HUMAN CAPITAL:

EDUCATION
AND EARNINGS

1 Hewi-Lin Chuang, Ph.D.

2010/04/08
台灣 1000 大企業最愛大學生調
查
文／吳永佳 2010 年 3 月 Cheers 雜誌

長期作為企業與人才之間的橋梁 ，
《 Cheers 》雜誌 14 年來執行「 1000 大企業
人才策略與最愛大學生」調查 ﹕
 調查對象﹕ 2009 年《天下》雜誌 1000 大企業人資主管
 有效樣本﹕寄出 1,192 份問卷，回收 628 份，回收率 53 ％
 調查期間﹕ 2010 年 1 月 14 日 ~2010 年 2 月 6 日
 調查執行：《天下》雜誌調查中心

2
表 1﹕1000 大企業最愛大學生總排
名
排名 學校 2009
2009 年
年 2008
2008 年
年 排名 學校 2009
2009 年
年 2008
2008 年
年
排名 排名 排名 排名

1 台灣大學 1 1 11
11 中央大學 11
11 13
13

2 成功大學 2 2 12
12 中原大學 12
12 11
11

3 交通大學 3 3 13
13 輔仁大學 13
13 14
14

4 清華大學 4 4 14
14 東吳大學 14
14 16
16

5 政治大學 5 5 15
15 台北大學 15
15 19
19

6 台灣科大 7 8 16
16 中興大學 18
18 18
18

7 中山大學 6 9 17
17 元智大學 17
17 12
12

8 淡江大學 8 7 18
18 東海大學 19
19 17
17

9 台北科大 9 6 19
19 中正大學 16
16 15
15
3
10
10 逢甲大學 10
10 10
10 20
20 文化大學 - -
表 2﹕1000 大企業最愛私校與技職排名

私校排名 整體排名 學校名稱 技職排名 總排名 學校

1 8 淡江大學 1 6 台灣科大

2 10 逢甲大學 2 9 台北科大

3 12 中原大學 3 21 高雄應用

4 13 輔仁大學 4 23 雲林科大

5 14 東吳大學 5 24 高雄第一

6 17 元智大學 6 28 南台科大

7 18 東海大學 7 30 高雄餐旅

8 20 文化大學 8 31 屏東科大

9 22 銘傳大學 9 34 朝陽科大
4
10 26 世新大學 10 36 文藻外語
看人才，態度面最重要，最擔心的是穩定度

企業最重視新鮮人的能力﹕
 「學習意願與可塑性」（ 73.5 ％）
 「穩定度與抗壓性」（ 65.7 ％）
 「專業知識與技術」居第 3 （ 56.5 ％ )
企業認為新鮮人最待加強的能力﹕
 「穩定度與抗壓性」（ 67.0 ％）
 「解決問題的能力」 （ 51.3 ％）
 「具有國際觀與外語能力」（ 38.9% ）

資料來源﹕
http://www.cheers.com.tw/doc/page.jspx?id=40288abc276148df0127ae79cce833e8&nu 5
mber=1
新世代最嚮往 100 大企業
文／史書華 2010 年 4 月 Cheers 雜誌

《 Cheers 》雜誌第五年「新世代最嚮往 100 大企

業」大調查﹕
 調查對象﹕全台灣 160 個大學科系應屆畢業生 ( 包含大學 77.98 ％
、研究所 22.02 ％ )
 有效樣本﹕寄出 4200 份問卷，回收 2,834 份 ，回收率 67.48%
 調查期間﹕ 2010 年 1 月 5 日 ~2010 年 3 月 5 日
 調查方法﹕根據教育部公佈 98 學年度各大專院校的人數與科系分
佈狀況，採分層比例抽樣法進行郵寄問卷調查
 調查執行：《天下》雜誌調查中心
資料來源﹕ 6
http://www.cheers.com.tw/doc/page.jspx?id=40288abc276148df0127ae6fdb1a33de&number=1
科技業退位，服務業竄前
 歷經一場全球經濟風暴後，今年調查共有 3 大改
變、 3 大發現﹕
 【改變 1 】統一企業擠下去年榜首中華電信和科技業霸主鴻海
 1. 統一企業， 2. 誠品， 3. 統一星巴克 ( 參見表 1)
 【改變 2 】「工作穩定」仍是選擇企業首要標準，但「搶抱鐵
飯碗」開始退燒
 畢業後打算至公營企業比例﹕ 2006 年 (36.5%) ， 2007 年
(52.0%) ， 2008 年 (66.8%) ， 2009 年 (81.4%) ， 2010 年 (70.7%)
 【改變 3 】數位新媒體超越傳統徵才管道，成為新鮮人認識企
業的重要來源
 數位新媒體﹕如 YAHOO! 奇摩知識 + 、 PTT (4)
 傳統徵才管道﹕企業徵才活動 (9) 就業輔導單位 (10) 徵才網站 (11) 7
( 參見表 2)
表 1﹕ 新世代最嚮往企業 Top 10 表 2﹕ 大學生認識企業的重要依據排
名
2010 年排名 2009 年排名 企業名稱 排名 選項

1 2 統一企業 1 企業知名度

2 企業形象／產品廣告
2 3 誠品
3 7 統一星巴克 3 產業前景

4 5 鴻海精密 4 社群網路（如 YAHOO! 奇摩知識 + 、

PTT ）
5 1 中華電信
5 師長、親友評價
6 13 台灣積體電
6 企業門市與產品
路
7 4 中國鋼鐵 7 企業的營運數字

8 6 華碩電腦 8 看企業 CEO 是誰

9 企業徵才活動
9 21 王品台塑牛
排館 10 就業輔導單位
10 10 奇美電子 11 徵才網站
8
註：群創光電 3 月 18 日已與奇美電子、統寶光電合併為「奇美電
子」
科技業退位，服務業竄前
 【發現 1 】新鮮人什麼都懂，就是不懂企業
 超過 4 成的受訪者表示「不認識」自己筆下選填的企業
 【發現 2 】：大學生自認「身價不只 22,000 」
 新鮮人自評身價﹕研究所應屆畢業生 (33,734) ，大學應屆畢業
生 (27,701)
 【發現 3
】：新世代最愛企業家：王永慶蟬連兩年 ，林
百里重新進榜
 1. 王永慶 2. 郭台銘 3 比爾‧蓋茲（ Bill Gates ）

資料來源：
http://www.cheers.com.tw/doc/page.jspx?id=40288abc276148df0127ae6fdb1
a33de&number=1
9
HUMAN CAPITAL: EDUCATION
AND EARNINGS
 Wages will vary among workers because workers
are different. We each bring into the labor market
a unique set of abilities and acquired skills, or
human capital.
 We begin our study of human capital by focusing
on the decision to acquire formal education. The
skills we acquire in school make up an
increasingly important component of our stock of
knowledge.
10
INTRODUCTION

 People bring into the labor market a unique set of

abilities and acquired skills known as human
capital.
 Workersadd to their stock of human capital
throughout their lives, especially via job experience
and education.

11
EDUCATION: STYLIZED FACTS

 Education is strongly correlated with:

 Labor force participation rates
 Unemployment rates
 Earnings

12
1. EDUCATION IN THE LABOR
MARKET: SOME STYLIZED FACTS
表一 勞動參與率與教育程度 單位： %

教育程度 九十九年
七十年 七十五年 八十年 八十五年 九十年 九十五年 九十七年 九十八年 ( 一、二
月)
國中及以下 56.69 58.89 56.55 53.00 47.71 44.34 42.87 41.67 41.45
高中（職） 54.76 57.87 60.22 60.53 61.34 63.52 63.64 62.61 62.23
大專及以上 63.13 66.19 67.27 67.97 66.82 67.38 68.18 68.40 68.78

表二 就業者與教育程度 單位： %
九十九年
教育程度 七十年 七十五年 八十年 八十五年 九十年 九十五年 九十七年 九十八年 ( 一、二
月)

國中及以下 69.31 64.05 54.26 44.34 35.64 27.40 24.75 23.27% 22.80%
高中（職） 20.08 23.85 29.89 33.69 35.97 35.91 35.38 34.55% 34.18%
大專及以上 10.61 12.11 15.85 21.96 28.39 36.70 39.87 42.18% 43.02%
13
表三 失業率與教育程度 單位： %
九十九年
教育程度 七十年 七十五年 八十年 八十五年 九十年 九十五年 九十七年 九十八年 ( 一、二
月)

國中及以下 0.70 1.63 0.98 1.83 4.59 3.21 3.76 5.84 5.36

高中（職） 1.95 3.89 2.05 3.00 4.65 4.36 4.34 6.19 6.15

大專及以上 1.23 2.93 1.78 2.41 3.20 3.98 4.21 5.57 5.57

表四 薪資與教育程度 單位：元 ( 平減 95 年 CPI)

教育程度 八十五年 九十年 九十一年 九十二年 九十三年 九十四年 九十五年 九十六年 九十七年 九十八年

國中及以下 29,035 28,119 27,866 27,244 27,423 27,410 27,294 27,235 26,739 24,597

高中（職） 31,390 31,217 31,059 30,858 30,523 30,363 30,123 29,773 28,709 27,820

大專及以上 43,925 44,396 43,882 43,905 42,942 41,851 41,409 40,510 39,241 38,223
14
2. THE SCHOOLING MODEL
 What factors motivates some workers to remain
in school while other workers dropout before
they finish high school? We assume that
workers acquire the skill level that maximizes
the present value of lifetime earnings.

15
2. THE SCHOOLING MODEL
 Real earnings (earnings adjusted for inflation).
 Age-earnings profile: the wage profile over a worker’s
lifespan.
 The higher the discount rate, the less likely someone
will invest in education (since they are less future
oriented).
 The discount rate depends on:
 the market rate of interest.
 time preferences: how a person feels about giving up
today’s consumption in return for future rewards. 16
(1) PRESENT VALUE CALCULATIONS

 Presentvalue allows comparison of dollar amounts

spent and received in different time periods. (An
idea from finance.)

 Present Value = PV = y/(1+r)t

r is the per-period discount rate.
y is the future value.
t is the number of time periods.

17
 The present value of the earnings stream if the worker
only gets a high school education is:

wHS wHS wHS 46

wHS
PVHS  wHS    ...  
1  r  1  r 2
1  r  t 0 1  r t
46
(1
)
The parameter r is the worker’s rate of discount. There
are 47 terms in this sum, one for each year that elapses
between the ages of 18 and 64.
18
The present value of the earnings stream if the worker gets
a college diploma is:
H H H wCOL wCOL wCOL
PVCOL  H       ... 
1  r  1  r  1  r  1  r  1  r 
2 3 4 5
1  r 46

Direct Costs of Attending College Post-College Earnings Stream

3
H 46
wCOL
   (2)
t 0 1  r t
t 4 1  r t

The first four terms in this sum give the present value of
the direct costs of a college education, while the remaining
43 terms give the present value of lifetime earnings in the
19
post-college period.
We assume that a person’s schooling decision maximizes the
present value of lifetime earnings. Therefore, the worker attends
college if the present value of lifetime earnings when he gets a
college education exceeds the present value of lifetime earnings
when he gets only a high school diploma, or:

Dollars PVCOL  PV HS (3)

wCOL Goes to College

wHS Quits after High School

Age
18 22 65
-H

20
FIGURE 1 Earning Streams Faced by a High School Graduate
(2) THE WAGE-SCHOOLING LOCUS
 Thesalaries firms are willing to pay workers
depends on the level of schooling.
 Properties of the wage-schooling locus.
 The wage-schooling locus is upward sloping.
 The slope of the wage-schooling locus indicates the
increase in earnings associated with an additional year of
education.
 The wage-schooling locus is concave, reflecting
diminishing returns to schooling.
21
(2) THE WAGE-SCHOOLING
LOCUS
The wage-schooling Dollars

locus gives the salary

that a particular worker 30,000
would earn if he
completed a particular 25,000
23,000
level of schooling. If the
20,000
worker graduates from
high school, he earns
$20,000 annually. If he
goes to college for 1
year, he earns $23,000.
And so on.
22
0 Years of
12 13 14 18
Schooling
 Defn. The Marginal Rate of Return to Schooling

The percentage change in earnings resulting from

one more year of school is defined to be the
marginal rate of return to schooling.

23
(3) The Stopping Rule, or When Should I
Quit School?
Suppose that the worker has a rate of discount r that
is constant; that is, it is independent of how much
schooling the worker gets. The stopping rule that
maximizes the worker’s present value of earnings
over the life cycle is given by:

Quit school when the marginal rate of return to

schooling = r

24
This stopping rule maximizes the worker’s present
value of earnings over the life cycle.
Rate of interest

r’

MRR

s’ s* Years of Schooling
25

FIGURE 3 The Schooling Decision

SCHOOLING AND EARNINGS WHEN
WORKERS HAVE DIFFERENT RATES OF
DISCOUNT
Rate of Dollars
Interest

wHS
PBO

rAL
wDROP PAL
rBO

MRR

11 12 Years of 11 12 Years of
Schooling Schooling

26
 SCHOOLING AND EARNINGS WHEN
WORKERS HAVE DIFFERENT ABILITIES
Rate of Dollars
Interest Z
wHS Bob

wACE Ace

wDROP
r PACE

MRRBOB
MRRACE
11 12 Years of 11 12 Years of
Schooling Schooling

Ace and Bob have the same discount rate (r) but each worker faces a different wage-
schooling locus. Ace drops out of high school and Bob gets a high school diploma.
The wage differential between Bob and Ace (wHS - wDROP) arises both because Bob
goes to school for one more year and because Bob is more able. As a result, this wage 27
differential does not tells us by how much Ace’s earnings would increase if he were
to complete high school (wACE - wDROP).


3. IS EDUCATION A GOOD
INVESTMENT?
(1) Is Education a Good Investment for Individuals?
The rate of return typically estimated for the U.S.
generally fall in the range of 5-15 percent.

 At first glance, an investment in education is

about as good as an investment in stocks, bonds,
or real estate.

28
EDUCATION AND THE WAGE GAP
 Observed data on earnings and schooling does not
allow us to estimate returns to schooling.

 In
theory, a more able person gets more from an
additional year of education.

 Abilitybias: The extent to which unobserved ability

differences exist affects estimates on returns to
schooling, since the ability difference may be the true
source of the wage differential.
29
ESTIMATING THE RATE OF
RETURN TO SCHOOLING
 A typical empirical study estimates a regression of
the form:
Log(w) = a·s + other variables
w is the wage rate
s is the years of schooling
a is the coefficient that estimates the rate of
return to an additional year of schooling

30
 31

資料來源：中央產經論文 - 台灣地區大學教育報酬率時間變化之分析 ( 邱麗
 32

資料來源：中央產經論文 - 台灣地區大學教育報酬率時間變化之分析 ( 邱麗
A. The Upward Bias

The typical estimates of the rate of return on further

schooling overstate the gain an individual student
could obtain by investing in education because they
are unable to separate the contribution that ability
makes to higher earnings from the contribution
made by schooling.

i.e., Some of the added earnings college graduates

typically receive would probably be received by an
equally able high school graduates who did not
attend college. 33
 Methods to correct this bias:
a. Separating effects of ability and schooling by
including the aptitude-test scores such as IQ.
b. Controlling for all the unmeasured aspects of ability
by using data on twins.
 Part of earnings differentials associated with higher
levels of schooling are due to inherently abler
persons obtaining more schooling.

34
B. The Downward Bias
a. Some benefits of college attendance are not
necessarily reflected in higher productivity.
b. Most rate-of-return studies fail to include fringe
benefits.
 Fringe benefits, usually as a fraction of total
compensation, tend to rise as money earnings rise.
a. Some of the job-related rewards of college are
captured in the form of psychic or nonmonetary
benefits.
35
C. Selection Bias
The measured rate of return on a college
education may understate the actual return for
those who choose to attend college. Likewise,
the measured rate of return may overstate the
return that would have been received by those
terminating schooling with higher school had
they instead chosen to attend college.

36
Measured benefit Bt: Bt  E cc,t  E hh,t

E cc,t : the earnings in a college-level job of those who

choose to go to college.
E hh,t : the earning in a high school-level job of those who
choose not to go to college.
Let E hc,t : the earnings in a high school-level job of those
who choose to attend college terminating
schooling at high school.
E hc,t is perhaps less than E hh,t
37
  Btc  E cc,t  E hc,t  Bt  measured benefit

E ch,t :the earnings in a college-level job of those

who choose not to attend college were to
alter their decisions.
E ch,t is perhaps less than E cc,t

 Bth  E ch,t  E hh,t  Bt  measured benefit

→When abilities are diverse, the principle of
comparative
advantage is an important factor in making choices
about schooling and occupations. 38
 SCHOOL QUALITY AND THE RATE
OF RETURN TO SCHOOLING
8 8
Rate of return to schooling

7 7

Rate of return to schooling

6 6
5 5
4 4

3 3

2 2
15 20 25 30 35 40 0.5 0.75 1 1.25 1.5 1.75 2
Pupil/teacher ratio Relative teacher wage

Source: David Card and Alan B. Krueger, “Does School Quality Matter?
Returns to Education and the Characteristics of Public Schools in the United
States,” Journal of Political Economy 100 (February 1992), Tables 1 and 2.
The data in the graphs refer to the rate of return to school and the school 39
quality variables for the cohort of persons born in 1920-1929.
台大「惠我良多」？
—論各大學畢業生初出校門的表現
 運用「七十二年與七十三年專上畢業生調查」與「七
十三年與七十四年專上畢業生調查」資料，分析台灣
各大學院校初出校門的畢業生，在薪資決定、就業與
進修決擇上是否有顯著差異。
 校際間的差異主要顯現在「就業比例」、「進修比例
」上，然而薪資水準並無顯著差異。
 未控制「聯考最低錄取分數」這項能力指標下，台大
畢業生在勞動市場的表現相當優異，薪資水準高於多
數大學院校。
 然而在控制「聯考最低錄取分數」下，台大畢業生在
勞動市場的表現並不甚佳，就業比例低於多數學校， 40
薪資水準亦未必優於他校。
台大「惠我良多」？
—論各大學畢業生初出校門的表現

41
台大「惠我良多」？
—論各大學畢業生初出校門的表現
 若嘗試由「教育生產函數」理論提出解釋，由於台大畢
業生多屬聯考中的佼佼者，在未控制「能力」之下，可
能使得台大的貢獻被高估。
 值得注意的是，此研究所使用的資料為「專上畢業生調
查」是自學校畢業或役畢未滿一年的樣本。依 Farber
and Gibbons (1991) 的說法，初入勞動市場者的「薪資
」與「生產力」間無必然關係，因而各大學院校對初入
勞動市場者薪資的影響，未必能反映對生產力的影響。
 資料來源：于若蓉 , 朱敬一 , “ 台大「惠我良多」 ? －
論各大學畢業生初出校門的表現” , 《經濟論文叢
刊》 , 1998, 26: 65-89. 42
98 年大專應屆畢業青年平均薪資

43
資料來源：行政院青年輔導委員會 -98 年大專青年就業力現況調查報告
98 年大專應屆畢業青年平均薪資

44

資料來源：行政院青年輔導委員會 -98 年大專青年就業力現況調查報告

(2) Is Education a Good Social Investment?
Some critics have suggested that to a large degree,
education acts merely as a “sorting device”.
→Schooling does nothing to alter productive
characteristics. The education is simply as a filter
that has the effect of “signaling” which people are
likely to be most productive.
Note: Even if schooling were only a screening
device, it could have social value. Employers need a
reliable method by which to select employees.
Investment in schooling sends a signal to the labor
market that one has a certain level of ability. 45
→School would have net social value if the decision to
attend and the success one attained in school sent
accurate signals about productive characteristics to
employers in the least costly way.

Either education does enhance worker productivity or

it is a cheaper screening tool than any other that firms
could use. In either case, the fact that employers are
willing to pay a high price for an educated work force
seems to suggest that education produces social
benefits.
46
SCHOOLING AS A SIGNAL
 Education reveals a level of attainment which
signals a worker’s qualifications or innate ability to
potential employers.
 Informationthat is used to allocate workers in the
labor market is called a signal.
 There could be a “separating equilibrium.”
 Low-productivity workers choose not to obtain X years of
education, voluntarily signaling their low productivity.
 High-productivity workers choose to get at least X years
of schooling and separate themselves from the pack. 47
 EDUCATION AS A SIGNAL
Dollars Dollars

Costs

300,000 300,000 Costs

250,001 y 

200,000 Slope = 25,000 200,000

20,000 y Slope = 20,000

0 y Years of 0 y Years of
Schooling Schooling
(a) Low-Productivity Workers (b) High-Productivity Workers

Workers get paid $200,000 if they get less than y years of college, and
$300,000 if they get at least y years. Low-productivity workers find it
expensive to invest in college, and will not get y years. High-productivity 48
workers do obtain y years. As a result, the worker’s education signals if he is a
low-productivity or a high-productivity worker.
 EDUCATION AS A SIGNAL
 The low-productivity worker will not attend college if
$200,000  $300,000  ($25,001 y ) (6-16)
Solving for y implies that
y  3.999 (6-17)
 The high-productivity workers get y years of college whenever
$200,000  $300,000  ($20,000  y ) (6-18)
Solving for y yields
y5 (6-19)

49
IMPLICATIONS OF
SCHOOLING AS A SIGNAL
 Education
is more than a signal, it alters the stock of
human capital.
 Socialreturn to schooling (percentage increase in
national income) is likely to be positive even if a
particular worker’s human capital is not increased.

50
4. POST-SCHOOL HUMAN
CAPITAL INVESTMENTS
 Some important properties of age-earnings profiles:
 Highlyeducated workers earn more than less educated
workers.
 Earnings rise over time at a decreasing rate.
 The age-earnings profiles of different education cohorts
diverge over time (they “fan outwards”).
 Earnings increase faster for more educated workers.

51
 AGE-EARNINGS PROFILES
Men

2600
2300
2000 College Graduates
Weekly Earnings

1700
1400
1100 Some college
800 High school graduates
500 High school dropouts
200
18 25 32 39 46 53 60
52
Age
 AGE-EARNINGS PROFILES
Women

1400
1200 College Graduates
1000
Weekly Earnings

800 Some college

600
High school graduates
400
High school dropouts
200
18 25 32 39 46 53 60

Age 53
ON-THE-JOB TRAINING
 Mostworkers augment their human capital stock
through on-the-job training (OJT) after completing
education investments.
 Two types of OJT:
 General: training that is useful at all firms once it is
acquired.
 Specific: training that is useful only at the firm where it is
acquired.
54
IMPLICATIONS
 Firmsonly provide general training if they do not
pay the costs.
 Inorder for the firm to willingly pay some of the
costs of specific training, the firm must share in the
returns to specific training. Engaging in specific
training eliminates the possibility of the worker
separating from the job in the post-training period.

55
THE ACQUISITION OF HUMAN
CAPITAL OVER THE LIFE CYCLE
Dollars The marginal revenue of
an efficiency unit of
MC human capital declines as
the worker ages (so that
MR20, the marginal
MR20
revenue of a unit acquired
at age 20, lies above
MR30
MR30). At each age, the
worker equates the
marginal revenue with the
marginal cost, so that
more units are acquired
0 Q30 Q20 Efficiency Units when the worker is 56
younger.
AGE-EARNINGS PROFILES AND OJT
 Human capital investments are more profitable the
earlier they are taken.
 The Mincer earnings function:
 Log(w) = a·s + b·t – c·t2 + other variables.
 The “overtaking age” is t* and indicates the time
when the worker slows down acquisition of human
capital to collect the return on prior investments so
as to “overtake” earnings of those that did not
undertake similar investments.
57
THE AGE-EARNINGS PROFILE IMPLIED
BY HUMAN CAPITAL THEORY
Dollars
The age-earnings profile is
upward-sloping and concave.
Older workers earn more
Age-Earnings
because they invest less in
Profile human capital and because
they are collecting the
returns from earlier
investments. The rate of
growth of earnings slows
down over time because
workers accumulate less
Age
human capital as they get 58
older.