What are the most promising solutions for closing the gap? Tony Moss Researcher / Data Analyst Kansas State Dept. of Education Gap Workgroup Meeting KSDE, 1 December 2014

What we’ll cover: Review the previous presentations – What are gaps and what causes them? – Where are the gaps greatest? What about schools—how do they contribute to gaps? What are the most promising solutions? – Early childhood remedies – Setting up a process to design, develop and test remedies: the Washington State Institute for Public Policy example What do we want to recommend? Next steps?

What are gaps? 1.Gaps are measures of the health and success of a society.  Bigger gaps => worse health, crime, more social costs, large opportunity costs  Smaller gaps => better health, less crime, fairer, lower social costs, higher incomes 2.Gaps are also leading indicators of a society’s economic future.  Bigger gaps => more economic, social, and political instability  Smaller => wider prosperity and greater economic and social stability

American gaps show in international comparisons: UNICEF’s rating of child well being places the U.S. near the bottom of developed countries. U.S. spends more but gets less in education & health care; no practical support for single moms.

What causes gaps? Because early experience shapes genetic expression and the architectures of many biological systems, both health and academic gaps originate in early child development and social interactions, especially with primary caregivers. The stressors and risks tied to lower social and economic status (SES) have become much stronger predictor of gaps than ethnicity or race. Once the architectures of developing brains and biological systems are shaped by stressors and social interactions, gradients in health and achievement persist. Plasticity is limited. e.g. Japanese soldiers; MAOA genes

Poverty Trends within Ethnic Groups Kansas public schools only, all grades and preschool,

African-AmericanHispanicWhite Mixed, Other, Unknown n% of pop.n n n Child Characteristics males9, , , , females8, , , , male/female ratio Kansas Newborns by Ethnicity & Characteristics, from Kansas Birth Records, meansdmeansdmeansdmeansd birth weights males3, , , , females3, , , , Maternal Characteristics age yrs. of ed percentages % unmarried % foreign born

Different poverty environments explain differences in academic gradients.

Low SES prompts poor postnatal responsiveness Low parental SES is associated with parental irritability, depression, anxiety Parents typically invest less in children when the parents’ survival seems threatened (parental investment theory) Parental stress also leads to reduced nurturance (less warmth, less sensitivity, less language play) and higher likelihood of harsh and inconsistent discipline, neglect, and abuse The quality of early parenting can predict children’s emotional and behavior patterns years later, and risks, like dropping out of school

We are talking about probabilities, not inevitabilities. Poor parents can moderate the effects of poverty, but the odds are against them.

Early stressors interfere with bonding and attachment, the foundation of later social skills. Insecure bonding is biologically embedded in the HPA axis, which predicts more reactive stress responses and future social conflict.

Quality of Parenting as Measured by the HOME-Short Form Brooks-Gunn & Markham, The contribution of parenting to ethnic and racial gaps in school readiness.

The brain is sculpted by experience, especially so from birth through age 5. This foundational architecture shapes some capacities for life.

Like language skills, health gradients emerge in early childhood and persist:

Nationally, white males live, on average, to age 75, while black males live to 68. In Kansas, the gap is a little larger, with the average white males living to 76, and the average black males living to a little more than 68 years. Earlier gaps have their counterpart at the end of life, in life expectancy.

return per $1 invested Investments in high quality early childhood return 3 X as much as school investments, and 8 X as much as adult remediation programs. Investments in high quality early childhood return 3 X as much as school investments, and 8 X as much as adult remediation programs.

Another cause of gaps is inferior early child care. Of the 45 percent of 2-year-olds in low- quality day care, the majority are very poor, and African- American or Hispanic. G.M. Mulligan and K.D. Flanagan, Age 2: Findings from the 2-Year-Old Follow-Up of the Early Childhood Longitudinal Study, Birth Cohort (ECLS-B), U.S. Dept. of Education, NCES, August 2006

The rapid growth in the proportion of working and single mothers is a global phenomenon. Countries that have high-quality Early Education and Child Care (EECC), have reduced the gaps associated with poverty and single-parenthood.

Increasing economic inequality is also increasing academic gaps.

As income and wealth gaps have increased over the last 3 decades, the upper classes are investing more in their children, increasing academic gaps by 30 to 60 percent.

In sum, the 3 most important early factors for a child’s future academic success are: Rich language & intellectual stimulation and reciprocity (play) Warm, positive bonding & attachment & emotional reciprocity (play) Normal, not chronically stressful or neglectful, social environments

3 Profound implications: 1.The quality of early childhood, in language immersion, responsive care, play, bonding and attachment, experience, nutrition, and health, are the foundations of academic and social achievement, and adult health. 2.If we want to reduce health and academic gaps, we need to reduce the stresses of maternal and child poverty and broadly improve the quality of parenting, and of early childcare and education. 3.While the up-front costs of high-quality childcare and parental programming can be expensive, the savings and social improvements are much greater than the expense.

And one big implication for schools: The success of the whole educational enterprise is even more dependent on developmental events, especially in early childhood, than was previously known. Working memory and cognitive controls Students’ social predilections and skills Students’ behaviors (persistence, engagement, motivation) Especially language skills—the primacy of reading Some disabling conditions—antisocial behaviors, ADHD— all originate in early development.

But what may tip the balance sheet in favor of high- quality early childcare and pre-kinder, are savings in justice, health and other costs. From Perry Preschool:

Because of the long time-horizon, the Return-on-Investment grows larger as participants become adults.

Re-allocation toward prevention rather than remediation could make government more efficient.

In the 2 nd presentation, we answered 2 questions: In Kansas, is poverty a stronger predictor of lower academic results than ethnicity? Which ten Kansas districts have the greatest numbers of Hispanics, African- American, and lower SES students?

Does ethnicity or poverty better predict academic growth? We used a generalized mixed model to separate school effects from individual effects.

When we selected the districts with the largest numbers of below proficient and low-income students, regardless of race, we get these districts: Number of Below proficient, Lower-Income Students

Overlap of Kansas Districts with the Largest Numbers of Below-Proficient & Low-Income Students by Ethnic Group

Questions? Comments? Break?

Yes, after: 1.poverty’s effects on development, 2.inferior child care, 3.increasing investments in children by the middle and upper classes, disadvantaged children are often in: 4.schools and neighborhoods with higher teacher turnover, and negative peer influences, among other disadvantages. What about schools? Don’t they contribute to gaps?

Washington State’s study of schools’ contributions to gaps is typical: Inequitable distribution of skilled, experienced teachers Insufficient and inequitable school funding Inadequate, obsolete, and unbalanced distribution of facilities, technology and instructional materials Inequitable access to demanding, rigorous pre-college coursework Institutional racism Lack of cultural competence among teachers and staff Washington State Institute for Public Policy’s 2008 review of the achievement gap literature

Ed. Testing Service did 2 correlational studies that identified these school factors: Curriculum rigor Teacher preparation Teacher experience & attendance Teacher turnover Teacher expertise in the subject taught Class size Less instructional technologies Less safety Less parental involvement Based on other studies, we could add attendance and engagement strategies and policies. Paul Barton, Parsing the Achievement Gap, I & II, 2003 & 2009 respectively

Let’s focus on teacher turnover rates. How do the 10 districts we identified compare to national teacher turnover rates?

International comparisons suggest that nationally, U.S. schools have a problem with teacher turnover, teacher selection, and teacher retention. Successful reformers—Finland, Singapore, and South Korea—have yearly turnover rates from 1 to 3 percent. But they are highly selective and carefully control their teacher labor markets.

Improving the Quality of Teachers Peter Dolton & Oscar Marcenaro-Gutierrez, 2011, Teacher’s Pay and Pupil Performance, s/download/cp352.pdf s/download/cp352.pdf

What do teachers earn compared to others with similar educational levels? In OECD’s Education at a Glance (2012), only Austrian teachers were making less.

International Comparisons of Policies Aimed at Attracting and Retaining Teachers √ significant priority in the country, best-in-class practice policies to attract/retain top teachersSingaporeFinlandS. KoreaU.S. 1 selective admissions to teacher training √√√most programs not selective 2government paid teacher training√√ students finance own education 3 government regulates supply of teachers to match demand √√√oversupply of teachers 4professional working environment√√√variable working conditions 5competitive compensation√ √ compensation not attractive to many students 6 cultural respect accorded to teaching √√√ respect not comparable to other nations 7teaching considered as a career√√√ relatively high attrition in early years 8 robust opportunities for career advancement √ limited opportunities for advancement 9performance pay for teachers√ √limited performance pay Source: Interviews, McKinsey research, Auguste, B., Kihn, P. and Miller, M. (2010). Closing the talent gap: Attracting and retaining top-third graduates to careers in teaching.

Kansas teachers’ wages were even lower—70 percent of comparable occupations in 2006.

Among labor market economists, feminization of a profession is a marker of lower status and lower pay.

KSDE counts of teacher registrants to teacher hires from the Educators’ Education and Employment tables. Only new teachers are included—latent supply of already certified teachers is ignored. Lower pay is related to a lack of selectivity and over-supply:

The 3 comparative countries select only the top 20 percent or higher. Finland selects for social skills, too.

McKinsey’s researchers did an international comparative study of teacher quality. Improving teacher starting salaries to somewhere above $65K per year, would quadruple the percentage of applicants from the top third of university cohorts. Current salaries of beginning teachers in the 10 school districts we identified is $43K. Auguste, Kihn, & Miller (2010). Closing the Talent Gap: Attracting and Retaining Top-Third Graduates to Careers in Teaching

What would it take to put the most talented teachers into the highest need schools and retain them?

Should adolescents be taught about optimum child development and how to make their future children smarter? Beyond putting, and keeping, the best teachers in the neediest schools, there are lots of promising but untested ideas.

Other promising, unfolding ideas: 2 Generation Programs combine services for both the parent and the child, workforce development + high quality EECC; maternal health care + EECC Chase-Lansdale & Brooks-Gunn, Two-Generation Programs in the Twenty-First Century

There are now curricula designed to strengthen particular parts of the brain needing further development.

We are really talking about recommending an experimental design and testing service: 1.Use well-developed theory to design interventions. 2.Engagement & retention of parents is crucial—we must give them strong reasons to invest in the service. 3.Use controlled, randomized efficacy pilot trials & tweak as needed. 4.In natural settings, use randomized effectiveness trials, & tweak again. 5.Move on to replication trials, again with randomized controls, in new settings or with different populations. 6.Measure benefits and costs across development (longitudinal data).

Questions? Comments? Break?

An example of a well-timed intervention: KWWo#t=68 KWWo#t=68

Early Childhood Remedies 1.High quality, center-based care (EECC) large latent effects on many child outcomes, including academic, and improved parenting / personnel quality matters 2.Home-visits by well-trained nurse practitioners large positive effects on parental nurturance / latent child effects 3.Family literacy programs like 30 Million Words higher language scores than controls / no randomized studies of 30 Million Words yet 4.Changing parental behaviors to fix child behavior problems like Incredible Years (often used in Head Start) large positive effects on reducing problem behaviors 5.Pre-Kindergarten with well-paid teachers with 4-year degrees in Early Childhood and Education

Some big cautions: Parenting interventions can work but many have failed because of poor design, poor testing and development, and especially because of lower- quality personnel. Programs must give parents reasons to care and participate (e.g., sector training + EECC; health services + EECC). The work we’re doing here as a committee doesn’t qualify us to make well-informed, scientifically defensible recommendations. But we could recommend creating a process that does.

High-Quality Early Education Example: Carolina Abecedarian Project 111 infants born between 1972 & 1977 Random assignment, treatment group in preschool for 5 years, 6 to 8 hrs. / day, 5 days per week Game-based activities that cultivated social, emotional, and cognitive development, especially language Control group got nutritional supplements, social services and health care 67

Abecedarian Results at age 21: Half-year more education 42% vs. 20% continuing their education 36% vs. 14% attending college 47% vs. 27% in skilled jobs 26% vs. 45% teen parents Mothers of participants, esp. those who were teen moms, achieved higher education and employment status 68

Chicago Child-Parent Centers (CPC) are the most realistic model for states. One of the largest, publically-funded (Title I) early childhood interventions. 4 decades of operation. 20 CPCs in lowest income neighborhoods of Chicago. Depending on the methods used, and the age of the participants, the per-dollar return on investments have run from $7 at age 19 to $22 at age 26. Even larger returns might be realized by experiments in the design, e.g. coupling with a visiting nurse program; adding a Thirty-Million-Words curriculum; starting at age 1 or 2 through age 3 or 4.

Theme: the Quality & Training of Personnel Magnuson & Waldfogel’s Optimum Scenario: If We Increase the Quality of Head Start and Other Preschools, & Admit All Children Below 200% Poverty, at 100% Enrollment Quality Effect in Standard Deviations Increases in Population Average Decrease in Gap BlacksHispanicsWhites Black / White Hispanic / White.111%16%5% 12%20%.217%22%8% 18%28%.323%28%10% 24%36% Katherine A. Magnuson & Jane Waldfogel, 2005, Early Childhood Care and Education: Effects on Ethnic and Racial Gaps in School Readiness

Integrating and Rationalizing Existing Services & Adding Effective New Services? servicepurposefunding strong evidence of reducing gaps 1Private Child Carechild careparents, ? 2Healthy Startprevent infant mortality & disabilitiesfederal in health 3Parents as Teachersparent education, optimize developmentstates, foundations ? 4Healthy Familiesparent education, supportstate, local ? 5Early Head Startcare, parent ed, preventionfederal, state ? 6Part C – tiny-kbirth to 3 children w disabilitiesfederal, state, local ? 7Head Startoptimize developmentfederal yes* 8Sec 619 Early Childhood Special Ed3 to 5 children w disabilitiesfederal ? 94-Year-Old At-Riskchild care, educationstate yes* 10Pre-Kindergartenchild care, education parents, federal, state, local yes* 11Migrant Family Literacyfamily literacy, educationfederal ? 12Women, Infants, and Childrennutrition, preventionfederal in health 13Nurse Home-Visits for Prenatal + 2 Yearsprevention, optimize development? yes, but not academics 14High-Quality Early Ed and Child Careeducation, optimize development? yes* 15Family Literacy (e.g. 30 Million Words)optimize development, educationfoundations, ? ? * = effectiveness depends on the quality and training of staff; ? = reviewer unaware of the results of randomized trials or similar evaluations

Experiment: One-stop services Lower income parents may not have the time or energy to get all the services their children need when the services are spread out Concentrated expertise in the services can realize economies of scale & mutual support, e.g. WIC is well positioned to provide more health services, e.g., screenings and referral for maternal depression, child screenings past the first year. Transitions between early childhood services, kindergarten and elementary school can be smoothed by known staff

The Washington State Institute for Public Policy

They’ve combined a Consumer Reports-like reporting method with an investment portfolio:

What would the ideal continuous improvement model look like in Kansas? Which agencies should be on the board? – KDHE oversees early childhood care – Many federal initiatives, Head Start, WIC – How should the legislature & governor be involved? Who should be on the Technical Advisory Committee? National & state experts? Who should actually do the work? What about the hosting agencies?

Questions? Comments? What do we want to recommend? Next steps?