Profiles and Transitions of Groups at Risk of Social Exclusion: Lone Parents

5. Labour Market Activity and Low Income

5.1 Introduction

The relatively high incidence of low income among lone mothers is not very surprising, given that typically there are no other income earners in the family and the fact that the presence of young children is both a cost and a potential barrier to full-time/full-year employment.

This section focuses on the following question: why are some lone mothers in low income while others are not? Is it mostly because they do not work enough hours, or is it mostly because their hourly earnings are low? And what are the main factors that may explain low hours (e.g. presence of pre-school age children) or low hourly earnings (e.g. low education)?

To answer the above questions, we first compared the hours of work and hourly earnings of low income and non-low income lone mothers. Then, we used multivariate analysis to understand the reasons for the low hours of work or low earnings.

The analysis focuses on lone mothers only. The reason is that lone mothers have more than twice the incidence of low income of lone fathers. As a result, the significant gender difference between low income and non-low income lone parents will tend to distort the influence of other factors (such as hourly earnings) that tend to be closely correlated with gender.

5.2 Labour Market Comparisons between Low Income and Non-Low income Lone Mothers

Low income among lone mothers is mostly the result of low hours of work. For example, if all lone mothers worked at least 1,750 hours annually (the average for non-low income working mothers), their low income rate would have dropped from 39% to 8%.

This sub-section compares key labour market factors between low income and non-low income lone mothers. Table 5.1 shows that the most significant difference between low income and non-low income lone mothers is their employment rate. Thus while 89% of non-low income lone mothers worked for pay in 1998, the corresponding rate among low income lone mothers was 37%.

Table 5.1 Work and earnings of lone mothers, 1998
		Low income	Non-low income
1.	All lone mothers	246,407	384,324
2.	Had some work in 1998	90,944	340,148
2.	Employment rate (2/1)	36.9%	88.5%
3.	Average weeks employed (per employed)	33.8	49.5
4.	Had some unemployment in 1998	36,080	44,316
	Unemployment rate (4/2)	39.7%	13.0%
	Average weeks unemployed (per unemployed)	19.8	16.0
5.	Average weekly hours of work (per employed)	29.7	35.3
6.	Average hourly earnings (per employed)	$8.13	$16.59
7.	Average annual earnings (per employed)	$5,874	$27,360
8.	Average annual earnings (all lone mothers)	$3,012	$25,656

Low income lone mothers worked, on average, fewer weeks and fewer weekly hours, than non-low income lone mothers. They also were more likely to experience unemployment. At the same time, their hourly earnings, when they worked, were less than half the hourly rates for non-low income lone mothers.

As a result of all these differences, the average earnings for all low income lone mothers (working and non-working) in 1998 were $3,012, compared to $25,656 for all non-low income lone mothers - a gap of $22,644.

(a) Impact of Hours and Wages on the Earnings Gap

Table 5.2 shows the impact on the earnings gap between low income and non-low income lone mothers if low income lone mothers had similar hours of work and similar hourly earnings to those of non-low income lone mothers. It shows that:

impact of increasing work effort: if low income lone mothers had the same employment rate and the same annual hours of work as non-low income lone mothers, the earnings gap between the two groups of lone mothers would have dropped by 43.2%;
impact of increasing hourly earnings: on the other hand, if only the hourly earnings of low income working lone mothers were raised to match the average hourly earnings of non-low income working lone mothers, the earnings gap between the two groups would have dropped by 14.9%.

The point of the above simple exercise is to demonstrate that low hours of work is the main factor behind the earnings gap between low income and non-low income lone mothers. However, it must be recognized that there is an interaction between hours of work and hourly earnings. In other words, although the direct cause of low income is typically low hours of work or no work at all, a not uncommon underlying cause is low potential hourly earnings.

Table 5.2 Impact of hours of work and wage rates on earnings gap between low income and non-low income lone mothers, 1998
	Earnings Gap	Percent drop
Earnings gap between low income and non-low income lone mothers before any adjustments.	$22,644
Earnings gap if low income and non-low income lone mothers had the same employment rate.	$18,201	19.6%
Earnings gap if low income and non-low income lone mothers had the same employment rate and average hours of work.	$12,851	43.2%
Earnings gap if low income and non-low income lone mothers had the same hourly earnings, with no change in the hours of work.	$19,277	14.9%

(b) Impact of Hours and Wages on the Incidence of Low Income

Table 5.3 presents estimates of the impact on the incidence of low income among lone mothers of raising their work effort or hourly earnings. The results show that:

impact of increasing the employment rate: if all non-working low income lone mothers were working for pay (at the same average hours and same wage rates as the average working low income lone mother), the incidence of low income in 1998 would have dropped from 39.1% to 23.0% (a 16 percentage point reduction);
impact of increasing the hours of work: in addition, if no working low income lone mother worked less hours than the average non-low income working lone mother, the incidence of low income in 1998 would have dropped to 8.1% (an additional 15 percentage point reduction);
impact of increasing hourly earnings: on the other hand, if there was no change in work effort among low income lone mothers but only their hourly earnings were raised so that no one earned less than the average rate of non-low income lone mothers, the incidence of low income would have dropped from 39.1% to 29.5% (one-third of the potential drop of increasing work effort).

Table 5.3 Impact of hours of work and wage rates on the incidence of low income among lone mothers, 1998
	Estimated number	Incidence	Drop in incidence
All lone mothers with children under 18.	630,731
Number of low income before any adjustments.	246,407	39.1%
Number of low income if all non-working low income lone mothers worked as many hours as the average working low income lone mother.	144,945	23.0%	16.1%
Number of low income if all poor lone mothers worked at least as much as the average working non- low income lone mother.	50,842	8.1%	31.0%
Number of low income if the only change was to raise the hourly earnings of working low income lone mothers to be at least as high as the average hourly earnings of non-low income lone mothers.	186,378	29.5%	9.5%

The above calculations imply that:

if all low income lone mothers had worked at least 1,000 hours in 1998 (at current wage rates for low income lone mothers), 40% of the low income problem would have disappeared; and
if they had worked 1,750 hours (i.e. the average hours of non-low income working mothers) most low income (80%) would have disappeared.

The above results should not be interpreted as implying that skills upgrading and earnings supplementation are not also important. After all, both of them can have a positive effect on work effort (by, respectively, improving marketable skills and providing work incentives). Rather, the main message from these results is that the main focus of policy should be to encourage a stronger labour force attachment among lone mothers. This could be achieved by, for example, providing more employment services (e.g. job referrals) and making paid work more attractive (e.g. though earnings supplementation or a more generous treatment of earnings under provincial social assistance programs).

5.3 Determinants of Probability of Working for Pay

Among all lone mothers (low income and non-low income), the three strongest factors associated with a low probability of working for pay, in descending order of significance, were: the presence of at least one high risk characteristic (recent immigrant; disabled; aboriginal); being a high school drop-out; and living in the Atlantic region.

This sub-section takes a closer look at the factors that affect the labour market behaviour of lone mothers (both low income and non-low income). Given that over one-half of low income lone mothers are not working for pay, the main focus of the discussion here is on the factors that affect the work participation of lone mothers.

Table 5.4 compares the incidence of work, annual hours of work and hourly earnings of lone mothers by various characteristics, while Table 5.5 presents the results of a further probing using logit regression analysis of the probability of working among lone mothers. The logit regression results in Table 5.5 show that, in descending order of importance, the three main influences on the probability of work are:

recent immigrant, aboriginal or disabled: membership in at least one of these three high risk groups had the most negative influence on the probability of working (26%); this result suggests that lone mothers with any of these additional characteristics should be a high priority for public policy;
education: high school drop-outs have a 12% lower probability of working than those with a high school diploma; the difference is even greater relative to those with a post-secondary certificate or degree (21%); and
region: living in the Atlantic region reduces the probability of working relative to Ontario by 12% and relative to the Prairies by 24%; this result likely reflects inter-provincial differences in labour market conditions; however, it may also reflect differences in the design of social assistance programs among provinces (issues related to social assistance are probed in a later section).

As seen in Table 5.5, one surprising result was that the presence of pre-school age children did not have a negative effect on the probability of working. However, further probing showed that among high school drop-outs the probability of working is low regardless of the age of the youngest child; on the other hand, among higher education levels, there was a negative relation between the presence of young kids and the probability of working. Unfortunately, the sample is too small for more definitive conclusions.

Another surprising result was that the presence of earnings from other members of the family, investment income, or alimony income did not have a negative effect on the probability of working. One would have expected a negative relation, because the presence of other income sources tends to lessen the need for work. In fact, such a negative relationship has been confirmed in the literature with respect to married women. The most likely explanation why this negative correlation was not confirmed here is that, in most cases, other sources of market income among lone mothers were not significant enough to influence their work behaviour.

Table 5.4 Cross-sectional work profile of all lone mothers, 1998¹
	Percent who worked	Average hours of work	Average hourly earnings
*Age in 1998*
16-29	47.8%	1,242	$10.86
30-55	73.2%	1,657	$15.42
*Age when first child was born*
Under 20	50.0%	1,473	$13.84
20 or more	72.6%	1,613	$15.09
*Marital status when first child was born*
Not in a union	59.0%	1,443	$13.66
Married	76.6%	1,707	$15.96
Common law	67.9%	1,660	$13.33
*Age of youngest child*
0-5	59.6%	1,411	$13.55
6-11	69.7%	1,580	$13.75
12-17	75.9%	1,780	$16.94
*Student during the year*
Yes	56.2%	1,242	$13.96
No	71.0%	1,661	$14.96
*Level of education of non-students*
Less than high school	50.5%	1,565	$10.64
High school diploma	73.1%	1,735	$12.98
Some post-secondary	66.6%	1,389	$16.55
Post-secondary degree	79.7%	1,731	$16.29
*Immigrant, aboriginal, or disability²*
Yes	48.6%	1,284	$12.77
No	74.2%	1,663	$15.20
*EI region employment rate*
At, below average	61.6%	1,538	$13.64
Above average	73.0%	1,639	$15.49
*Broad region*
Atlantic	56.5%	1,434	$12.62
Quebec	67.2%	1,606	$14.56
Ontario	67.9%	1,636	$16.71
Prairie	76.9%	1,661	$12.38
B.C.	69.7%	1,505	$14.25
*Other earnings or market income in 1998*
Yes	78.7%	1,696	$15.90
No	61.3%	1,519	$13.85
*All*	68.3%	1,602	$14.80
(1) Sample of lone mothers, age 16-55, with at least one child under 18 in 1998. (2) Immigrated in last 10 years; or aboriginal origin; or work limiting disability.