Excitement over new early childhood education initiatives is building at the State Capitol. No wonder. After years of frustration and failure in trying to shrink our state's yawning K-12 racial and income achievement gap, there's hope that we may have found a new, and simpler, route to success.

On one level, the new approach has much to recommend it. Advocates support state-funded scholarships that would give low-income parents both access to high-quality preschool programs and the information they need to choose them. More parental choice, more information, more accountability -- that's a model we should be expanding across the entire education spectrum.

But I worry that, in our search for a silver bullet, we may have oversold the likely benefits of early childhood education. Some advocates, for example, assure us that every dollar we spend on high-quality preschool programs for poor children will save a whopping $16 down the road through higher graduation rates, better wages and lower incarceration rates.

Minnesotans are likely to be disappointed if they expect a payoff anything like this.

Research suggests that low-income students often make cognitive gains, generally modest, in high-quality preschool programs. The challenge is to sustain these gains beyond early elementary school.

Unfortunately, few studies show lasting gains, and even fewer show lasting gap reductions, according to Chester Finn, an expert in education policy and author of "Reroute the Preschool Juggernaut." So where do claims about a 16-to-1 return come from? From studies of a handful of "hothouse" pre-K programs from the 1960s and '70s -- primarily the Perry Preschool project in Michigan and the Abecedarian project in North Carolina.

These were, in Finn's words, "truly exceptional," "ultra-intensive" programs that differed greatly from the ones most low-income children in Minnesota would attend today. Both involved fewer than 200 children and cost about four times as much per child as the scholarships being proposed here (at least $15,000 vs. $4,000). Neither has been duplicated so results can be tested. In fact, with their multifaceted components, they would be almost impossible to bring to scale.

Moreover, there is serious debate about whether they actually produced the long-term impact that some studies have found. In a 2005 appraisal of Perry data, for example, analysts found that "exposure to Perry explains less than 3 percent of all the variation in [participants'] earnings ... and 4 percent of the variability in school attainment levels." Abecedarian -- a highly intensive birth-to-kindergarten program that delivered 13 times as many "preschool intervention service" hours per child as Head Start does -- produced lasting cognitive gains, but not statistically significant differences in graduation rates, employment or criminal activity.

Head Start's track record illustrates the difficulty of producing enduring gains for poor children. It costs about $8,000 per student per year. Yet its participants' cognitive and socio-emotional gains -- minimal to begin with -- fade completely by third grade and yield no lasting benefit, according to federal studies.

Why are low-income youngsters' gains so hard to sustain? Over time, says Finn, these students' chaotic homes and neighborhoods -- and the low-quality schools they generally attend -- tend to swamp initial benefits. Further, our K-12 schools are not structured to build on whatever gains children bring with them. Since it's so hard to differentiate instruction, teachers often find themselves compelled to teach to the lowest common denominator.

Similar challenges afflict all-day kindergarten -- the subject of another big push at the Legislature. In 2001, about 60 percent of all U.S. kindergartners were in full-day programs. A 2005 RAND study found that while such programs produced greater achievement gains during the year itself, there were no significant differences in academic performance by the end of third grade.

Can we design preschool programs that will reliably raise long-term academic performance? Finn sees value -- for the neediest children and coupled with the right K-12 reforms -- in intensive, heavily cognitive and rigorously evaluated programs, starting perhaps before birth, and enlisting and assisting the parents "from day one." He adds that a makeover of compulsory K-3 grades to stress early literacy and vocabulary, and to dramatically increase instruction time for disadvantaged children, might benefit poor children far more than adding voluntary preschool programs.

In the end, there's only one reliable way to shrink the learning gap. That's to ensure -- somehow -- that far more low-income boys and girls come from homes with two dedicated parents who give them the love and support they need to succeed in school. Without a cultural sea change of that kind, our learning gap is likely to bedevil us for a long time to come.

---------------------------

Katherine Kersten is a senior fellow at the Center of the American Experiment. The views expressed here are her own. She is at kakersten@gmail.com.