Senate Committee on Agriculture, Nutrition, and Forestry
March 12, 1998
 

 Thank you for the invitation to testify today.  I am Robert Greenstein, executive director of the Center on Budget and Policy Priorities, a nonprofit organization that studies fiscal policy issues and social programs at both the federal and state levels, with emphasis on issues affecting low- and moderate-income families.  The Center does not represent interests or constituents.  It is a public policy organization funded almost entirely by foundations.

 In the late 1970s, I served as Administrator of the Food and Nutrition Service in the U.S. Department of Agriculture, while in 1994, I served as a member of the Bipartisan Commission on Entitlement and Tax Reform chaired by Senator Bob Kerrey and former Senator Jack Danforth.  I have been involved with the child nutrition programs for about a quarter century.

The Challenge Facing the Committee — Crafting a Deficit-neutral Child Nutrition Bill

 It now seems clear the Committee will have to develop a child nutrition reauthorization bill that is deficit-neutral.  To be able to move a bill that increases child nutrition entitlement costs, the Committee would need to receive a grant of new entitlement authority through the Congressional budget resolution.  With the Budget Committees searching high and low for mandatory savings to offset increased ISTEA costs, there appear to be no prospects for the Agriculture Committee to receive an increased allocation enabling it to increase child nutrition entitlement spending.

 This does not mean, however, that no improvements can be made in child nutrition programs.  Some modest improvements should be possible; they can be paid for through other modest child nutrition changes that produce offsetting savings without harming either the programs or the children they serve.  Doing so, however, is not an easy task.  There are a number of proposals in circulation, and with only limited dollars likely to be available, these proposals essentially compete with each other.  The Committee will have to establish priorities among them.  In this testimony, I would like to discuss both what I regard the most important improvements to make and some options for offsetting savings to finance them.  I also would like to discuss several other proposals entailing increased child nutrition costs that I would caution against.

 Before doing so, I first would like briefly to review the current condition of some of the child nutrition programs and how they were affected by the 1996 welfare law.
The Current State of the Programs

 The child nutrition programs perform an outstanding service.  They improve the nutrition of participating children, including children who are in school, child care settings, and summer recreation programs.  Their most important function is the strong role they play in reducing hunger and undernutrition among low-income children.  Without the child nutrition programs, many poor children would not be adequately nourished and not have as strong a chance of reaching their full potential.

 Today, most of the child nutrition programs are healthy.  For example, although concerns have been voiced at points in the past about whether significant numbers of schools would drop out of the school lunch program, that is not occurring.  The number of schools in the program has risen a small amount nearly every year since 1984 and stood last year at about 95,000 schools.  While a few schools may leave the program here or there, the number of such schools is small and is more than outweighed by new schools entering the program.  (It also should be noted that the small number of schools that have departed contain few children eligible for free or reduced meals.)

 In addition, school lunch participation by children is rising.  In school year 1997, some 26.9 million children participated in the school lunch program on an average day.   That is 1.3 million more children than in school year 1994 and 2.4 million more than in school year 1991.  School lunch participation last year was the second highest on record.

 Particularly striking is the growth of the school breakfast program.  In 1977, some 20,000 schools served breakfast.  In 1987, some 37,000 schools did.  Last year, nearly 69,000 did.  During the past 10 years, the number of children receiving school breakfasts on an average day rose 96 percent, climbing from 3.6 million to 7.1 million.

 Figures are sometimes cited that can create a misimpression that the breakfast program is not doing well.  It is sometimes said that only 40 percent as many children receive free or reduced-price school breakfasts as receive free or reduced-price lunches.  This figure can create an impression that most low-income schoolchildren do not have access to school breakfast, but that is not the case.  A principal reason that the number of low-income children eating breakfast at school is only two-fifths the number of such children who eat school lunches is that large numbers of these children eat breakfast at home with their families.  That is an option not generally available for these children with regard to lunch.

 To determine the reach of the breakfast program today, the most relevant question is not what percentage of low-income children eat breakfast at school, but rather what proportion of the schools that offer the school lunch program also offer the breakfast program, and what proportion of low-income school children attend a school with a breakfast program.  The answers to these questions are very encouraging.
 Ten years ago, only about 40 percent of the schools with a lunch program also offered the breakfast program.  Today, 73 percent of schools do.  Moreover, since the breakfast program is much more heavily concentrated in low- and moderate-income areas than in affluent areas, it would appear that more than 80 percent of the low-income children attending a school with a lunch program are enrolled in a school that also offers breakfast.  And since the breakfast program is concentrated more heavily in elementary than secondary schools, it is likely that somewhere in the vicinity of 85 percent of low-income children enrolled in elementary school attend a school offering the breakfast program.  This is remarkable progress.

 Nor has the progress ended; the school food programs have continued to grow.  Even the Summer Food Service Program for Children has continued to expand, despite reductions in the meal reimbursement rates enacted under the welfare law and contrary to fears that those reimbursement rate reductions would cause summer food sponsors to leave the program in large numbers and program participation to deteriorate.

 The overall pattern in the nutrition programs is a striking one — food stamp participation has been falling sharply, declining substantially faster than can be explained by the economy and the new welfare law, while participation in most child nutrition programs has been rising.  (CBO devotes a full page in its recent report on the budget and the economy to a discussion of the fact that food stamp participation is falling by much more than it had forecast.)  Between December 1995 and December 1997, food stamp participation dropped by about 5.5 million people, or more than 20 percent.  In the last 15 months alone, participation has plummeted by 4.5 million people, including about two million children.  Census data show that this isn’t simply due to economic growth; between 1995 and 1996, the number of poor children fell by 200,000 as the economy improved, but the number of children receiving food stamps fell by 700,000, more than three times as much.  The proportion of poor children receiving food stamps now appears to be dropping significantly.

 The opposite pattern appears in the school food programs — participation continue to rise.  Table 1 and Figure 1, on pages ____ and ______ of this testimony, show the disparity between the food stamp and child nutrition trends.  Over the three years ending in September 1997, the number of people receiving food stamps plunged 23 percent.  By contrast, between school year 1993-1994 and school year 1996-1997, school lunch participation rose five percent, and school breakfast participation rose 21 percent.  In the summer feeding program, participation increased six percent over the past three years.

 The one child nutrition program where a different pattern is emerging is the program affected most heavily by the welfare law changes — the family day care home component of the Child and Adult Care Food Program.  Prior to the welfare law, this was the one child nutrition program that was not means-tested, and studies indicated that about 70 percent of the children served — and about 70 percent of the federal funds involved — went to serve children with family incomes above 185 percent of the poverty line.  The welfare law maintained the current meal reimbursement structure for meals served in family day care homes that either are located in a low-income area or operated by a low-income provider, but reduced the rates by about half for “non-low-income” homes.

 This change did not take effect until July 1, 1997, and not many data are yet available on its impact.  Preliminary data indicate that between September 1996 and September 1997, the number of homes in the program dropped 7.5 percent, while the number of children attending homes participating in the program fell two percent.  (It is unclear why these percentages differ.)

 There are many things we don’t know yet about what is happening in the family day care program.  Has the decline that has occurred to date in the number of participating homes been primarily among homes serving middle-income children or has it affected a significant number of low-income children as well?  Is the 7.5 percent figure likely to be the full measure of the decline in the number of participating homes or will that number continue to grow?  As part of the welfare law, Congress wisely mandated a study of the effects of the sweeping changes in this program.  When the results become available, they may indicate a need for mid-course corrections.  Such results will not be available in time for this year’s reauthorization.

Issues Related to Reauthorization

 The committee is likely to have limited funds available for child nutrition improvements since improvements need to be financed with offsetting child nutrition savings measures.  Among the proposals that have been advanced which would entail new cost are proposals to make the school breakfast program universal in elementary schools, to reinstate funds for child nutrition outreach grants, to restore summer feeding meal reimbursement rates, and to expand after-school food assistance programs.  As noted, with limited resources available, these proposals essentially compete with each other.  Choices will need to be made, and priorities set, among them.

 In my view, the proposals relating to after-school programs have the most merit.  They are where I would recommend putting the resources.  I believe the other proposals should be accorded much lower priority.  I also believe some of the proposals are unwise.  Before discussing issues relating to the after-school programs, I would first like to share with you my concerns about some of the other proposals, especially the proposal to make the school breakfast program free for elementary school children at all income levels.
Universal Breakfast Proposal

 As someone who has been involved in issues relating to the school breakfast program for more than 20 years — especially issues related to program expansion — I am particularly interested in the universal free breakfast proposal.  This proposal is in the longstanding tradition of efforts to expand and strengthen the program.  But the conclusion I reach is that despite its excellent intentions, this proposal would not represent a wise use of limited funds.  Both inside and outside the child nutrition area, there are other initiatives to which I would accord much higher priority.  The reasons I would recommend against this proposal include the following:

 • As noted, the school breakfast program is healthy and has been growing rapidly under current law.  It does not need a major infusion of new federal money to expand.  The number of schools in the program has skyrocketed from 20,000 in 1977 to 37,000 in 1987 to 69,000 in 1997.  Most schools in low-income areas are already in the program, especially elementary schools.  It appears that 80 percent to 90 percent of all low-income elementary school children already attend a school with a breakfast program.

 • School breakfasts already are free to all children in these schools with family incomes up to 130 percent of the poverty line ($20,865 for a family of four) and is available for a nominal price (schools may charge no more than 30 cents) for children with incomes between 130 percent and 185 percent of the poverty line, now about $30,000 for a family of four.  In addition, there already is an option in law under which schools that serve mostly low-income children can offer breakfasts free to all children, and schools are beginning to make increased use of this option.

 • Consequently, the proposal to make school breakfasts free to children at all income levels in all schools would primarily subsidize middle-income children who do not need free breakfasts.  These middle-income children consist principally of two groups.  One is children who already eat a school breakfast and pay the regular charge; as a result, large amounts of new federal cost would be incurred in “buying out the base” (i.e., in providing a deeper federal subsidy for the school breakfasts these children already are eating).  The other group of middle-income children whose meals would be completely subsidized are children who now eat breakfast at home with their families but would begin eating at school instead if school breakfasts were free.  In my view, “buying out the base” and inducing some middle-income children to eat at school rather than at home do not constitute social advances sufficiently important as to justify a claim of several billion dollars over five years when there are more pressing needs to be met in addressing poverty and hunger.  I believe that the limited resources that may be available for child nutrition improvements should be targeted primarily on the children who need assistance most, those with low incomes.

 • Another argument sometimes made by proponents of the proposal is that it will result in many fewer children missing a morning breakfast.  Large research evaluations funded by USDA in the early 1990s, however, do not support this contention.  The studies show that 94 percent of children in kindergarten through third grade already eat breakfast and that the presence of the school breakfast program does not increase this number.  When a breakfast program is established in a school, the studies found, the proportion of children eating at home falls by essentially the same amount that the proportion eating at school rises.  The studies did find that school breakfasts are nutritionally superior, on average, than other breakfasts, a point that should not be minimized.  But the claim that making breakfasts universally free will close a meal gap among large numbers of children who miss breakfast altogether is not borne out by the available evidence.

 • A final argument is that a study in Minnesota and another by Harvard researchers have found there to be significant educational benefits from providing universal free school breakfasts.  This argument is not strong.  The Minnesota study does not appear to be a very rigorous study; conducted in a small number of schools in Minnesota, it consists largely of the impressions of teachers, parents, and children who were interviewed.  It does not contain data produced under the types of methodologies that rigorous evaluations employ. The material I have seen from the “Harvard study” is not actually a study; rather, it is a series of overheads produced for USDA in November by two Harvard Medical School faculty members (and the work for which was financed by the Kellogg cereal company and the Mid-Atlantic Milk Marketing Association).  The overheads present data from several schools where breakfasts were provided free to all children, but one cannot determine from the overheads how some of the data were collected or how rigorous the methodologies were.  (Moreover, the text of the overheads appears to recommend greater use of the option already available under current law through which schools with mostly poor children can serve breakfasts free to all children, an approach I strongly support.  That is different from making school breakfast available for free to children at all income levels and in schools in all areas, including affluent ones.)

 I would recommend two courses of action here.  First, greater effort should be devoted to promoting to schools, and encouraging and assisting schools in adopting and implementing the provision already in law that enables schools with a high percentage of low-income children to offer both breakfasts and lunches free to all children.  (Under this provision of law, a school receives federal reimbursement based on the proportion of school meals served free, reduced-price, and paid before this options is implemented.  If a high percentage of meals are served free or at a reduced-price, the school is able to cover the added cost of serving all meals free through offsetting administrative savings.)

 Specifically, I would recommend inclusion in this year’s child nutrition legislation of grants to states of $1 million to $2 million a year for states to promote and provide assistance to schools in adopting this option and also in adopting another option provided under the National School Lunch Act, the “direct certification” option. Under that option, schools may streamline the process of certifying poor schoolchildren for free meals by matching school enrollment to names of children receiving food stamps or TANF cash assistance.  Where it has been used, this option has been successful both in increasing the proportion of children eligible for free lunches and breakfasts that actually are certified for free meals and in reducing schools’ administrative costs.

 Providing these small grants to expand use of these two options should help bring free breakfasts to more low-income children and do so in a way that is much better targeted and far less costly than providing breakfasts free at all income levels in all locations.  Greater use of these options can help more children receive free school lunches as well.

 Second, the Committee might work to consider authorizing a universal free breakfast pilot project, with a rigorous evaluation to test the claims that universal free breakfast will significantly improve school performance.  I mention the possibility of such a pilot with some hesitation.  On some occasions in the past, a pilot project label has been used as a cover for starting a project in many more locations than needed for a pilot.  The history of such endeavors is that once these projects are established in a number of areas, they never end, even if the evaluation does not find especially positive results.  The so-called pilots get extended each time the programs come up for reauthorization, and the Committee must make cuts elsewhere to finance the continuation of projects in “pilot areas” even though Congress has long since determined not to turn these projects into national policy.

 Such a development ought not be allowed to happen here.  Any such pilot should be limited in funding and scope to what is needed to conduct an evaluation of this approach.  Given the difficulty the Committee is likely to face in ending universal free meals in any pilot area after the pilot ends, the pilot should be no larger than is needed to test and evaluate this idea.  While I do not know precisely how much such a pilot would cost, past FNS evaluation efforts in the child nutrition area suggest that no more than $5 million over several years should be needed.  If proposals emerge for a $25 million, $50 million, or $100 million pilot test of universal free breakfasts, they should be seen not as pilot test proposals but as efforts to take the first steps toward implementation of this proposal even before testing it.  Such proposals could ultimately harm poor children; if the Committee must find another $25 million, $50 million, or $100 million to continue universal free breakfasts in the pilot areas each time child nutrition programs are reauthorized, the Committee could find itself forced to squeeze nutrition programs primarily aiding poor children in future years to maintain free breakfasts for children in relatively affluent areas long after the pilot test and accompanying evaluation have been completed.
 

Raising Summer Feeding Reimbursement Rates

 Another proposal that I would recommend against pursuing would be to reinstate in whole or in part the welfare law’s reduction in reimbursement rates for meals served in the Summer Food Service Program for Children.  We now have had a year of experience with the lower reimbursements rates, and this experience indicates that raising the rates back up should not be a priority use of limited child nutrition resources.

 There was concern at the time of the welfare law’s passage that the reduced rates might induce many sponsors to drop out of the summer program.  Not only did that not occur last summer, but the program grew, expanding at about the same rate as it did the previous summer when the reimbursement rates were higher.  The number of children served, the number of meals served, the number of participating sponsors, and the number of participating sites all increased.  I would note that while the welfare law reduced the meal reimbursement rates, the rates remain somewhat above the rates provided for free meals in the school lunch program.

 Given the limited resources the Committee is likely to have available, raising summer food reimbursement rates ought not rank as a priority.  There may, however, be a few small and much less expensive steps the Committee could take (such as modest changes in the site limitations imposed on nonprofit organizations serving as summer food sponsors) that could address some current difficulties with bringing more sites and children into the program.

Outreach Funding

 Another proposal that has been mentioned is to reinstate the approximately $5 million a year in school breakfast outreach grants to states that the welfare law ended.  These outreach funds were instituted in 1989 and appear to have helped spur some of the recent growth in the number of schools in the breakfast program.  But there is significant question as to whether these funds still are needed very much.  They may have largely accomplished their intended purpose.

 As noted earlier, the school breakfast program has grown tremendously over the past two decades.  Most schools in low-income areas now operate a program.  It is not clear that further breakfast outreach funds are needed.

 The doubts about the need for further outreach funds are strengthened by what occurred in FY 1996, the last year these grants were provided.  Some of the breakfast outreach funds went unused that year, because states could not find sufficient qualifying schools that were not already in the breakfast program and that would use the funds to help them enter the program.

 The doubts about a continued strong need for these funds led to the inclusion of a provision ending these funds in most of the major welfare proposals in 1996.  Provisions ending these funds were included in the Clinton budget that year, the Daschle welfare substitute on the Senate floor, and the Castle-Tanner bipartisan welfare bill in the House, as well as in the Republican welfare bill that ultimately became law.

 With the need for school breakfast outreach funds beginning to wane, the 1994 child nutrition reauthorization broadened the allowable use of these funds to include summer feeding outreach.  We do not know whether summer food outreach grants were effective.  So little of this money was used for summer food outreach that it is difficult to tell.  It may be noted, however, that the summer food program expanded as much in the summer of 1997 — when no outreach grant money was available — as in the summer of 1996 when a small amount of the money was used for this purpose.

 All in all, this does not add up to a persuasive case for making reinstatement of the outreach grants a priority use of funds.  I would recommend according higher priority to other uses, such as the institution of grants to states to promote the free meal and direct certification options, as discussed above.
 

After-school Programs

 In 1996, the nation adopted legislation making a profound transformation in the nation’s welfare system.  Strict time limits were applied to welfare receipt, as well as tough work requirements, backed up in many states by sanctions that terminate assistance not only to non-compliant parents but also to their children.

 While there has been a substantial decline in welfare caseloads, the toughest challenges for states in implementing welfare reform lie ahead.  The families that have left the welfare rolls appear, by and large, to be the more tractable cases.  As a result, while caseloads are smaller, the proportion of the caseload that consists of long-term cases with more serious barriers to employment appears to have risen.

 There are only a few states where the time limits have yet taken effect.  But they soon will.  As growing numbers of the more difficult cases approach the time limits, the imperative to move them into employment will become stronger.

 Adequate care during after school hours for school-age is important in this regard.  This is underscored by the findings from a project in Chicago known as the Gautreaux project that is well known among poverty researchers.

 The Gautreaux project began in 1976 as a result of a court desegregation order related to housing.  Under this project, several hundred families living in Chicago housing projects are moved each year, with some moving to apartments in the suburbs and some to apartments largely in low-income areas of the city.  Researchers from Northwestern University have been studying these families for years.  The researchers found that the families that moved to the suburbs and those that moved within the city matched each other very closely before the moves.  As a result, by following these two groups of families, the researchers have essentially been able to monitor what is essentially a “natural experiment” on the differential effects of enabling poor single-parent families to move to the suburbs as compared to these families remaining in low-income parts of the city.

 In recent years, the researchers have published a number of articles on their findings in academic research journals.  They found that the families that moved to the suburbs fared significantly better on a number of dimensions.  One is that the single female parents who moved to the suburbs became employed at higher rates than those who remained in poor areas of the city.

 When the researchers conducted follow-up interviews with these women to learn the reasons for these differences in employment, one of the things they learned was that a number of the mothers with school-age children who lived in the poor inner city neighborhoods would not take jobs that did not enable them to be home by 3 pm when school let out.  These mothers expressed strong fear about crime, violence, gangs, drugs, and the dangers or allures of the street and said they had to be home when school let out to keep their children safe and out of trouble.  Needless to say, this adversely affected these mothers’ employment prospects.

 The need for after-school programs, especially in poor areas, is becoming increasingly understood.  This is true not just for elementary school children but for teenagers as well.  In middle-class neighborhoods, there often are various types of activities available after school, and the dangers of street violence and gangs are low.  In impoverished areas, by contrast, the number of such programs and activities often are fewer, especially in schools, while the dangers of the streets are much greater.

 The lack of after-school activities in poor areas can have deleterious consequences.  Crime statistics indicate that the peak hours for teenage crime are not late at night, but between 3 pm and 8 pm.

 To assemble the resources needed to create and maintain after-school programs, schools, community groups, and churches often need to combine resources from several sources.  One source can be the Child and Adult Care Food Program.  The CACFP plays a role here for another reason as well — children are not likely to find attractive an after-school program that fails to provide an afternoon meal supplement.  Providing nutrition assistance to after-school programs through the Child and Adult Care Food Program thus can be thought of as part of a larger national effort to provide adequate child care assistance, including adequate after-school care, to children from low- and moderate-income working families.

 There are, however, two barriers to CACFP support for such programs.  These barriers could be addressed in reauthorization legislation.  The two barriers are the CACFP age restrictions and the CACFP application process for after-school programs.

The Age Restriction

 The Child and Adult Care Food Program provides reimbursement for meals served in after-school programs to children through age 12.  In non-low-income areas, that appears sufficient, as teenage children are in much less danger of gangs and street violence.  But in low-income areas, there is a need for strong, viable programs that will attract teens after school and keep them off the streets — so that they can be safe and can be kept away from crime and drugs and so that their parents do not feel they can’t accept jobs that do not allow them to be home by 3 pm.  Accordingly, I would recommend that CACFP support meals served to children up to age 18 in approved after-school programs in low-income areas.  This should not have an especially large cost.

 I would note that it is important for this provision to cover after-school programs located outside schools as well as in schools.  If a local YMCA, YWCA, boys’ or girls’ club, community center, or a church in a poor area offers such an after-school program, it should qualify for meal reimbursements just as an after-school program in a school should.  This may seem obvious, but I raise this point because I have seen a proposal in recent weeks that would limit the increase in age to programs run in schools.  In fact, the proposal I have seen would deny these meal reimbursements to after-school programs run by community institutions and churches in low-income areas — where the need for subsidized after-school programs for teens is substantial — while providing such meal reimbursements for after-school programs for teens in middle-class and even affluent areas so long as the programs in these areas are run in schools.  Given that suburban high schools already operate all sorts of extracurricular activities after school, it is not clear why there is a need to provide federal funds to support meals served to teenagers after school in these locations.

The Application Process

 A second issue involves the mechanisms through which schools can receive meal reimbursements for meals served in after-school programs.  Schools are entitled to receive such reimbursements if they apply for them; for most schools, this entails submitting an application to the Child and Adult Care Food Program, which in some states is run by the same agency that operates the school lunch program and in other states is operated by a different state agency.

 It would be easier for schools if they could apply through the school food programs for meal reimbursement for meals provided in after-school programs, rather than applying (and submitting reimbursement claims) through CACFP.  If the application and reimbursement processes were streamlined in this fashion, it is likely that somewhat more schools would provide these meals.

 A key question here is to what extent streamlining this process would cause more schools to provide after-school care programs and thereby increase the availability of after-school care.  This is a difficult question to answer; there is little hard evidence on it to examine.  Some who are familiar with these issues believe that streamlining this process would encourage more schools to institute or expand after-school programs.

 Given the need for more after-school care, especially in low-income areas, this would seem a good step to take.  This proposal would be more expensive than the proposal just mentioned to raise to 18 the age for after-school programs in low-income areas.  Unlike the proposal to raise the age to 18, the proposal to streamline the application and reimbursement processes need not be limited to low-income areas; there is a logic in allowing schools in all areas to apply for after-school meal reimbursements through the school food programs.  On the other hand, the need for more after-school programs — and hence for this change — is clearly greatest in low-income areas.  If the Committee cannot afford to extend this provision to all schools, it should enable schools in low-income areas to use it.

 I would note that the targeting of provisions in the National School Lunch Act to low-income areas has a long history.  The Summer Food Service Program for Children has been limited to such areas since its establishment in 1968.  The 1996 welfare law also established a meal reimbursement category in the family day care home portion of the Child and Adult Care Food Program that consists primarily of homes located in low-income areas.

 A final issue in the CACFP is whether to reinstated the so-called “fourth meal.”  Prior to the welfare law, child care centers (although not homes) could receive federal reimbursement for four meals a day served to children maintained in care for more than eight hours a day.  The welfare law reduced this to three meals.

 Program rules have gone back and forth on this issue.  Until 1981, the fourth meal was reimbursed.  Congress halted the fourth-meal reimbursement in 1981, reinstated it with some limitations in 1988, and eliminated it again in 1996.  There isn’t much evidence on the effects one way or the other.  During the 1981-1988 period, this was not an issue that provoked loud complaint.

 On the one hand, if resources permit, it would be nice to reinstate reimbursement for the additional meal; most of the children involved have low incomes and they are in care for long periods during the day.  On the other hand, there is not much indication that this part of the welfare law is causing serious problems for which there is a pressing need for a remedy.  It is possible that in a number of cases, the fourth meal is still being served.  I would accord the after-school proposals significantly higher priority than this proposal.
 

A Few Very Low-cost but Desirable Steps

 There are a few other extremely cheap, but still significant steps I would urge the Committee to take.  The homeless pre-schoolers food assistance program — a program that is operated in a limited number of locations and subsidizes meals for homeless pre-school children in shelters — will expire if not renewed.  Renewal would cost only about $1 million to $2 million a year.  These young children are the poorest of the poor — they don’t even have a home.  This program should definitely be extended.

 I also would urge the Committee to invest a very small sum — perhaps $2 million on a one-time basis — in developing a better system for family day care sponsors to get timely, accurate information on whether day homes are located in low-income areas.  The ability of states and sponsors to administer the sweeping changes the welfare law made in the family day care program are dependent upon the availability of this information.  An improved system of providing this information to day care sponsors is needed.
 
 
 

Financing Child Nutrition Improvements

 How could modest child nutrition improvements be financed?  I would like to discuss two possible offsets.  I would strongly recommend the first of these offsets, and recommend for consideration, but less strongly, the second.

 A. Fixing a Costly Anomaly Enacted in 1994

 In 1994, a provision was added to the child nutrition reauthorization bill at the request of a few House members that was thought to be innocuous and have no cost.  (CBO and USDA both scored it as having no cost.)  But the provision has turned out to have the unintended effect of increasing school food commodity entitlement costs to higher levels than Congress intended.  Addressing this matter and dropping the unintended subsidies would save about $200 million over five years.

 The National School Lunch Act entitles schools to a specific amount of commodities for each lunch served, and this amount is indexed annually for inflation.  The provision added in 1994 has had the unintended effect of slightly increasing the commodity entitlement above the longstanding — and quite adequate — entitlement level the law has long contained.  Modifying the 1994 provision so it does not have this unintended effect would leave in place the longstanding commodity entitlement level and full, annual indexation.

 I would note that the 1995 Senate reconciliation bill, the 1995 Senate welfare bill, the FY 1997 Clinton budget, and the bipartisan Castle-Tanner and Chafee-Breaux bills in 1996 all contained a provision to avoid these unintended costs.  In addition, in June 1996, the American School Food Service Association, Food Research and Action Center, and Center on Budget and Policy Priorities jointly recommended including a provision to avoid these unintended costs in the welfare bill and using the savings to lessen other child nutrition reductions in the bill.  (This offset is discussed in more detail in the box on page ____.)

 B. Rounding Down Reimbursement Rates

 Nearly all federal entitlement benefits are rounded down to the nearest whole dollar.  This is true for popular programs from Social Security to veterans compensation to SSI.  It also is true for food stamps.

 The child nutrition programs are something of an anomaly here.  The reimbursement rates for meals served to non-low-income children are rounded down to the nearest whole cent, as a result of a provision in the welfare law.  But reimbursement rates for free and reduced-price meals continue to be rounded to the nearest quarter-cent.  This is probably not necessary.
 It also adds an element of complexity.  There no longer is a single “section 4" cash reimbursement rate covering all lunches — free, reduced-price, and paid alike.  Because one set of rounding rules is used for the section 4 rate for paid meals and a different set of rounding rules is used for the section 4 for fee and reduced-price meals, there now are two slightly different section 4 rates in most years.

 Rounding down all child nutrition reimbursement rates to the nearest whole cent would produce some modest savings.

Achieving Balance Between Offsets and Improvements

 Finally, I would urge the Committee to make sure that the combined effect of any offsets and expansions it considers is not to take money from poor schools and school districts and shift it to more affluent ones.  Middle- and upper-middle-class children do not suffer from hunger or impoverishment.  I would hope the Committee can fashion a package of improvements and offsets that strengthens the child nutrition programs overall.  But a bill that shifted resources from poor schools and poor areas to more affluent ones would, in my view, be unwise and make less sense than a straight reauthorization without changes.