman, H. et al. (2001), “Attrition in Longitudinal Household Survey Data”, Demographic Research, Vol. 5, pp. 79-124, https://doi.org/10.4054/demres.2001.5.4.
[22] Aleksynska, M., J. La and T. Manfredi (2023), “Transitions to and from formal employment and income dynamics: Evidence from developing economies”, OECD Development Centre Working Papers, No. 349, OECD Publishing, Paris, https://doi.org/10.1787/bc186b3b-en.
[28] Berg, J. (2015), Labour Markets, Institutions and Inequality: Building just societies in the 21st century, International Labour Organization, Geneva, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/---publ/documents/publication/wcms_347249.pdf (accessed on 3 July 2023).
[25] Cleaver, F. (2005), “The inequality of social capital and the reproduction of chronic poverty”, World Development, Vol. 33/6, pp. 893-906, https://doi.org/10.1016/J.WORLDDEV.2004.09.015.
[17] Danquah, M., S. Schotte and K. Sen (2019), Informal work in sub-Saharan Africa: Dead end or steppingstone?, UNU-WIDER, https://doi.org/10.35188/unu-wider/2019/743-9.
[32] Dercon, S. and J. Shapiro (2007), “Moving On, Staying Behind, Getting Lost: Lessons on poverty mobility from longitudinal data”, Global Poverty Research Group, No. GPRG-WPS-075, Global Poverty Research Group, https://ora.ox.ac.uk/objects/uuid:5703a9c0-ed5c-4878-b48f-222b875373f6/download_file?file_format=pdf&safe_filename=gprg-wps-075.pdf&type_of_work=Working+paper.
[16] Folawewo, A. and O. Orija (2020), Informal–formal workers’ transition in Nigeria: A livelihood analysis, UNU-WIDER, https://doi.org/10.35188/unu-wider/2020/903-7.
[24] Funkhouser, E. (1997), “Mobility and Labor Market Segmentation: The Urban Labor Market in El Salvador”, Economic Development and Cultural Change, Vol. 46/1, pp. 123-153, https://doi.org/10.1086/452324.
[30] ILO (2015), Towards a South African Minimum Wage, ILO, Pretoria, https://www.lrs.org.za/wp-content/uploads/2021/01/Towards-a-South-African-National-Minumum-Wage.pdf.
[9] Instituto nacional de Estadística e Informática (INEI) (2020), Encuesta Nacional de Hogares Panel 2016-2020.
[18] Lin, C., L. Ye and W. Zhang (2020), Transforming informal work and livelihoods in China, UNU-WIDER, https://doi.org/10.35188/unu-wider/2020/907-5.
[8] Malawi National Statistical Office (2019), Fifth Integrated Household Survey 2019-2020.
[7] Malawi National Statistical Office (2016), Fourth Integrated Household Survey 2016-2017.
[5] Malawi National Statistical Office (2010), Third Integrated Household Survey 2010-2011.
[6] Malawi National Statistics Office (2013), Third Integrated Household Survey (Short-Term Sub Panel) 2013.
[20] Michiels, S., C. Nordman and S. Seetahul (2021), “Many Rivers to Cross: Social Identity, Cognition, and Labor Mobility in Rural India:”, The ANNALS of the American Academy of Political and Social Science, Vol. 697/1, pp. 66-80, https://doi.org/10.1177/00027162211055990.
[19] Natarajan, R., S. Schotte and K. Sen (2020), Transitions between informal and formal jobs in India: Patterns, correlates, and consequences, UNU-WIDER, https://doi.org/10.35188/unu-wider/2020/858-0.
[1] OECD (2023), Informality and Globalisation: In Search of a New Social Contract, OECD Publishing, Paris, https://doi.org/10.1787/c945c24f-en.
[27] OECD (2021), OECD Economic Surveys: Indonesia, 2021, OECD Publishing, Paris, https://doi.org/10.1787/fd7e6249-en.
[15] OECD (2019), SIGI 2019 Global Report: Transforming Challenges into Opportunities, Social Institutions and Gender Index, OECD Publishing, Paris, https://doi.org/10.1787/bc56d212-en.
[21] OECD (2018), A Broken Social Elevator? How to Promote Social Mobility, OECD Publishing, Paris, https://doi.org/10.1787/9789264301085-en.
[29] OECD (2008), “Declaring Work or Staying Underground: Informal Employment in Seven OECD Countries”, in OECD Employment Outlook 2008, OECD Publishing, Paris, https://doi.org/10.1787/empl_outlook-2008-4-en.
[4] RAND Institute (2015), Indonesian Family Life Survey (IFLS) - Wave 5, Indonesian Family Life Surveys, https://www.rand.org/labor/FLS/IFLS.html (accessed on 22 June 2018).
[3] RAND Institute (2007), Indonesian Family Life Survey (IFLS) - Wave 4, Indonesian Family Life Surveys, http://www.rand.org/labor/FLS/IFLS.html.
[2] RAND Institute (2000), Indonesian Family Life Survey (IFLS) - Wave 3, Indonesian Family Life Surveys, http://www.rand.org/labor/FLS/IFLS.html.
[26] Rungo, P. and A. Pena-López (2018), “Persistent inequality and social relations: An intergenerational model”, The Journal of Mathematical Sociology, Vol. 43/1, pp. 23-39, https://doi.org/10.1080/0022250X.2018.1470511.
[14] Southern Africa Labour and Development Research Unit (SALDRU) (2016), National Income Dynamics Study (NIDS) Wave 5.
[13] Southern Africa Labour and Development Research Unit (SALDRU) (2014), National Income Dynamics Study (NIDS) Wave 4.
[12] Southern Africa Labour and Development Research Unit (SALDRU) (2012), National Income Dynamics Study (2012) Wave 3.
[11] Southern Africa Labour and Development Research Unit (SALDRU) (2010), National Income Dynamics Study (NIDS) Wave 2.
[10] Southern Africa Labour and Development Research Unit (SALDRU) (2008), National Income Dynamics Study (NIDS) Wave 1.
Annex 3.A. Data description
The analysis of this paper is based on the panel data available for four developing and emerging economies: Indonesia, Malawi, Peru and South Africa. For Indonesia, the Indonesian Family Life Survey (IFLS) is used, with three waves, IFLS-3 (2000), IFLS-4 (2007‑08) and IFLS-5 (2014‑15); the interval between each wave is seven years. For Malawi, the Malawi Integrated Household Panel Survey (IHPS) is used, with waves collected in 2010, 2013, 2016 and 2019; the interval between each wave is three years. For Peru, the Peruvian National Household Survey (Encuesta Nacional de Hogares (ENAHO) Panel) is used, with five waves and a one-year interval between each wave, spanning the period from 2016 to 2020. For South Africa, the National Income Dynamics Study (NIDS) is used, with data collected between 2008 and 2017 over five waves and at intervals of two years.
Annex Table 3.A.1. General data description
Source: Authors’ calculations.
The four panel datasets used are unbalanced, meaning that individuals are observed inconsistently from one survey wave to another, and over the entire period covered by all waves. The average wave-to-wave attrition rate ranges from 15.9% in Malawi to 30.6% in Peru, with the average attrition rates lower in the subsamples of the working-age population compared with the general population. They are quite standard (or even lower) compared with other panel data available in developing and emerging economies (Alderman et al., 2001[31]; Dercon and Shapiro, 2007[32]). In contrast, the attrition rates over the total time span of each panel dataset are substantial, thus precluding the use of standard longitudinal analysis techniques, such as survival or duration analysis. Instead, the analysis shows the averages of wave-to-wave pooled data.
Because each country has a different number of survey waves, and a different interval between each wave, we have treated the data for each country separately. Nevertheless, the same principles have been applied for the rest of the analysis.
Workers included in this sample reported information on their employment status at the beginning and at the end of the relevant data collection time period. Given that labour market transitions are the subject of this study, only working-age individuals (aged 15‑65 years) are considered for the analysis. Individuals close to retirement age are included in the Sankey graphs, with some of them moving to the “out of employment” category at the end of any studied period. If these workers remain in employment at the end of the period, they are also included in the regression analysis, but they are automatically excluded from the regression analysis if they move to the “out of employment” category for any reason, including retirement. Samples in the regression analysis are restricted to the employed. The restriction is necessitated by the fact that the information about labour incomes is only available for working individuals and is not available for those not in employment. Consequently, the regression analysis only considers moves between formal and informal employment statuses, disregarding moves into and out of employment. In each country, this sample is further restricted to the individuals with available data on employment status and labour income at the beginning and at the end of each data collection time period, as well as to individuals with non-missing data on other socio-economic characteristics, including age, gender, education, civil status, number of household members, status in employment (employee or self-employed), sector of economic activity and place of residence (urban or rural).
For more details, see (Aleksynska, La and Manfredi, 2023[22]).
Annex 3.B. Transitions between two consecutive waves of data
Annex Figure 3.B.1. Transitions between three employment states, Indonesia
Annex Figure 3.B.2. Transitions between three employment states, Malawi
Annex Figure 3.B.3. Transitions between three employment states, Peru
Source: Authors’ calculations based on data from Peru’s National Institute of Statistics and Information (2020[9]) (Encuesta Nacional de Hogares (ENAHO) Panel 2016, 2017, 2018, 2019 and 2020).
Annex Figure 3.B.4. Transitions between three employment states, South Africa
Source: Authors’ calculations based on data from SALDRU (2008[10]; 2010[11]; 2012[12]; 2014[13]; 2016[14]) (National Income Dynamics Study (NIDS) 2008, 2010, 2012, 2014 and 2016).
Notes
← 1. This does not mean that workers did not change jobs.
← 2. Following the 20th International Conference of Labour Statisticians (ICLS), a new definition of International Classification of Status in Employment (ICSE 18) as well as an International Classification of Status at Work (ICSaW 18) reorganised previously used categories of self-employment and paid employment into two new categories: independent and dependent workers. The data presented in this section were collected prior to the adoption of these new definitions. For this reason, the old classifications are used here: self-employment, which includes employers, own-account workers and contributing family workers; and paid employment, which includes employees. The figures presented in this section examine five employment states: i) workers in formal paid employment, ii) workers in informal paid employment, iii) formal self-employed individuals, iv) informal self-employed individuals and v) non-employed individuals.
← 3. Figure 3.8 reports the odds ratios. The odds ratios are the likelihood that an event will occur, compared with the likelihood that it will not occur, if the condition (measured by the independent variable transitions between labour market states), is met. As such, odds ratios are always positive. Odds ratios greater than one indicate that the event is more likely to occur than it is to not occur. Odds ratios below one indicate that the event is less likely to occur than it is to not occur.
← 4. Figures presented here show the odds ratios. Statistically, the odds ratio is the probability of the event occurring divided by the probability of the event not occurring. It indicates how likely an event is to occur relative to it not happening. To convert from odds to a probability, one needs to divide the odds by one plus the odds. For example, the odds ratio of experiencing a large income gain after transitioning to formal employment in Indonesia is equal to 5.48. This is the same as saying that the probability of experiencing a large income gain is 0.84 (out of 1), which is quite a likely event. In contrast, the odds ratio of exiting the bottom income decile in Indonesia after transitioning to formal employment is 0.72. This is the same as saying that the probability is 0.41 (out of 1), which means that the event is unlikely to happen.