Over four decades of research has consistently found that non-school factors affect students' academic performance much more than do in-school factors.  As a result, at least half of the achievement gap is formed before students start school and at least another quarter forms during summer vacations.

During those same four decades, however, interventions designed to narrow the achievement gap have focused almost exclusively on changing schools.  While it may be possible to significantly narrow the achievement gap by changing only schools these reforms, collectively, have utterly failed to do so.  The racial achievement gap narrowed in the 70s and 80s, but has been steady ever since.  The income achievement gap, meanwhile, has steadily grown and the gap between upper- and lower-income students may now be twice as large as the gap between Black and White students.

The obvious response to the research and our policy failures is to re-focus our efforts on social policy interventions designed to change the home lives of students.  Indeed, newly formed Promise Neighborhoods are beginning to do just that (though they also include significant in-school interventions).  It remains unclear, however, exactly which environmental conditions and social factors experienced by those living in urban poverty are most significantly affecting academic performance and on which social policy can act.

I answer the first question by conducting an exhaustive synthesis of the research from across multiple fields and disciplines.  Over the past four years, I collected and coded hundreds of articles that estimate the impact of various social factors and environmental conditions on achievement and attainment.

In all but a few categories, research remains sparse and clear answers remain elusive.  I argue, however, that the evidence points us toward a collection of 20 factors/conditions upon which researchers and policymakers should focus their efforts.  I divide these factors and conditions into three categories: neighborhood/housing, health/health care, and family/home.

Neighborhood/housing factors include: crowding; disorder; homeownership; mobility/homelessness; noise; physical distress; social organization; and crime/violence.

Health/health care factors include: fitness; mental health, nutrition; prenatal care; stress; and toxins/pollution.

Family/home factors include: attachment; home environment; home resources; language exposure; parenting style; and parental warmth

Other factors may deserve attention as well, but this represents a good starting point for researchers to explore factors more in-depth and those designing Promise Neighborhoods and other social policy interventions to address.