III. The Lost Boys of Sudan

In 1983, civil war broke out between the northern Arabic-speaking government of Sudan and southern tribes seeking autonomy. Since then, nearly two million people have died, and five million have been displaced. Among these are at least 20,000 children, mostly boys, between 7 and 17 years of age who were separated from their families. This extraordinary exodus has its origins in traditional forms of migration. After initiation into manhood, young adolescent boys in southern Sudan have generally been quite mobile. Organized into small groups of their peers, they would leave home for a period to look after cattle. Some would head for the towns or cities to go to school or to seek their fortune before eventually returning home. In addition, at times of stress, families all over Africa have sent their children elsewhere to find safety, food, work, and schooling.

After civil war broke out, this process escalated dramatically. Factions began to attack peaceful villages, kidnapping young males to use as cannon fodder in battle zones or to walk through minefields. Fearing they would be targeted as potential combatants, many boys left their villages for refugee camps in Ethiopia. Some traveled with friends or relatives; others slipped away on their own at night. Few had any idea of what lay ahead of them, believing that their journey would last only a few days. Continually under threat, they fled for their lives, losing their way in the wilderness. They lost everything en route to soldiers, swindlers, or bandits. Many fell victim to lethal diseases. Others were so weakened by hunger and lack of sleep that they could go no further and sat down by the roadside, prey for lions and other wild animals.

The survivors who reached refugee camps in Ethiopia began to lead relatively peaceful lives again. But this was not to last. Following the change of government in Ethiopia in May 1991, the Sudanese youths were forced to flee again. This time the journey occurred during heavy rains, and many perished crossing the swollen rivers or were hit by aerial bombardment. Hungry, frightened, and weakened by sleeplessness and disease, they made their way to camps in Sudan, where they received help from the International Committee of the Red Cross. From there, they then traveled on foot to safety in northern Kenya. Since 1992, UNICEF has been able to reunite nearly 1,200 boys with their families. But thousands more have remained in the dusty, fly-ridden refugee camp at Kakuma, where they have had to scrape for food and struggle for education.

In 1999, the United Nations High Commissioner for Refugees, working in collaboration with the Department of State, referred over 3,300 of these youth, called the Lost Boys of Sudan, to the U.S. for resettlement processing. Once the U.S. agreed to admit this group, UNHCR undertook formal identity and background checks. The U.S. then began formally processing the group for resettlement in the fall of 2001. ORR began to track the progress of this group beginning with the 2001 survey. This report continues a five-year longitudinal assessment of the Lost Boys assimilation into U.S. society.

Geographic Location of Lost Boys

The Lost Boys refugees have settled in eighteen States including: Arizona, California, Colorado, Florida, Illinois, Kentucky, Massachusetts, Maine, Michigan, Minnesota, Missouri, North Carolina, New York, Pennsylvania, Texas, Utah, Virginia, Washington. The largest number of Lost Boys initially resettled in Texas (106), followed by Massachusetts (65), California (53), Pennsylvania (49), Washington (45), Utah (37), Arizona (33), New York (20), and Virginia (20).

Economic Adjustment

In 2004, ORR completed its 3rd annual survey of a random sample of Lost Boys selected from a population of at least 3,300 Sudanese children, mostly males between 7 and 17 years of age, who had been separated from their families and subsequently arrived in the U.S. between May 1, 2001 and April 30, 2002. The survey collected basic demographic information such as age, education, English language fluency, job training, labor force participation, work experience, and barriers to employment of each adult member of the household of the selected person. The survey also collected household income, housing, and welfare utilization data.

To evaluate the economic progress of this subset of refugees, ORR used several measures of employment effort frequently used by economists. The first group of measures relates to employment status in the week before the survey and includes the employment‑to‑population ratio (EPR), the labor force participation rate, and the unemployment rate. In addition, data on work experience over the past year and typical number of hours worked per week were analyzed, as well as reasons for not working. Data are also presented on the length of time from arrival in the U.S. to first employment and self-sufficiency.

Employment Status

Table III-1 presents the employment rate (EPR) ¹⁶ in October 2004 for refugees 16 and over within the Lost Boys population. The survey found that the overall EPR for the Lost Boys in the 2004 sample was 92 percent (94 percent for males and 56 percent for females). This rate is much higher than the rate recorded for the non-Lost Boys refugee population (63 percent) (See Table II-2). The male cohort of the Lost Boys population (94 percent) far exceeded the non-Lost Boy refugee cohort (71 percent), and for the first time the female Lost Boy cohort (56 percent) exceeded (by three percent) the overall female refugee cohort, which was 53 percent. As a point of further reference, the employment rate for the non-refugee U.S. population was 62 percent in 2004, 69 percent for males and 56 percent for females. Finally, the overall employment rate of Lost Boys in the 2004 survey was higher than the rate in the survey of the previous year for the same population (86 percent).

Thus, the males in the survey exceeded their counterparts in the U.S. general population by 25 percentage points, while the females in the survey were essentially equal to the U.S. female cohort. This 38 point difference between the male and female Lost Boy cohorts clearly shows that the path to self-sufficiency has been much more difficult for the female members of the Lost Boys population. Notwithstanding the lower employment rate of the female cohort, the overall picture appears quite favorable, with the females making substantial gains in 2004. This picture is particularly notable for a newly arrived population with few family members awaiting them in their designated communities, no work history other than herding cattle, and little in the way of transferable skills. As for the lower employment rate of the female cohort, only a decade ago this employment rate was even lower for the entire refugee population. The 1993 survey, for example, reported that refugees in the five-year population were employed at about half the rate of non-refugee U.S. population (33 percent versus 64 percent).

Table III-1 also contains data on labor force participation rate ¹⁷ for refugees 16 and over in the five-year population. This rate is closely related to the employment rate, except it includes individuals looking for work as well as those currently employed. In December 2004, the overall labor force participation for the Lost Boy population was extremely high (96 percent) (98 percent for males and 67 percent for females). This overall rate is 27 points higher than that of the five-year refugee population (69 percent) and 30 points above the non-refugee U.S. population (66 percent). The male cohort of the Lost Boys (98 percent) also far exceeded both the corresponding male refugee rate (77 percent) and the non-refugee male rate (73 percent). The female rate (67 percent) exceeded for the first time both the corresponding rate of non-Lost Boy female refugees (60 percent) and the U.S. female rate (59 percent). As with the EPR, the labor force participation rate of refugees increases with time in the U.S. The labor force participation rate for the 2004 Lost Boys was 96 percent, as compared to 91 percent for the same group in the 2003 survey. As time passes, this population should yield a greater presence in the workforce if the present pattern is maintained.

The unemployment rate continues this pattern. The overall unemployment rate for the Lost Boys was four percent in this year’s survey. This is a great improvement over the previous survey (12 percent) and compares favorably with the 2004 general refugee employment rate (7 percent) and the U.S. rate (six percent). Perhaps the most noteworthy finding is the dramatic decrease in the female Lost Boy unemployment rate, from 40 percent in the 2003 survey to 11 percent in the 2004 survey. Part of the large swings recorded for this group may be rooted in the extremely small numbers interviewed (only ten females in 2003).

It is clear that the overall pattern is that males are joining the work force at a far greater rate than other refugee males or the U.S. male population as a whole. In addition, the females in this cohort are making progress, having decreased their unemployment rate by 29 points since 2003.

Table III-2 shows that 98 percent of the Lost Boys population has worked at some point since arriving in the U.S. Seventy-seven percent in this year’s survey claimed to have worked at least 50 weeks during the past year. The average number of weeks worked was 48 weeks. Last year, 79 percent reported that they had worked at some point since arrival. Only 54 percent had worked as many as 50 weeks, and the average number of weeks worked was about 46. Noteworthy in Table III-3 is the very high percentage of female Lost Boys that have worked at some point since arrival (67 percent). This seems to suggest that retaining, rather than finding, employment may be their problem in gaining financial independence.

Reasons for Not Looking for Work

The survey also asked Lost Boys age 16 and over who were not employed why they were not looking for employment. Attending school accounted for the largest proportion (almost 50 percent) with an associated median age of 18. Couldn’t find job, Childcare/family responsibility, and poor health each accounted for approximately 17 percent.

Figure 1. Reason not looking for Work for Lost Boys: 2004 Survey.

Note: Limited to Lost Boys who did not work in previous year and are not looking for work at the time of the survey.

Limited English has almost been eliminated as a reason for not looking for work, compared to the 2002 survey when 90 percent of the Lost Boys not in the work force declared limited English proficiency as the reason for unemployment. The Lost Boys seem to recognize the importance of education and English language training in their acclimation to U.S. society, because the largest proportion of Lost Boys (nearly 50 percent) stated attending school as a reason for not looking for employment. It is noteworthy that 34 percent of the Lost Boys did not give a reason for not looking for employment.

Table III-1 – Employment Status of the Lost Boys
	Employment Rate (EPR)			Labor Force Participation Rate			Unemployment Rate
	All	Male	Female	All	Male	Female	All	Male	Female
Lost Boys	92.2%	94.1%	55.6%	96.1%	97.6%	66.7%	3.9%	3.5%	11.1%
U.S. Rates	62.3	69.2	56.0	66.0	73.3	59.2	5.5	5.4	5.6
Note: As of December 2004. Not seasonally adjusted. Data refers to Lost Boys 16 and over. U.S. rates are for 2004.

Table III-2 - Work Experience of Adult Lost Boys
	Number	Percent Distribution
Total Lost Boys 16 years and older	655	100.0
Worked*	641	97.8
50-52 weeks	501	76.5
Full-time	421	65.7
Average weeks worked	48.4
*Lost Boys who worked in the year prior to the survey. **As of December, 2004.

Table III-3 – Employment Status of Selected Lost Boys by Sex: 2004 Refugee Survey
Employment Measure	Percent
Employment Rate (EPR)	92.2%
-Males	94.1
-Females	55.6
Worked at any point since arrival	97.8
-Males	99.4
-Females	66.7
Labor Force Participation Rate	96.1
-Males	97.6
-Females	66.7
Unemployment Rate	3.9
-Males	3.5
-Females	11.1
Note : As of December 2003. Not seasonally adjusted. Data refers to Lost Boys 16 and over.

Figure 2: Elapsed Time to First Job for Lost Boys who have ever worked

Elapsed Time to First Job

How soon do Lost Boys find work after coming to the U.S.? The 2004 survey indicates that of those who have worked at all since coming to the U.S. (about 98 percent of refugees 16 years old and over in that survey), approximately 15 percent found work within one month of arrival, an additional 34 percent after two to three months, nine percent within 4 to 6 months, while another nine percent took 7 to 12 months and only five percent took more than a year (refer to Figure 2). Thus, approximately 60 percent found employment within 6 months of arrival, and 50 percent of refugees reported finding first employment within 3 months of arrival in the current survey.

Factors Affecting Employment

The average number of years of education for all arrivals was approximately ten (refer to Table III-4). It was reported in the 2004 survey that six percent of the Lost Boys had no formal education before coming to the U.S.; this is an increase from the response given the year before when 3 percent of the Lost Boys population reported having no formal education prior to coming to the U.S. It is currently unclear why the percentage doubled. Almost forty percent of the respondents indicated that they had a primary school education, which is very similar to what was reported the previous year (42 percent). Respondents with a high school education made up nearly 20 percent of the Lost Boys population.

The migratory nature and age of the Lost Boys group undoubtedly account for the lower levels of education in this population. Beginning in 1997, many of the Lost Boys were driven from their homes seeking refuge from the ongoing war. Since arriving in the U.S., the Lost Boys have attempted to improve their educational standing. The 2004 survey shows that more than 68 percent of the Lost Boys population are attending some type of school or university, 20 percent are attending high school, 38 percent are in an associate degree program, and 7 percent are in a bachelor’s degree program. Some degree of caution is necessary when interpreting education statistics as these data are self-reported.

Just over two percent of Lost Boys in the 2004 survey reported speaking no English when they arrived in the U.S. (refer to Table III-4). But this was reduced to zero by the time of the survey interview, when all Lost Boys reported speaking English. At the time of arrival, 39 percent of the Lost Boys reported not speaking English well; similarly, 31 percent reported speaking English well or fluently upon arrival in the U.S. The ability of the Lost Boys to quickly gain fluency in English, as evidenced by the statistics, could account for the high EPR for this population. By the time of the interview, only a little more than seven percent of the Lost Boys still reported speaking little English while 83 percent report speaking English well or fluently. The ability to speak English is one of the most important factors influencing the economic self-sufficiency of refugees; the Lost Boys who came here have clearly worked hard at bettering their prospects through intensive language study (refer to Table III-5).

Table III- 4 – Education and English Proficiency Characteristics of Lost Boys
Education and Language Proficiency
Average Years of Education before U.S.	9.7
Highest Degree before U.S.
None	6.0%
Primary School	39.6
Training in Refugee Camp	0.0
Technical School	1.6
Secondary School (or High School)	19.8
University Degree (Other than Medical)	1.6
Medical Degree	0.0
Other	0.0
Attended School/University (since U.S.)	68.7%
Attendance School/University (since U.S.) for degree/certificate	68.7%
High School	19.8%
Associates Degree	37.9
Bachelor’s Degree	6.6
Master’s/Doctorate	0.0
Professional Degree	0.0
Other	0.0
Degree Received	6.6
At Time of Arrival
Percent Speaking no English	2.2%
Percent Not Speaking English Well	39.0
Percent Speaking English Well or Fluently	30.8
At Time of Survey
Percent Speaking no English	0.0
Percent Not Speaking English Well	7.7
Percent Speaking English Well or Fluently	83.0
Note: Data refer to Lost Boys 16 and over. These figures refer to self-reported characteristics of Lost Boys. Professional degree refers to a law degree or medical degree.

Unsurprisingly, the survey found that the Lost Boys who spoke no English upon arrival continued to lag behind the other groups. Their employment rate was 25 percent, far below the overall employment rate of 92 percent. There was little difference, however, between those speaking English well or fluently upon arrival (91 percent) and those who claimed that they could not speak English well (96 percent).

Those who spoke English fluently by the time of the interview reported a very high employment rate (95 percent), but their counterparts who still could not speak English well reported an employment rate 24 points lower (71 percent).

Historically, most refugees improve their English language proficiency over time, and those who do not are the least likely to be employed. During the past 12 months, 28 percent of all Lost Boys attended English Language Training (ELT) outside of high school (refer to Table III-6). For the same period, the proportion of refugees who have attended job-training classes (seven percent) lags far behind those in ELT. About 34 percent of all survey respondents currently attend language instruction, either through high school curriculum or through ELT classes. In the 2003 survey, by contrast, only 19 percent of the Lost Boys were attending some sort of language class at the time of the survey.

Earnings and Utilization of Public Assistance

The earnings of employed refugees generally rise with length of residence in the U.S. (refer to Table III-7). In 2004 the average hourly wage was $9.42. ¹⁸ The corresponding hourly wage in the 2003 survey of employed refugees in the Lost Boys population was $8.56. Table III-7 also details the economic self‑sufficiency of the Lost Boys population in 2004. According to the 2004 survey, almost 93 percent of all refugee households in the U.S. had achieved economic self‑sufficiency—up five points since the 2003 survey when 88 percent were financially independent of public assistance. An additional 8 percent had achieved partial independence, with household income a mix of earnings and public assistance.

Table III – 5 English Proficiency and Associated EPR by Year of Arrival
Percent Speaking No English (EPR)	Percent Not Speaking English Well (EPR)	Percent Speaking English Well or Fluently (EPR)
At the time of arrival
2.2 (25.0)	39.0 (95.8)	30.8 (91.1)
At the time of survey
0.0 (0.0)	7.7 (71.4)	83.0 (94.7)
Note: As of December 2004. Not seasonally adjusted. Data refers to Lost Boys. These figures refer to self-reported characteristics of Lost Boys.

In both 2004 and 2003, no Lost Boys households reported that their household income consisted entirely of public assistance. By way of contrast, the 2002 survey found that nearly 33 percent reported both earnings and public assistance and five percent reported only public assistance. This is a remarkable achievement for a refugee group three years after relocating in the U.S.

The absence of public assistance only cases, coupled with an increase in earnings only cases, confirms that the Lost Boys have adjusted well to the U.S. workforce.

Table III-8 details several household characteristics by type of income. Households that have a mix of earnings and assistance income average approximately four members and three wage earners. Households that receive no public assistance also average four members with three wage earners. However, 25 percent of households with at least one member under age 16 received both public assistance and earnings, whereas only six percent of households with earnings only had at least one family member under 16.

Medical Coverage

Overall, about 19 percent of adult refugees surveyed lacked medical coverage of any kind throughout the year preceding the survey (refer to Table III-9). Also, 60 percent of the Lost Boys received medical coverage through their employer and only seven percent of the sample received medical coverage from Medicaid or RMA. In 2003 the figure was somewhat similar, with five percent receiving medical coverage from Medicaid or RMA. By way of contrast, the 2002 survey indicated that 48 percent of the Lost Boys relied on Medicaid or Refugee Medical Assistance during the year while seven percent had no medical coverage in any of the past 12 months.

Welfare Utilization

Table III-10 presents cash and non-cash welfare utilization data on the Lost Boys sample.

Dependence on cash assistance in 2004 (eight percent) continued to decline from the 2003 figure of 11 percent, albeit at a much slower rate than between 2003 and 2002, when the cash assistance dependence rate was 37 percent. Four percent of Lost Boy households received TANF in the last 12 months. This represents an increase of four percent over 2003, which contrasts with the 16 percent decline from 2002 to 2003.

Table III-6 – Service Utilization by Lost Boys
Type of Service Utilization	Percent
ELT since arrival Inside High School	11.9%
ELT since arrival Outside of High School	28.2
Job training since arrival	6.8
Currently attending ELT Inside High School	11.9
Currently attending ELT Outside of High School	22.0
Note: Data refer to Lost Boys. In order that English language training (ELT) not be confused with English high school instruction, statistics for both populations are given.

Table III-7 – Lost Boys Hourly Wages, Home Ownership, and Self-Sufficiency
Hourly Wages of Employed – Current Job	Own Home or Apartment	Rent Home or Apartment	Public Assistance Only	Both Public Assistance and Earnings	Earnings Only
$9.42	18.9%	75.5%	0.0%	7.5%	92.5%
Note: Data refers to Lost Boys. These figures refer to self-reported characteristics of refugees.

Table III-8 – Characteristics of Lost Boys Households by Type of Income
Lost Boys Households with:
Household Characteristics	Public Assistance Only	Both Public Assistance and Earnings	Earnings Only	Total Sample
Average Household Size	0.0	4.3	3.6	3.7
Average Number of wage earners per household*	0.0	2.8	3.3	3.2
Percent of households with at least one member:
Under the age of 6	0.0%	25.0%	4.1%	5.7%
Under the age of 16	0.0	25.0	6.1	7.5
Fluent English Speaker	0.0	75.0	93.9	92.5
*Data refer to Lost Boys. Lost Boys households with neither earnings nor assistance are excluded.

Table III-9 Source of Medical Coverage for Lost Boys
Source of Medical Coverage	Percent
No Medical Coverage in any of past 12 months	18.6%
Medical Coverage through employer	59.9
Medicaid or RMA	6.8
Note : As of December 2004. Data refer to Lost Boys 16 and over

Approximately 4 percent of sampled households received RCA in 2004, down from 8 percent in the previous survey. Two percent of Lost Boys households received Supplemental Security Income (SSI) in the past twelve months.

General Assistance (also called General Relief or Home Relief in some States) is a form of cash assistance funded entirely with State or local funds. It generally provides limited assistance to single persons, childless couples, and families with children that are not eligible for TANF. The 2003 survey reported that about seven percent of Lost Boys households received some form of GA during the previous twelve months; in the current survey, GA utilization dropped to zero.

Table III-10 Public Assistance Utilization of Lost Boys
Type of Public Assistance	Percent
Cash Assistance
Any Type of Cash Assistance	7.5%
TANF	3.8
RCA	3.8
SSI	1.9
General Assistance	0.0
Non-Cash Assistance
Medicaid or RMA	6.7
Food Stamps	11.3
Housing	18.9

Note : Medicaid and RMA data refer to adult Lost Boys age 16 and over. All other data refer to Lost Boys households and not individuals. Many households receive more than one type of assistance.\

Non-cash welfare utilization increased by nearly six percentage points between 2003 and 2004. About 11 percent of Lost Boys households reported receiving food stamps in 2004. This was roughly equivalent to the 2003 figure; however, non-cash assistance for housing increased three percentage points to 19 percent, and Medicaid/RMA increased by two percentage points to seven percent (Refer to Table III-10).

Improvement in self-sufficiency was also recorded for medical coverage in Table III-9. The 2003 survey reported that 46 percent of Lost Boys received their medical coverage through their employer. The 2004 survey shows a marked improvement—with 60 percent of Lost Boys now covered by their employer. Although, Lost Boys showed an overall improvement in medical coverage they showed a slight increase in their housing utilization increasing from 16 percent in 2003 to 19 percent in 2004.

Employment and Welfare Utilization Rates by State

The 2004 survey also reported welfare utilization and employment rate by State of residence. Table III-11 shows the EPR and utilization rates for various types of welfare for twelve States, as well as the nation as a whole. Unlike Table III-10, which computes welfare utilization rates for entire households, Table III-11 presents data on utilization by individual refugees (including children).

In the five-year population of non-Lost Boys refugee households, the EPR was generally high where welfare utilization was low and vice versa. In the Lost Boys sample, however, there was not always a distinct correlation between EPR and welfare dependency. The welfare utilization rates in States with the highest refugee employment rates— Texas (100 percent), Massachusetts (100 percent), Illinois (100 percent), Connecticut (100 percent), Nebraska (100 percent), and Utah (100 percent)—were zero for all but Massachusetts, which had a utilization rate of 33 percent for RCA.

Some States with lower employment rates reported no welfare utilization. For example, California (75 percent) and North Carolina (85 percent) had lower employment rates but also had welfare utilization rates of zero.

The findings from ORR's 2004 survey indicate (as in the previous year) that the Lost Boys faced significant problems upon arrival in the United States, especially the female members of this group. They have, however, made significant strides in achieving independence. Although many spoke at least some English upon arrival, they have enrolled in ELT in large numbers, and their gains in English language fluency have been substantial. A great majority (83 percent) report speaking English well or fluently now, with virtually all able to speak at least some English. Their cash assistance utilization, never high in their first year, has continued to drop to where it is approximately eight percent now. They have eagerly sought employment, and the few that are currently not working are actively looking for employment.

Table III-11 – Lost Boys Employment-to-Population Ratio (EPR) and Welfare Dependency for Top Twelve States
Percent of Individuals (vs. Households) on Welfare
State	Arrivals*	EPR	TANF	RCA	SSI	GA	Total**
Texas	(95)	100%	0.0%	0.0%	0.0%	0.0%	0.0%
North Carolina	(84)	85.0	0.0	0.0	0.0	0.0	0.0
Georgia	(73)	90.0	0.0	0.0	0.0	0.0	0.0
New York	(62)	93.8	33.3	33.3	0.0	0.0	33.3
Massachusetts	(59)	100.0	0.0	33.3	0.0	0.0	33.3
Michigan	(51)	92.9	0.0	0.0	0.0	0.0	0.0
Utah	(48)	100.0	0.0	0.0	0.0	0.0	0.0
Illinois	(37)	100.0	0.0	0.0	0.0	0.0	0.0
California	(33)	75.0	0.0	0.0	0.0	0.0	0.0
Tennessee	(29)	60.0	50.0	0.0	0.0	0.0	50.0
Connecticut	(26)	100.0	0.0	0.0	0.0	0.0	0.0
Nebraska	(22)	100.0	0.0	0.0	0.0	0.0	0.0
Other States	(91)	87.0	0.0	0.0	9.1	0.0	9.1
All States	(710)	92.2	3.8	3.8	1.9	0.0	7.5
*The State arrival figures are weighted totals. **The column totals represent percent of individuals who received any combination of AFDC, RCA, SSI and/or GA, e.g., if an individual received AFDC, RCA, SSI, and GA, he/she is counted four times. Note: As of December 2003. Not seasonally adjusted. Welfare utilization refers to receipt of public assistance in at least one of the past twelve months. The listed utilization rate for each type of public assistance is the ratio of the number of individuals (including minor children) receiving such aid to the total number of individuals in the sample population residing in that State. Because some refugees have difficulty distinguishing between GA and AFDC/TANF, some GA utilization may reflect AFDC/TANF utilization. For data on welfare utilization by household, see Table III-10.

Technical Note: The Lost Boys of Sudan Survey, with interviews conducted by DB Consulting Group, Inc. in the fall of 2002, 2003, and 2004, is a subset of the Annual Survey of Refugees conducted by ORR since 1975. Although respondents from Sudan have traditionally been included into the Annual Survey of Refugees, this is the first time that a single population has been surveyed to track their adjustment to resettlement in the U.S.

For the 2004 survey, 65 of the 305 Lost Boys in the sample were contacted and interviewed. Of the remaining 240 cases, one moved abroad, one went into the army, one refused, and 238 could not be traced in time to be interviewed.

---