5. FURTHER EDUCATION AND TRAINING

5.1 Introduction

Literacy and numeracy skills tend to be chiefly acquired in the early years of life through the formal educational system, but the learning activities that are undertaken during the remaining years, at work or outside of work, can also contribute to the retention or further development of these skills. Birth cohort studies that follow individuals over their life course have shown that literacy and numeracy skills tend to decline with age (Willms and Murray, 2007), but are less likely to do so if tasks requiring literacy and numeracy skills are regularly undertaken. For this reason, it is interesting to examine the learning activities of workers with relatively low levels of literacy or numeracy.

As part of the Government's strategy to raise literacy and numeracy skills, literacy and numeracy teaching and assessment will increasingly be embedded into vocational training (Tertiary Education Commission, 2008, p.9). In this way, people studying for a certificate in an education setting or participating in industry training will be able to improve their reading, writing and mathematical skills in the course of their vocational learning. This policy direction also provides an impetus for considering what can be learnt from ALL on the further education and training participation patterns of less literate adults at the time the ALL survey was carried out. We focus particularly on the learning activities of workers aged 25 and over, because we are most interested in the 'further' education and training that adults do after they have made the transition from full-time education to employment.

The education and training questions in ALL distinguish between courses or programmes of study that lead to a qualification and courses that do not lead to a qualification. The questions were not restricted to courses completed at work or for work purposes. However, the vast majority of learners said that their main reason for studying towards a qualification or taking a course was a job or career-related reason. This does not necessarily mean that the study or course was related to the job they currently held. Many tertiary students work in part-time or temporary jobs that are unrelated to their course of study, for financial reasons.

Because the reference period for the education and training questions was the previous 12 months, the learning that was recorded in ALL could have been carried out in a school or tertiary institution in the period before the worker was recruited to the job they currently held or in a previous job in a different industry or occupation. However, any type of formal learning has the potential to raise or maintain literacy and perhaps numeracy skills, and the study and course participation rates discussed here give an indication of how many workers in each industry or occupation had recently undertaken some form of structured education or training.

It seems likely that the ALL survey's measure of 'studying for a qualification' will cover participation in provider-based tertiary education courses and participation in workplace-based industry training programmes. Most industry training programmes lead to qualifications at levels 1-4 in the National Qualifications Framework. ALL's measure of 'other courses not leading to a qualification' is likely to cover short courses that are delivered through workplaces and funded by employers, as well as short courses that are undertaken by adults at their own initiative.

Section 5.2 summarises the education and training rates of the working-aged population and workforce as a whole to provide a context for the rest of the analysis. Section 5.3 compares the further education and training rates of workers with low literacy skills and workers with higher literacy skills. Section 5.4 analyses the demographic and job characteristics of the workers who participated in further education or training, to identify factors that may increase the likelihood of participation. Section 5.5 concludes.

5.2 Education and training rates

Compared with other countries that have also conducted ALL, the rate of participation in programmes of study leading to a qualification was relatively high in New Zealand. Details are given in Table 5. Twenty-seven percent of all adults aged 16-65 did some study towards a qualification in the previous 12 months, compared with 16-21 percent in Canada, Norway, Switzerland and the United States. However, average hours of study per participant appear to be lower in New Zealand than elsewhere.^[5] This could be because the courses or study programmes are shorter on average in New Zealand than in other countries or because a higher rate of participation is associated with a higher rate of non-completion.

 

Table 5: Education and training of the working-aged population during the past year - New Zealand compared with other OECD countries

	Study towards a qualification			Other courses
Country	Participation rate (%)	Average hours per participant	Average hours per capita	Participation rate (%)	Average hours per participant	Average hours per capita
New Zealand	27	526	143	31	54	17
Canada	16	595	94	25	63	16
Norway	21	895	185	31	48	25
Switzerland	20	640	125	40	61	15
United States	20	574	114	21	65	14

Note: The data for New Zealand were collected in 2005-07. The data for the other countries were collected in 2003 and sourced from Rubenson et al, 2007.

Evidence presented later in this paper indicates that adults with higher levels of educational attainment are more likely than those with lower levels of educational attainment to participate in further education or training, even when other characteristics, including the individual's literacy skill level, are taken into account. As discussed in Section 2, more highly educated adults appear to be somewhat over-represented in the ALL sample, and it is possible that this feature of the sample composition is contributing to a relatively high measured rate of participation in programmes leading to qualifications.

The rate of participation in courses that do not lead to a qualification in New Zealand, and the average hours spent on such courses by participants, were broadly comparable with the rates reported in the other countries: New Zealand is in the middle of the distribution. Although the average number of hours spent on courses by participants may seem high at 54 hours, this represents the total time spent on up to three courses. The average duration of an individual course was 30 hours, and the median duration was 10 hours.

Adults who were currently employed were less likely than the working-aged population as a whole to have studied towards a qualification in the last 12 months (see Table 6), but they were more likely to have taken a course or courses. Twenty-three percent of the employed did some study towards a qualification,^[6] and 36 percent took a course that was not intended to lead to a qualification. Overall, 52 percent of employed people reported that they participated in a programme of study, a course or both types of learning during the previous 12 months.

Table 6: Education and training rates of working-aged and employed adults, previous year

	Study towards a qualification			Other courses
	Participation rate (%)	Average hours per participant	Average hours per capita	Participation rate (%)	Average hours per participant	Average hours per capita
All working-aged adults	27	526	143	31	54	17
All employed	23	315	73	36	42	15
Employed and aged 25 or over	20	210	41	38	40	15
Employees aged 25 and over
All courses	22	214	47	39	42	17
Courses with employer funding	12	120	15	28	NA	NA

Note: The duration of employer-funded courses is not shown in the table because information on the sources of funding was not collected separately for each individual course.

Per employed participant, the average number of hours of study towards a qualification was 315. Per worker, the average number of hours was 73. To put these figures in context, a full-time full-year educational course is expected to require at least 1,200 hours of study. It is likely that many participants were either studying part-time or for part of the year only.

The rate of participation in study programmes leading to qualifications is slightly lower if the population is restricted to employed people who were aged 25 or over, but not much lower (20 percent rather than 23 percent). However, the average hours of study per participant are considerably lower when 16-24 year olds are excluded (210 rather than 315). The rate of participation in courses not linked to a qualification was slightly higher among workers aged 25-65 than among all workers.

People who did some study or courses were asked to say who contributed to the costs (for example, their employer, the government or themselves). Only employees can receive employer funding, and therefore we restrict the analysis to employees aged 25 years or over and exclude the self-employed. Information on all education and training courses undertaken by employees and the courses they took that were identified as receiving funding from employers is shown in the final two rows of Table 6.

Because employees do not always know how the courses they attend are financed, it is possible that there is some under-estimation of the total volume of employer-funded education and training. Nevertheless, the results indicate that about 55 percent of employees who did qualifications-oriented study or training were partly or fully employer-funded, and around 70 percent of employees who did other courses (not leading to qualifications) received some employer funding.

5.3 Education and training rates of workers with different levels of literacy or numeracy

For the rest of this section, we examine the learning of workers aged 25 years and over, excluding the under-25s as a crude method of excluding those who have not yet completed the first phase of their formal education. We are interested in the 'further' education and training that adults do when they have already made a transition from full-time education to employment.

The relationship between level of literacy or numeracy skills and rates of participation in further education and training, for workers aged 25 years and over, is explored in Table 7. In general, rates of study towards a qualification do not vary greatly by foundation skill level: lower skilled adults were about as likely to have undertaken some study towards a qualification as adults with higher levels of literacy or numeracy skills. Workers with level 1 literacy skills are a partial exception to this statement. The 'volume' of studying time per participant is also fairly constant across skill levels. In other words, average hours of study per participant does not show much of a literacy or numeracy skill gradient.

 

Table 7: Education and training rates of workers aged 25 years and over by their level of literacy or numeracy skill

	Participation rates (%)			Average hours per participant
	Study towards a qualif-ication	Other courses	Study and courses	Study towards a qualif-ication	Other courses
All education and training
All employed	20	38	51	210	40
Level 1 literacy	14	17	28	214	58
Level 2 literacy	21	28	43	194	45
Level 3 literacy	21	44	57	210	37
Level 4/5 literacy	19	50	61	228	39
Level 1 numeracy	18	16	32	192	44
Level 2 numeracy	23	34	50	237	46
Level 3 numeracy	18	42	54	196	38
Level 4/5 numeracy	18	52	63	194	38
Employer-funded education and training
All employees	12	28	38	120	NA
Level 1 literacy	8	14	19	127	NA
Level 2 literacy	13	19	30	105	NA
Level 3 literacy	13	33	43	124	NA
Level 4/5 literacy	13	40	49	129	NA

The absence of a stronger literacy skill gradient in study rates may partly reflect the role of industry training in New Zealand, which provides workplace-based training to workers who typically do not have high levels of education. In 2006, for example, around 176,000 individuals participated in industry training programmes (Tertiary Education Commission, 2007, p.3), of whom 122,500 were aged 25 years or over. Those workers represented about 7 percent of the total number of employed persons aged 25 years and over who were employed on average during 2006.^[7] However, the majority of industry training participants were workers who did not already have post-school qualifications, and the vast majority of the qualifications they gained were at levels 1-4 of the National Qualifications Framework. Workplace-based training opportunities are likely to be particularly important in engaging workers with low foundation skills because much of the learning is undertaken in work time and employers often facilitate the enrolment process.

In contrast to this pattern, workers with low literacy or numeracy skills were substantially less likely to have participated in courses that were not linked to a qualification than workers with higher skills. For instance, only 17 percent of workers with level 1 document literacy skills reported taking part in a course, compared with 50 percent of workers who were at level 4 or 5. This finding is not surprising, as this category of learning is dominated by relatively short courses that are funded by employers (and probably undertaken in work time). Studies of the incidence of employer-funded training that is delivered within workplaces almost universally find that workers with higher levels of educational attainment tend to do more of this type of training. This pattern has also been found in analyses of ALL data for other countries. For example, Rubenson et al (2007, p.37) reports there is a strong positive relationship between the literacy skills of individuals and their rates of participation in organised forms of learning in four countries that carried out ALL surveys (Canada, the US, Switzerland and Norway).

Figure 7: Further education and training participation rates for workers aged over 25, by literacy skill level

Data table for Figure 7

Table 7 also indicates that the proportion of employees who received employer funding for study towards a qualification (which was 12 percent overall) did not vary much by the worker's literacy skill level. In contrast, employees with higher literacy skills were far more likely than employees with lower skills to have taken courses that were not linked to qualifications, with employer funding. Although this disparity in course participation may put workers with low literacy or numeracy skills at a disadvantage when it comes to retaining or developing their skills, the total volume of learning time associated with such courses is considerably less than the volume of learning time associated with qualifications-linked courses.

5.4 Who is more likely to participate in further education or training?

In this section of the paper, we examine the further education and training patterns of workers with relatively low literacy skills in more detail, as well as those of all workers aged 25 and over. The main objective is to better understand which types of worker - viewed in terms of their demographic, educational and job characteristics - are more likely to undertake further education and training, despite having relatively low literacy skills. 'Low literacy skills' is defined here as level 1 or 2 on the document literacy scale.

The education and training rates of employed people aged over 25 are likely to be influenced by a variety of factors including their own preferences and choices, their family circumstances and the learning opportunities that are made available to them, particularly through work. Individuals of different ages and educational levels have different incentives to undertake further education and training. Some occupations and careers require or reward a higher level of continuing education than others. Firm characteristics such as size, profitability, capital intensity and technology will influence both the need and the incentives that employers have to provide training to their workforces. Government subsidies will also influence the incentives that firms in different industries have to provide training.

5.4.1 Education and training rates by worker characteristics - descriptive statistics

We begin by presenting simple summary statistics on the further education and training rates of workers in the ALL survey, disaggregated by their demographic, socio-economic and job characteristics. These summary statistics are set out in Table 8 (studying for qualifications) and Table 9 (participation in other courses). Each table shows the participation rates for all workers on the left and the participation rates for workers with low literacy skills on the right, along with the confidence intervals associated with each rate. For some socio-economic groups, the sample sizes in ALL are relatively small, especially when we focus on the 'low literacy' sub-sample, leading to relatively imprecise estimates of studying and course rates with fairly large sampling errors. We comment only on the differences between population groups that are statistically significant at the 95 percent confidence level.

For all workers aged 25 and over, the descriptive statistics indicate the following:

• Rates of studying for qualifications are similar for men and women, but women were more likely than men to have taken courses that were not linked to qualifications.
• Older workers were less likely to study for a qualification than younger workers. There is not a clear age gradient in rates of undertaking courses that are not linked to qualifications, however.
• Maori were more likely to have studied for a qualification than Europeans. Their rate of undertaking other types of courses was not significantly different from that of Europeans.
• Workers with a Pacific ethnic affiliation were significantly less likely than Europeans to have participated in a course that was not linked to a qualification, suggesting they were less likely to have received employer-funded training.
• Workers who had a tertiary qualification already (including level 1-3 qualifications) were more likely to undertake further education than those who did not. This is true of both study for qualifications and participation in other courses.

 

Table 8: Rates of studying for a qualification, by worker and job characteristics

	All workers aged 25 or over				Workers with low literacy skills (level 1 or level 2)
	Partic-ipation rate	95 percent confidence intervals			Partic-ipation rate	95 percent confidence intervals
		Lower	Upper			Lower	Upper
Sample size	4561				1885
Estimated population size (000s)	1668.4				638.7
All demographic and labour force groups	19.6	18.5	20.6		18.7	16.4	21.1
Gender and age group
Males	19.2	17.6	20.9		19.7	16.3	23.0
Females	19.9	18.7	21.1		17.8	15.1	20.4
Aged 25-34	26.0	23.5	28.5		21.8	16.9	26.7
Aged 35-44	21.2	18.8	23.6		19.0	15.1	22.9
Aged 45-54	16.2	14.6	17.8		18.7	14.8	22.6
Aged 55-65	13.8	10.8	16.8		15.4	10.4	20.3
Ethnic group
European	18.1	17.0	19.3		15.9	12.9	18.9
Maori	31.2	25.9	36.5		30.5	22.6	38.3
Pacific	24.0	17.6	30.4		22.0	15.2	28.7
Asian	20.7	16.8	24.6		20.6	14.1	27.1
Birthplace and language
Born in New Zealand	19.2	17.9	20.5		18.0	15.2	20.9
Recent immigrant	25.6	19.3	31.8		20.7	12.4	29.0
Speaks English as second language	22.0	18.6	25.3		20.4	16.2	24.5
Education
Completed 5th form /year 11 or less	13.7	11.4	16.0		12.2	9.4	15.1
Upper secondary education	15.4	12.1	18.7		11.8	8.5	15.1
Post-school level 1,2 or 3 qualification	24.1	19.4	28.7		26.8	19.4	34.2
Level 4 qualification	20.3	15.5	25.2		19.5	14.1	24.9
Level 5, 6, or 7 qualification	24.3	21.0	27.5		19.7	14.2	25.2
Degree	21.2	19.1	23.3		32.5	24.0	41.0
Job characteristics
Employee	21.9	20.6	23.1		20.4	17.7	23.0
Self-employed	10.9	9.0	12.8		10.8	6.7	14.9
Full-time employed	20.4	19.0	21.8		19.1	16.3	21.9
Part-time employed	16.2	13.2	19.3		17.5	12.3	22.6
Enterprise size (number of employees)
Firm size 1-19 employees	14.7	12.9	16.6		14.0	10.5	17.5
Firm size 20-99 employees	17.7	14.7	20.7		17.9	12.0	23.7
Firm size 100-9999 employees	22.1	19.0	25.2		18.5	12.8	24.2
Firm size 1000 or more employees	26.7	24.1	29.4		28.0	24.4	31.6
Industry
Agriculture, forestry and fishing	16.7	12.8	20.5		14.4	7.9	21.0
Manufacturing	16.7	13.5	20.0		14.1	10.5	17.7
Wholesale and retail trade	11.9	8.4	15.4		7.8	4.1	11.6
Food services and accommodation	20.8	12.2	29.3		12.7	5.6	19.9
Transport and communications	15.0	10.2	19.7		8.8	1.4	16.2
Finance and business services	12.0	9.1	14.9		10.9	4.0	17.8
Public administration and defence	29.5	22.9	36.0		35.0	17.8	52.2
Education and training	26.7	22.5	31.0		40.1	31.9	48.4
Health and community services	32.5	28.1	37.0		33.0	26.4	39.6

Note: Respondents could give more than one ethnic affiliation, and if they did so, they are counted in each applicable ethnic group. Recent immigrants are people who were born outside New Zealand and moved to New Zealand in 2001 or more recently.

 

Table 9: Rates of participation in courses that were not linked to a qualification, by worker and job characteristics

	All workers aged 25 or over				Workers with low literacy skills (level 1 or level 2)
	Partic-ipation rate	95 percent confidence intervals			Partic-ipation rate	95 percent confidence intervals
		Lower	Upper			Lower	Upper
Sample size	4561				1885
Estimated population size (000s)	1668.4				638.7
All demographic and labour force groups	37.7	36.1	39.3		24.2	21.6	26.8
Gender and age group
Males	34.5	32.0	37.1		21.7	18.9	24.6
Females	41.3	39.1	43.5		26.9	23.4	30.4
Aged 25-34	35.7	32.8	38.5		22.6	16.7	28.5
Aged 35-44	38.9	36.2	41.5		25.9	20.5	31.3
Aged 45-54	41.1	38.0	44.2		21.6	17.2	26.0
Aged 55-65	33.2	29.9	36.4		27.2	21.7	32.7
Ethnic group
European	39.5	37.4	41.7		25.5	22.5	28.4
Maori	35.2	30.4	40.0		27.1	20.7	33.5
Pacific	19.8	14.3	25.3		14.1	9.0	19.1
Asian	27.5	20.9	34.1		19.5	12.7	26.3
Birthplace and language
Born in New Zealand	38.6	36.9	40.3		25.2	22.6	27.9
Recent immigrant	34.5	27.1	41.9		20.3	10.9	29.7
Speaks English as second language	26.7	22.1	31.4		19.1	13.0	25.1
Education
Completed 5th form /year 11 or less	23.2	20.5	25.9		17.6	14.5	20.7
Upper secondary education	30.1	25.5	34.7		20.3	13.7	27.0
Post-school level 1,2 or 3 qualification	31.8	26.3	37.3		25.5	17.9	33.1
Level 4 qualification	37.4	32.2	42.5		26.1	17.9	34.2
Level 5, 6, or 7 qualification	46.1	41.7	50.5		30.3	21.8	38.7
Degree	51.0	47.4	54.6		38.3	30.4	46.2
Job characteristics
Employee	39.2	37.6	40.8		24.9	21.9	28.0
Self-employed	31.9	28.4	35.5		20.9	16.4	25.4
Full-time employed	38.3	36.5	40.1		24.4	21.4	27.4
Part-time employed	35.2	30.7	39.8		23.7	18.5	28.9
Enterprise size (number of employees)
Firm size 1-19 employees	29.7	27.5	31.8		17.2	9.7	24.6
Firm size 20-99 employees	42.4	38.6	46.2		16.5	9.2	23.7
Firm size 100-9999 employees	41.2	36.7	45.8		17.4	12.9	21.9
Firm size 1000 or more employees	46.3	42.6	50.0		16.9	5.3	28.4
Industry
Agriculture, forestry and fishing	24.9	19.5	30.4		17.4	10.6	24.1
Manufacturing	29.4	25.7	33.2		16.9	12.1	21.7
Wholesale and retail trade	27.8	23.1	32.4		18.6	13.6	23.6
Food services and accommodation	27.0	20.1	34.0		27.7	16.3	39.1
Transport and communications	31.4	22.1	40.6		20.9	12.5	29.3
Finance and business services	41.5	34.7	48.4		30.4	20.7	40.2
Public administration and defence	57.0	50.0	64.0		36.4	22.8	49.9
Education and training	55.1	51.0	59.1		41.9	32.2	51.6
Health and community services	50.9	45.6	56.3		33.6	26.7	40.5

Note: Respondents could give more than one ethnic affiliation, and if they did so, they are counted in each applicable ethnic group. Recent immigrants are people who were born outside New Zealand and moved to New Zealand in 2001 or more recently.

• Employees were twice as likely to have studied for a qualification as the self-employed. They were also more likely to have taken other courses.
• Workers at larger enterprises (particularly those with more than 1,000 employees but also those with more than 100 employees) were more likely to have participated in both types of further education and training than those who worked at small or medium-sized enterprises.
• Using broadly defined industry groups, the industries with the highest rates of studying towards a qualification were health care and social services, public administration and defence, and education and training.
• For courses that do not lead to qualifications, there is also a clear pattern of higher participation by workers in the three industry groups that have the highest levels of public ownership or public funding - public administration, education and training, and health and community services - than elsewhere in the economy.

If we consider the further education and training rates of workers with low literacy skills (shown in the right-hand columns of each table), most of these general patterns also hold true. This suggests that an understanding of the drivers of participation in further education and training for workers in general is likely to be relevant for understanding the drivers of participation by less literate adults.

5.4.2 Multivariate analysis of further education and training participation patterns

When considering the relationship between any particular characteristic and education or training participation rates, it is important to control for other factors that may also be influencing the probability of studying or training. For example, the higher course participation rate of women may be due to differences in the occupational or industry distribution of women and men, rather than any gender-specific differences in the motivation to learn or the likelihood of being sent on training courses by employers. Women may tend to work in the types of occupations or industries in which there is a high level of emphasis on further education or training.

Binomial logistic regression models were used to explore the association between particular characteristics and studying or training rates, while holding all other variables constant. These regression models use information on the personal and job characteristics of individuals to predict the likelihood of studying or training. Using the model estimates, the impact (or marginal effect) of a change in one characteristic on the chance of participating, while holding all other measured characteristics constant at their mean values, can be estimated.^[8] If there are systematic differences in participation rates by personal and job characteristics, this could reflect differences in the opportunities that are open to workers, although other more direct evidence would be needed to verify whether this is the case.

The likelihood of participation in a) programmes leading to a qualification and b) other courses was modelled. The characteristics that were included in the regression models as explanatory variables were: gender; five-year age group; ethnic affiliation; whether born overseas and recently migrated to New Zealand; whether an ESOL speaker; level of educational attainment; literacy skills measured on the document literacy scale; whether self-employed; whether working part-time hours; firm size; and occupation and industry of employment. Most characteristics are categorical rather than numeric and were included in the regression model using a set of dummy variables. For each characteristic, we omit the group or 'category' whose studying or course participation rate was closest to the all-sample average. The omitted group becomes the reference group against which the results for the other groups are compared.

Summary results are presented in Table 10, and the full results are presented in Tables A8 and A9 in Appendix 3. Here, we summarise the main findings on the statistically significant effects.

Likelihood of studying or training for a qualification - all workers aged 25 and over

Estimates from a regression model of the probability of studying for a qualification that was estimated using the entire sample of workers aged 25 and over indicate the following:

• Workers aged 25-29 years were more likely to study for a qualification than workers aged 30 or over (and 9 percentage points more likely to do so than the omitted age group of 40-44 year olds, after controlling for the effects of other factors). Workers aged 50 and over were less likely to study towards a qualification.
• Workers with a Maori ethnic affiliation (alone or in combination with other ethnic groups) were more likely to study than Europeans (9 percentage points more likely).
• Workers with no qualifications or school-level qualifications only were less likely to have studied (about 6 percentage points less likely than workers with a level 4 qualification - the omitted educational group).
• The self-employed were less likely to have studied towards a qualification than employees (by about 5 percentage points).
• Workers who were part-time employed were less likely to have studied than the full-time employed (by about 5 percentage points).
• Those employed in small or medium-sized enterprises were less likely to have studied than those employed at larger firms (those with 100 or more employees). The estimated participation probability of workers who were employed by firms in the largest size group (1,000 plus employees) is 5 percentage points higher than that of workers in the smallest firm size group (1-19 employees).

Table 10: Marginal effects of worker and job characteristics estimated from the regression models of participation in further education or training

	Studied for a qualification					Other courses
	All workers		Workers with low literacy skills			All workers		Workers with low literacy skills
Female	-0.02		-0.06	*		0.04	*	0.03
Aged 25-29	0.09	*	0.07			-0.06		-0.02
Aged 30-34	0.04		0.00			-0.03		-0.08
Aged 35-39	0.05		0.04			-0.03		-0.04
Aged 45-49	-0.01		0.02			-0.01		-0.08
Aged 50-54	-0.06	*	0.00			-0.01		-0.07
Aged 55-59	-0.03		0.00			-0.09	*	-0.03
Aged 60-65	-0.07	*	-0.04			-0.04		0.01
Maori ethnic affiliation	0.09	*	0.11	*		0.02		0.01
Pacific ethnic affiliation	0.02		0.04			-0.11	*	-0.10	*
Asian ethnic affiliation	-0.02		0.01			-0.06		-0.08
Recent immigrant	0.04		0.01			-0.01		-0.02
Speaks English as second language	0.02		-0.01			-0.06		-0.05
Educational attainment
Completed 5th form/year 11 only	-0.06	*	-0.07	*		-0.11	*	-0.08
Upper secondary school	-0.06	*	-0.08	*		-0.08	*	-0.05
Level 1, 2, or 3 qualification	0.02		0.05			-0.05		0.01
Level 5, 6, or 7 qualification	0.01		-0.02			0.01		-0.01
Bachelors degree	-0.02		0.02			0.00		0.05
Higher degree	-0.03		0.09			0.03		0.10
Document literacy score in ALL/100	0.02		0.04			0.10	*	0.04
Self-employed	-0.05	*	-0.04			-0.01		0.03
Employed part-time	-0.05	*	-0.02			-0.05		-0.03
Firm size 1-19 employees	-0.03		0.00			-0.09	*	-0.12	*
Firm size 20-99 employees	-0.04	*	0.00			0.00		-0.05
Firm size 1000 or more employees	0.02		0.08	*		0.03		0.03
Industry
Agriculture	0.05		0.17			0.00		-0.09
Food manufacturing	0.03		0.06			0.01		-0.09
Other manufacturing	-0.05		0.01			-0.04		-0.12
Construction	0.02		0.14			-0.02		-0.05
Wholesale trade	-0.06		-0.02			-0.09		-0.13
Motor vehicle sales and service	0.01		-0.03			-0.03		-0.08
Retail trade (excluding motor vehicles)	-0.09	*	-0.09			-0.02		-0.09
Accommodation and food services	0.00		-0.02			0.02		0.07
Transport	-0.04		-0.04			-0.02		-0.05
Commmuications	-0.05		-0.03			0.03		-0.13
IT and scientific services						-0.05		-0.08
Business services	-0.01		0.01			0.01		-0.04
Public administration and defence	0.05		0.15			0.10		-0.08
Education and training	0.14	*	0.30	*		0.03		0.02
Health care and social services	0.16	*	0.23			0.12		0.00
Cultural and recreational services	0.01		0.13			-0.02		-0.01
Other services	-0.06		-0.08			0.18		-0.04

Note: The models also contained 18 occupational group controls and controls for 'industry not specified' and 'firm size not specified'. *Indicates that the marginal effect was statistically significant at the 95 percent confidence level. The underlying model estimates are given in Appendix 3.

• Workers in the education and training industry and in health and community services were more likely to have studied for a qualification than those employed in other industries, while workers in retail trade were less likely to have done so.

Likelihood of studying or training for a qualification - workers with low literacy skills

Fewer characteristics showed a statistically significant association with the likelihood of studying for a qualification when the model was restricted to workers with document literacy skills at level 1 or 2. This is partly because the total sample of workers with low literacy skills is relatively small (around 1,900 persons).

Statistically significant effects were found for the following factors:

• Females were less likely to have studied than males by around 6 percentage points.
• Maori were more likely to have studied or trained than Europeans by around 11 percentage points.
• Workers with school level qualifications only or no qualifications were less likely to have studied (7-8 percentage points less likely than workers with level 4 qualifications, the omitted group).
• Those employed in large enterprises with 1,000 employees or more were more likely to have studied or trained than workers at enterprises in any other size group. The estimated participation probability of workers employed by firms in the largest size group (1,000 plus employees) is 8 percentage points higher than the estimated participation probability of workers in the smallest firm size category (1-19 employees).
• Workers who were employed in the education and training industry were more likely to have studied or trained than those in other industries. Their estimated participation probability is 30 percentage points higher than that of workers in finance and insurance, the omitted industry group. It is possible that tertiary students who hold part-time jobs at the institution where they study are contributing to the high education and training participation rate of workers in this industry.

Likelihood of taking other courses - all workers aged 25 and over

Estimates from the regression model of the probability of taking a course that would not lead to a qualification, estimated using the entire sample of workers aged 25 and over, suggest the following:

• Females were more likely to have taken a course than males. However, the estimated gender difference in probability (4 percentage points) is smaller than the gender difference in the unadjusted course participation rates (7 percentage points), indicating that other correlated factors such as job characteristics were contributing to the overall difference in the descriptive statistics.
• Workers with a Pacific ethnic affiliation were less likely to have taken a course than Europeans, by around 11 percentage points. This is smaller than the 20 percent difference between Pacific and European workers found in the unadjusted course participation statistics, suggesting that correlated factors such as educational attainment or job characteristics were contributing to the total course participation gap between Pacific and European workers.
• Workers with no qualifications or school level qualifications only were significantly less likely to have taken a course than workers with a post-school qualification. The lowest educational attainment group was 11 percentage points less likely to have taken a course than workers with a level 4 qualification, the omitted educational group, and 14 percentage points less likely than workers with a higher degree.
• Workers with higher document literacy skills were more likely to have taken a course than those with lower literacy skills.
• People employed in small enterprises were less likely to have taken a course than those employed by large enterprises. Those in the smallest firm size group (1-19 employees) were 9 percentage points less likely to have taken a course than those in organisations with 100-999 employees, the omitted size group, and 12 percentage points less likely to have taken a course than those working in organisations with 1,000 or more employees.

Likelihood of taking other courses - workers with low literacy skills

For workers with low literacy skills, the marginal effect estimates indicate the following:

• Workers with a Pacific ethnic affiliation were less likely to have taken a course than Europeans by around 10 percentage points.
• Workers in very small enterprises (those with less than 20 employees) were less likely to take courses than those employed by larger enterprises (those with 100 employees or more). Those in the smallest firm size group (1-19 employees) were 12 percentage points less likely to have taken a course than those in organisations with 100-999 employees, the omitted firm size group, and 15 percentage points less likely than those in organisations with 1,000 or more employees.

5.4.3 Summary

The analysis in this section indicates, perhaps not surprisingly, that adults who already hold a post-school qualification are more likely to undertake further education and training, and this appears to be true of workers with level 1 or level 2 literacy skills as well as those with skills at level 3 or above.

There are some significant ethnic group variations in the further education and training participation rates measured in ALL that cannot readily be explained using information on educational attainment, age, occupation and industry. Maori workers were more likely than Europeans to have undertaken some education or training towards a qualification, and Pacific workers were less likely than Europeans to have participated in short courses that were not linked to a qualification. Although the reasons for these ethnic patterns are not clear, the Maori/European differential is consistent with documented patterns in the take-up of industry training programmes. In 2006, for instance, 18 percent of participants in industry training programmes were Maori (Tertiary Education Commission, 2007), higher than the proportion of Maori in the workforce. Statistics on students enrolled with tertiary educational providers for level 1-3 certificates also indicate that Maori were over-represented relative to their share in the workforce in 2006.

Recent immigrants and ESOL speakers were somewhat more likely to have studied for a qualification and somewhat less likely to have taken other courses than native-born residents and native-English speakers, but these differences were not statistically significant and tended to decline in size when other characteristics were taken into account.

Substantial firm size variations in the likelihood of a worker undertaking further education and training were identified, even after controlling for worker characteristics, occupation and industry. This was true of both training that is linked to qualifications and training that does not lead to qualifications. Focusing on workers with relatively low literacy skills, we estimated, for example, that the participation probability of workers who were employed by firms in the largest size group (1,000 plus employees) was 8 percentage points higher than that of workers in the smallest firm size group (1-19 employees). We also estimated that workers with low literacy skills in the smallest firm size group (1-19 employees) were 15 percentage points less likely to take a course that was not linked to a qualification than those in organisations with 1,000 or more employees. The model results suggest that firm size has a larger impact on the probability of undertaking training that is not linked to a qualification than on study or training within the qualifications system.

At the time the ALL survey was carried out, 39 percent of the workforce was employed by enterprises (or non-profit organisations) with 1-19 employees, 18 percent by firms with 20-99 employees, 20 percent by firms with 100-999 employees and 24 percent by firms with 1,000 employees or more. Workers with low literacy skills were distributed across firm size groups in a very similar way.^[9] The enterprise size patterns suggest that larger employers - those with 100 or more employees - tend to invest in workforce learning to a greater extent than small and medium-sized firms, leading to differences in opportunities across workers. It is also possible, however, that workers at smaller and larger firms differ in ways that contribute to the firm-size differences in education and training participation probabilities and are not fully controlled for in the regressions, such as the motivation to learn.

Differences in participation patterns by industry group were found in the analysis that suggest there may be a positive 'public sector ownership' effect on training rates, but it is difficult to be sure of this as the business sector of the job (public or private) was not measured directly in ALL.

5.5 Concluding comments

The ALL results indicate that workers with low literacy and numeracy skills have reasonably high rates of participation in further education and training courses that are linked to qualifications, that is, participation rates similar to those of workers with higher literacy or numeracy skills. In contrast, they are much less likely than workers with higher literacy and numeracy skills to receive or participate in education and training courses that are not linked to qualifications - courses that are much less likely to receive government funding, are frequently funded by employers and are more likely to be delivered in the workplace.

Although courses at tertiary institutions and industry training programmes have the potential to help to maintain or raise the literacy skills of adult learners, an obvious question is how effective they are at doing this in practice, particularly for adults whose existing literacy skills are relatively low. An important dimension of the Government's current strategy for raising the literacy, language and numeracy skills of the workforce is to provide more literacy and numeracy learning opportunities that are 'embedded' within mainstream vocational training courses. Embedded learning opportunities are believed to be more effective in encouraging the participation of less literate or numerate adults and in achieving good learning outcomes. At the time this paper was written, a least two evaluation studies of 'embedded' literacy programmes were underway, and in future, these are likely to provide more information on the effectiveness of those initiatives.

---