Controversies: Does Head Start Work?

This is part two of my early education series.

For additional information on the general impact of preschools, see part one: Is Preschool Necessary?


A lot has been written recently on universal prekindergarten in Oklahoma since President Obama’s State of the Union speech.  Oklahoma’s program has been seen as a national model, due to its broad reach (over 74% children in pre-kindergarten) and its adherence to National Institute of Early Education Research quality benchmarks

There are, of course, detractors. Small government idealists, for example, are against any expansion in the system, especially what can amount to glorified childcare in many cases – and center-based childcare, some research suggests, can do more harm than good for children’s emotional development. Other detractors point out that four-year-olds need to play, not sit at a desk and learn.

Those assertions are valid.  Any literature study on childcare and preschools will indicate that quality play-based education is the way to go for young children – and that this education does not need to be provided by the state, or by private educators. Parents, indeed, can provide all that a four year old needs.

Regardless of these facts, the discussion on prekindergarten is jumbled between mixing childcare, preschool, rote education, and programs for low-income students such as Head Start all into one jumble.

I address much of this in my first article, with numerous citations.  However, it all comes down to this: quality matters. When we’re talking about childcare vs. nanny care or familial care, it matters if the childcare is quality and if the home environment is toxic to the child. When we’re talking about the benefits of preschool, it matters if the preschool has teachers with degrees, high pay, low class ratios, and a belief in play-based education. And, when we’re talking about measurable results, it matters what we’re measuring. A child may not be reading at five years old – but may be showing other measures of early literacy that indeed might be more important at that age. So, the quality of the testing, indeed, matters.

In the end, I concluded what many have concluded before me. Is preschool necessary? Does it work? In the end,  no. It is not necessary. Your child may gain some short term benefit toward being exposed to a classroom environment and being with his or her peers, but keeping your child home from preschool will likely not hamper him or her in the long run. That is, if you are an involved enough parent to be researching this in the first place. But, by all means, for a short-term enrichment, and for your own sanity – do it.  And, if you are choosing between center-based care or a preschool with actual teachers – sure. Go for the preschool – there are fewer negatives.

I addressed all that to some extent in my article. However, I did not address Head Start or universal pre-kindergarten. Suddenly, we’re dealing with massive taxpayer-funded programs that have yet to show measurable results.  On the other hand, we are faced with disadvantaged children, those who may not be hearing enough words in their daily lives to assist in literacy, those who may be facing substandard childcare.

Does, the question goes, Head Start work?

By many indicators, no. Not on the national scale.

According to the Head Start impact study, a randomized experiment funded by the federal government, there was indeed some short-term impacts for three-year-olds and four-year-olds, specifically pre-reading, pre-writing, and vocabulary. However, there were no benefits on oral comprehension, pre-math skills, or phonological awareness.

When the study moved into the first-grade realm, the study showed diminished impacts, apart from vocabulary and oral comprehension. In fact, math skills were higher in the control group than in those in the Head Start group.

Finally, the most damning – the third grade followup.  Although there were significant improvements in indicators during the actual duration of the Head Start program, by third grade, most of the those improvements had dissipated.  Other than a few indicators, there were no differences between the Head Start group and the control group by third grade. The only real positive impact was for the three-year-old cohort in the third grade, with average reading/language arts proficiency scores favoring the Head Start group.

What about Head Start’s other missions?

Most people, including myself, focus on the cognitive development portion of Head Start’s mission. However, Head Start was founded on a ‘whole child’ model, including but not limited to medical, dental, and mental health care; nutrition services; helping emotional and social child development;  and parental training.

The Head Start impact study did in fact measures some of those indicators in addition to cognitive development. Alas, the same patterns applied: some modest gains early on, mostly erased by third grade.

So, Head Start is useless?

Probably a harsh judgment.   It may be true overall, but there are a few confounding factors.

For example, some of the children in the Head Start impact study did not actually enroll in Head Start, and some in the control group did enroll  (although usually in a Head Start center not involved in the study).  A 2008 reanalysis of the preschool outcomes obtained stronger positive effects for the Head Start group when accounting for this change.

In addition, not all Head Start centers are created equal. Some of the Head Start group attended centers with lower emphasis on language and literacy or math.  Forty percent of the students had teachers without a postsecondary degree. The majority had teachers with less than 25 hours of training the year before.

Third, the study did not measure the impact of Early Head Start.  One year of Head Start education likely will not provide lasting changes and/or undo the disadvantages children already acquire by age four. Even though there may be short term gains, in the long term, the home environment and a bad schooling environment for early elementary are going to take over. Some then focus on fixing the schools. Yes, but there are other lessons there, including earlier intervention and a focus on parenting education.

Early Head Start does appear to have some advantages, albeit modest. At the age of three, the program benefited families across child, parent, and family self-sufficiency outcomes. However, ,programs that fully implemented the Head Start performance standards and mixed-appprach programs (home- and center-based services) had broader and stronger impacts. Impacts also varied by sub-groups (race, length of time in program, and risk factors). In the prekindergarten years, students who continued with formal early childhood education exhibited an impact. A grade 5 followup (with a smaller sample size), however, experienced the same ‘fade out’ effect on the full group, but gains persisted in some subgroups. Thus, although short-term gains are stronger with earlier intervention, long-term gains remain problematic.

As a corollary to the failing schools argument: fourth, although the study showed lowered indications of math ability for the Head Start children in kindergarten, that indicator was based only on teacher reports of children’s performance and not supported by scores on three direct math assessments.  And, this is where the actual school attended for early elementary may matter – the control group children attended school reporting a higher percentage of students at or above a proficient level of math than the Head Start children’s schools.

However, the argument that it is the schools that are failing do not hold as much water in most cases. Perhaps the schools are erasing any and all cognitive gains, thereby re-leveling the playing field, but with small exceptions such as the math one above, one cannot argue that the control group children attended better schools and the Head Start children did not – these children were pulled in general from the same sociodemographic group.

It is important to understand that although failing schools is a policy concern, most gains in preschool and kindergarten education are erased by third grade. This is true with any type of preschooling – regardless of socioeconomic status – and it is true in the case of redshirting kindergartners.  I address this in my articles on whether preschool matters and on redshirting, and the Director of the National Institute for Early Education Research Stephen Barnett addresses this in his apologetic paper on Head Start where he claims right off the bat: “Head Start is Effective“. Although I do not make such a bold statement, we are in agreement over some of the problems: measures of improvement are flawed, and Head Start implementation is spotty.

In addition, as Barnett argues,  we have to look at fifth: Benefits of preschool are not necessarily measurable by test scores. Research has shown an increase in high school graduation rates, future economic earnings, and lower crime rates in participants who attended preschool.  Some of those studies are based on Head Start participants, but many are from small sample sizes with intensive programs such as the Perry Preschool Project and the Abcedarian experiment. As Megan McArdle points out, if you look at it from a market standpoint, yes, the return on investment for those programs is huge. However, also from a market standpoint, scaling up such programs leads to spending exorbitant amounts of money. Quality will get sacrificed for quantity. (See: Head Start) And, the small sample sizes of those two preschool darlings mean the results are more likely to be happenstance than statistically significant.

Sixth:  Another nod to Barnett’s article.  The research may be flawed here. In discussions with local educators and literacy professors, I’ve learned that most pre-reading tests do not fully embrace the five pillars of literacy, most importantly, comprehension.  Comprehension also gets shortchanged in most ‘teach-to-the-test’ instruction in the early grades. Hence, the third and fourth grade drop off in literacy – when, suddenly, reading is more than just stringing words together.

Seventh The study could not accurately separate the effects of Head Start versus familial care versus other other early education programs versus substandard childcare. We know that substandard care has a negative effect on children, while quality care has a positive effect on children when coming from bad home environments. The control group children did not stay exclusively in parental or familial care.  In fact, the control group were in non-parental care for more hours per week, on average.  The results of the impact study do not augur well for Head Start writ large as a program, but it would be incorrect to conclude that early childhood programs have no impact for disadvantaged children, as many in the control group also attended early childhood programs. Instead, the results indicate Head Start does not have a comparative advantage in early childhood education. That makes sense, given the haphazard way in which it is implemented (and, as small government advocates would likely point out, given that federal government does not by any means have an advantage in effective programs).

Therefore, what if Head Start and universal prekindergarten programs were implemented via best practices models?

We will explore that next, in looking at models from Oklahoma, Georgia, and Florida.

After that, policy proscriptions that come out of all this research, based on small government and market-driven principles.


4 thoughts on “Controversies: Does Head Start Work?

  1. Pingback: Is Tulsa a Beacon of Excellence in the Early Education Landscape? | Midtown Tulsa Parents

  2. Pingback: Universal Pre-Kindergarten: Evidence from the Field | Six Forty Nine

  3. Pingback: Welcome to the Carnival of Evidence-Based Parenting! First Edition: Preschool | Science of Mom

  4. Pingback: Why Johnny is Broke: Marrying Market-Driven Principles with Government Excess in Early Education | Six Forty Nine

Your thoughts?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s