Transcript what does the oecd alphabet soup say? wout ultee university of

WHAT DOES THE OECD
ALPHABET SOUP SAY?
WOUT ULTEE
UNIVERSITY OF HAIFA
NOVEMBER 18, 2012
THE ORGANIZATION FOR ECONOMIC
COOPERATION AND DEVELOPMENT
WAS FOUNDED IN 1961
IT WAS THE FOLLOW-UP TO THE AGENCY
RESPONSIBLE FOR THE DISTRIBUTION
AND
SURVEILLANCE OF THE MARSHALL AID
THE MEMBER STATES AT FIRST WERE THE
RICH WESTERN COUNTRIES OF THE WORLD
NOT LONG AGO ISRAEL BECAME A MEMBER
THE OECD DID NOT HAVE ANY
FORMAL
RESPONSIBILITIES
IT BECAME SOMETHING LIKE A
THINK TANK
AND IT AT FIRST TESTED THEORIES
OF RIGHT-WING US ECONOMISTS
SOMETIMES IT STILL DOES SO
UNEMPLOYMENT RATES AND LONG TERM
UNEMPLOYMENT RATES ARE HIGHER IN THE
EUROPEAN COUNTRIES THAN IN THE USA
BECAUSE UNEMPLOYMENT BENEFITS LAST LONGER
IN THE EUROPEAN COUNTRIES
BUT DID THE UNEMPLOYMENT RATE IN SOME
EUROPEAN COUNTRY MAINLY RISE BECAUSE
UNEMPLOYMENT BENEFITS BECAME EXTENDED?
MONEY FOR RETRAINING OF UNEMPLOYED
PERSONS LOWERS UNEMPLOYMENT RATES
MONEY FOR RETRAINING PERSONS ABOVE THE AGE
OF 50 YEARS DOES SO?
WHY IS NOT THE WHOLE WORLD
LIKE THE USA? QUESTIONS
VERSUS
WHY ARE THERE DIFFERENCES
BETWEEN EUROPEAN COUNTRIES?
QUESTIONS
THE OECD WAS PROBABLY THE FIRST
INTERNATIONAL ORGANIZATION TO
PUBLISH FOR ITS MEMBER COUNTRIES
TIME SERIES FOR
GROSS DOMESTIC PRODUCT PER CAPITA
IN
CONSTANT DOLLARS
THE OECD WAS CRITICIZED FOR THIS BECAUSE
MONEY DOES NOT BUY HAPPINESS
AND BECAUSE VARIOUS STATE MEASURES LIKE
SCHOOLS AND HOSPITALS CONTRIBUTE TO THE
LEVEL OF LIVING
IN THE 1980S THE UN (UNITED
NATIONS) STARTED
PUBLISHING THE HUMAN
DEVELOPMENT INDEX
THIS INDEX WEIGHS FOR COUNTRIES
MEASURES FOR PER CAPITA GDP WITH
MEASURES FOR EDUCATION,
SCHOOLING, HOSPITALS, PIPED WATER
ETC
THE RANKING OF THE COUNTRIES DOES
NOT DIFFER MUCH
BUT CUBA IS MUCH HIGHER
THE OECD AND THE ILO (INTERNATIONAL LABOR
ORGANIZATION) PUBLISHED TIME SERIES ON THE PERCENT
OF NATIONAL INCOME GOING TO SOCIAL SECURITY
EXPENDITURES
THIS MEASURE WAS HEAVILY CRITICIZED BY THE SWEDISH
SCHOOL ON WELFARE STATE RESEARCH RUN BY AT FIRST
WALTER KORPI AND LATER JOACHIM PALME
EXPENDITURES ON UNEMPLOYMENT BENEFITS IN A
COUNTRY MAY RISE, NOT BECAUSE UNEMPLOYMENT
BENEIFTS BECOME MORE GENEROUS
BUT BECAUSE THE UNEMPLOYMENT RATE ROSE WHILE
INDIVIDUAL UMEPLOYMENT BENEFITS WENT DOWN
THE SWEDISH SCHOOL CAME UP WITH SUCH INDIVIDUAL
MEASURES FOR THE OECD COUNTRIES
THE BEST CRITICISM OF MEASURES LIKE
GROSS DOMESTIC PRODUCT PER CAPITA
IS THAT IT PERTAINS TO AN AVERAGE
INHABITANT OF A COUNTRY
BUT THAT THE DISPERSION AROUND THE
MEAN IS NEGLECTED
IT MAY WELL BE THE CASE THAT A COUNTRY
WITH A HIGHER PER CAPITA GDP
HAS A HIGHER PERCENTAGE OF
INHABITANTS BELOW SOME LOW INCOME
LEVEL
THE INTERNATIONAL ORGANIZATION DOING THE MOST
AT FIRST FOR MAKING MEASURES ON INCOME
INEQUALITY AND POVERTY RATES AVAILABLE WAS THE
WORLD BANK
LATER THE OECD CAUGHT ON
POVERTY RATES ARE RATHER ARBITRARY
YET THE RANKING OF COUNTRIES AFTER POVERTY RATES
PROBABLY IS NOT AFFECTED BY DECISIONS ABOUT
WHERE EXACTLY TO PLACE THE POVERTY LINE
THE TWO PRIME MEASURES FOR INCOME
INEQUALITY ARE THE GINI AND THE QUINTILE
RATIO
THE GINI RUNS FROM ZERO TO 1, WITH ZERO
STANDING FOR FULL EQUALITY AND 1 FOR ONE
UNIT GETTING ALL THE INCOME
THE QUINTILE RATIO DIVIDES THE INCOME
SHARE OF THE RICHEST 20% BY THE INCOME
SHARE OF THE POOREST 20% OF UNITS
INCOME INEQUALITY USUALLY IS NOT
MEASURED AT THE INDIVIDUAL LEVEL (A LOT
OF INDIVIDUALS DO NOT HAVE INCOME)
BUT AT THE LEVEL OF HOUSEHOLDS, TAKING
THE INCOME OF ALL HOUSEHOLD MEMBERS
INTO ACCOUNT, EVEN IF THAT IS NILL
SO, IF BOTH THE HUSBAND AND THE WIFE
WORK, THEIR INCOMES ARE ADDED UP
TO DETERMINE THE INCOME OF A
HOUSEHOLD MEMBER, ALSO THE NUMBER OF
NON-WORKING CHILDREN IS TAKEN INTO
ACCOUNT (A CHILD OFTEN IS TAKEN FOR 0,4
ADULTS)
INCOME INEQUALITY PER HOUSEHOLD
MEMBER USUALLY IS NOT MEASURED AS
TAXABLE (PRE-TAX) INCOME
BUT AS POST-TAX (DISPOSABLE )INCOME
PER HOUSEHOLD MEMBER
SINCE THE 1980S THE STATISTICAL OFFICES OF MOST
OF THE RICH COUNTRIES IN THE WORLD
ON A YEARLY BASIS PRODUCE DIGITAL FILES FOR THE
INCOME OF THE HOUSEHOLDS IN THEIR COUNTRY
THIS OFTEN IS DONE FROM TAX REGISTERS, BUT
SOMETIMES SURVEY DATA ARE USED
LIS (THE LUXEMBURG INCOME STUDY) WAS THE FIRST
UNIVERSITY INSTITUTE TO USE THESE FILES AND
PRODUCE COMPARABLE INCOME INEQUALITY
DATA FOR RICH COUNTRIES
IN SOME COUNTRIES INTEREST ON MORTGAGES ARE
DEDUCTED FROM TAXABLE INCOME, IN OTHERS NOT
LIS AND NOW THE OECD DEAL WITH SUCH MATTERS
IN ONE OF ITS RECENT STUDIES THE OECD
SHOWED THAT, MEASURED BY GINIS AND
QUINTILE RATIOS, INCOME INEQUALITY ROSE
IN MOST OF THE OECD COUNTRIES BETWEEN
1985 AND 2005
ACCORDING TO THE OECD, THE RISE WAS
PARTICULARLY STRONG IN THE
UNITED STATES AND THE UNITED KINGDOM
ACCORDING TO THE OECD, INCOME
INEQUALITY ROSE A BIT IN THE NETHERLANDS
AND MORE IN ISRAEL
WHY DID INCOME INEQUALITY RISE IN MOST OECD
COUNTRIES?
THE OECD TESTED IN ONE OF ITS REPORTS
EXPLANATIONS INVOKING DEMOGRAPHIC CHANGE
IF RETIRED PEOPLE HAVE LOWER INCOME, AND IF
THE RETIRED INCREASE AS A PERCENTAGE OF THE
POPPULATION, INCOME INEQUALITY WILL INCREASE
WITHOUT ANY OF THE DEMOGRAHIC GROUPS OF A
COUNTRY’S POPULATION HAVING A CHANGE IN
INCOME
THE MOST INTERESTING DEMOGRAPHIC
EXPLANATION INVOKES INCREASED
FEMALE
LABOR FORCE PARTICIPATION AND WAS
PROPOSED BY THE SOCIOLOGIST GOSTA
ESPING-ANDERSEN
IF MARRIED WOMEN WITH A HIGHER LEVEL OF
EDUCATION ARE LESS LIKELY TO QUIT THEIR JOB
WHEN GIVING BIRTH TO A CHILD THAN MARRIED
WOMEN WITH A LOWER LEVEL OF EDUCATION
HOUSEHOLD INCOME INEQUALITY WILL RISE
AND IT WILL RISE EVEN MORE IF WOMEN WITH A
HIGHER LEVEL OF EDUCATION ARE MARRIED TO
MEN WITH A HIGHER LEVEL OF EDUCATION
(EDUCATIONAL HOMOGAMY)
HOWEVER, ALL THESE DEMOGRAPHIC EXPLANATIONS TAKEN
TOGETHER DO NOT FULLY EXPLAIN THE RISE IN
INCOME INEQUALITY
WAGE DIFFERENTIALS HAVE BECAME LARGER
NOT ONLY WAGE DIFFERENTIAL BETWEEN PERSONS WITH A
HIGHER AND A LOWER LEVEL OF EDUCATION
BUT ALSO WAGE DIFFERENTIALS BETWEEN ECONOMIC SECTORS
SUCH AS FINANCE PLUS BANKING VERSUS ALL OTHER SECTORS
WHAT ARE THE WAGE DIFFERENTIALS IN ISRAEL BETWEEN THE
HIGH TECH SECTOR AND THE OTHER SECTORS?
THE MOST COMMON THEORY IN SOCIOLOGY
EXPLAINING INCOME INEQUALITIES
BETWEEN COUNTRIES IS GERHARD LENSKI’S
SOCIAL-DEMOCRACY HYPOTHESIS:
IN COUNTRIES GOVERNED FOR A LONGER
TIME BY SOCIAL DEMOCRATIC
GOVERNMENTS
INCOME INEQUALITIES, EVEN AFTER TAKING
PROGRESSIVE TAXATION INTO ACCOUNT, ARE
SMALLER THAN IN COUNTIRES GOVERNED
FOR A LONGER TIME BY CONSERVATIVE
GOVERNMENTS
DOES THE OECD ADDRESS THIS EXPLANATION?
NOT REALLY!
THERE IS ONE THEORY DOWNPLAYING
THE FINDING THAT INCOME INEQUALITY,
OF ALL THE RICH
COUNTRIES, IS LARGEST IN THE USA
THE USA MAY HAVE LARGE INCOME
INEQUALITIES, BUT IT ALSO HAS A LOT OF
INCOME MOBILITY
INDEED, THE USA-INEQUALITIES PROVIDE
THE INHABITANTS OF THE USA WITH THE
MOTIVATION TO CLIMB UP IN INCOME
DOES THE OECD ADDRESS THIS RIGHTWING EXPLANATION? YES!
SINCE A BOOK BY WERNER SOMBART
FROM 1906
IT IS A COMMON THEORY THAT THE
USA HAS MORE SOCIAL MOBILITY THAN
EUROPEAN COUNTRIES
PARTICULARLY FATHER-SON
OCCUPATIONAL MOBILITY
THE SOCIOLOGIST SAYMOUR MARTIN
LIPSET REPEATED THIS THEORY IN A
BOOK IN 1959, AND EVER SINCE
HOWEVER, LIPSET FOUND IN 1959
THAT MOBILITY RATES ARE
PRETTY MUCH THE SAME IN
INDUSTRIAL COUNTRIES
IN 1978 JOHN GOLDTHORPE FROM
BRITAIN AND ROBERT HAUSER
FROM THE USA TOOK SOMBART’S
THESIS NOT TO BE ABOUT
ABSOLUTE MOBILITY RATES,
BUT ABOUT RELATIVE MOBILITY
CHANCES
RECENT DATA OF SOCIOLOGISTS SHOW
THAT RELATIVE MOBILITY CHANCES ARE
HIGHEST IN THE SCANDINAVIAN
COUNTRIES AND QUITE LOW IN THE USA
HOWEVER, SOCIOLOGISTS MEASURE
MOBILITY WITHIN A SYSTEM OF
CLASSES OR ALONG AN OCCUPATIONAL
PRESTIGE SCALE
ECONOMISTS LIKE TO STUDY INCOME
MOBILITY
THE FIRST STUDIES ON FATHER-SON
INCOME MOBILITY FOR
UNREPRESENTATIVE SAMPLES FOR THE
USA FOUND CORRELATIONS OF AT
MOST 0.2 FOR FATHER’S YEARLY
INCOME AND SON’S YEARLY INCOME
SOME DECADES LATER
THIS WAS MUCH LOWER THAN THE
CORRELATION OF AT LEAST 0.4 THAT
SOCIOLOGISTS FOUND WITH RANDOM
SAMPLES BETWEEN
THE OCCUPATIONAL PRESTIGE OF
FATHERS AND SONS
LATER STUDIES BY ECONOMISTS ON
INCOME MOBILITY COMPUTED THE
CORRELATION BETWEEN FATHER’S
INCOME AVERAGED OVER THREE YEARS
AND SON’S INCOME AVERAGED OVER
THREE YEARS
THEN THE CORRELATION ROSE TO
ABOUT 0.4
THE SOCIOLOGISTS HAD COME UP
WITH THE RIGHT VALUE
OCCUPATIONAL PRESTIGE STANDS FOR
LONG-RUN INCOME
THIS IS THE OECD RESULT ON INCOME
INEQUALITY AND
INCOME MOBILITY
FORGET THE RIGHT PART OF THE FIGURE
HIGHER INCOME INEQUALITY
GOES TOGETHER WITH LESS
INCOME MOBILITY
OF LATE, THE OECD HAS BEEN MAKING THE
HEADLINES WITH PISA
PISA = THE PROGRAM FOR INTERNATIONAL
STUDENT ASSESSMENT
FOR PISA, LARGE NUMBERS OF STUDENTS
AGED 15 YEARS FROM THE OECD MEMBER
COUNTRIES MAKE STANDARDIZED TESTS
ON LANGUAGE COMPREHENSION AND
MATHEMATICS
PISA ALSO COLLECTS DATA ON THE
SCHOOLS OF THESE STUDENTS, ON THEIR
TEACHERS, AND ON THEIR PARENTS
THE RANKINGS OF COUNTRIES AFTER
THE SCORES OF THEIR STUDENTS TEND
TO CAUSE POLITICAL UPROARS IN
SEVERAL COUNTIRES
FOR INSTANCE, GERMANY WAS
SHOCKED TWO YEARS AGO ABOUT ITS
LOW RANKINGS
THE HIGH RANKING OF FINLAND BELIED
THE HYPOTHESIS THAT IF TEACHER’S PAY
IS HIGHER, STUDENT’S SCORES ARE
HIGHER TOO
FINNISH TEACHERS ARE NOT PAID A LOT
PISA’S BOSS ANDREAS SCHLEICHER ANSWER TO A QUESTION OF
A JOURNALIST IN THE INTERNATIONAL HERALD TRIBUNE OF
NOVEMBER 12, 2012
Q: I WANT TO ASK YOU ABOUT A RESULT OF YOURS ON CLASS SIZE
A: IT IS CLEAR THAT IF EVERYTHING ELSE IS EQUAL, A SMALLER
CLASS IS BETTER THAN A LARGER CLASS. BUT THAT IS NOT THE
RIGHT QUESTION. THE RIGHT QUESTION IS: IF YOU HAVE ONE
DOLLAR EXTRA TO SPEND, DO YOU PUT IT INTO A SMALLER CLASS,
A BETTER TEACHER, MORE LEARNING TIME? WHAT OUR
RESEARCH VERY CLEARLY SHOWS IS THAT IF YOU HAVE TO MAKE A
CHOICE BETWEEN A GREAT TEACHER AND A SMALL CLASS,
CHOOSE THE GREATER TEACHER.
STUDIES LIKE PISA, STUDIES COLLECTING SCORES
FOR STUDENTS, CHARACTERISTICS THEIR
PARENTS, THE SIZE OF THE CLASS THEY ATTEND,
THE FINANCING OF THEIR SCHOOL,
WERE UNDERTAKEN IN THE USA IN THE 1960S
ACCORDING TO HAUSER, THEIR MODEL IS NOT
TO BE FOLLOWED, WHILE PISA DOES
HAUSER SAID SO AT SENATE HEARINGS
HE REPEATED THIS AT THE MEETING OF THE
RESEARCH COMMITTEE ON SOCIAL
STRATIIFICATION AND MOBILITY IN YALE IN 2009
HAUSER’S ARGUMENT IS THAT PISA IS CROSS-SECTIONAL
WHILE DYNAMIC DATA ON THE SAME STUDENT ARE
NEEDED
A STUDENT THAT MOVED FROM A BAD SCHOOL TO A
GOOD ONE IS TREATED IN THE CROSS SECTIONAL
ANALYSIS AS A STUDENT WHO HAS ATTENDED ALL HIS
LIFE THIS SCHOOL
BUT EXPOSURE TO FIVE YEARS OF A BAD SCHOOL IS NOT
WASHED OUT BY ONE YEAR AT A GOOD SCHOOL
SCHOOL EFFECTS SHOW UP GRADUALLY: THEY ARE WEAK
IN THE FIRST YEAR, AND BECOME STRONGER LATER
PISA’S STUDENTS ARE ALL IN THE SAME CLASS AND OF
THE SAME AGE
TO PARAPHRASE SCHLEICHER:
IF YOU AS A RESEARCH SOCIOLOGIST HAVE ONE
DOLLAR EXTRA TO SPEND
WOULD YOU SPEND IT ON REPEATING CROSS
SECTIONAL RESEARCH FOR OTHER STUDENTS
OR WOULD YOU SPEND IT ON RESEARCH FON
THE SAME STUDENTS SOME YEARS LATER?
ACCORDING TO METHODOLOGY TEXTBOOKS THE
ANSWER IS CLEAR: FOLLOW BEGINNING
STUDENTS THROUGH THEIR EDUCATIONAL
CAREER
UNCLARITIES AND ERRORS OF PISA
AND THE OECD
THIS IS THE MAIN DIAGONAL OF A MOBILITY TABLE
A CURIOUS READER WOULD ALSO LIKE TO SEE THE OFFDIAGONAL CELLS
MEN FROM THE BOTTOM
FIFTH HAVE FOUR STEPS TO
TAKE TO THE HIGHEST FIFTH,
MEN FROM THE MIDDLE
FIFTH TWO STEPS
THERE IS LESS MOBILITY OF
FOUR STEPS THAN OF TWO
STEPS IN BOTH COUNTRIES
IS THIS ENLIGHTENING?
MOBILITY OF FOUR STEPS IS
LESS COMMON IN THE USA
THAN IN DENMARK
HOW ABOUT MOBILITY OF ONE STEP UP
AND ONE STEP DOWN?
IS IT POSSIBLE FOR A PERSON IN THE
TOP QUINTILE TO GO ONE STEP UP?
IS IT POSSIBLE FOR A PERSON IN THE
BOTTOM QUINTILE TO GO ONE STEP
DOWN?
OECD, EDUCATION AT A GLANCE 2012, PAGE 104
CHART 6.1
PROPORTION OF 20-34 YEAR OLD STUDENTS IN
HIGHER EDUCATION WHOSE PARENTS HAVE LOW
LEVELS OF EDUCATION
PROPORTION OF PARENTS WITH LOW LEVELS OF
EDUCATION IN THE TOTAL POPULATION
THE ODDS OF BEING A STUDENT IN HIGHER
EDUCATION IF PARENTS HAVE LOW LEVELS OF
EDUCATION
WHAT ARE ODDS?
COMPUTE ODDS
RATIOS!