There are so many ways to measure poverty that even a school district superintendent can get a little confused about how many of his students are officially considered "poor". In a February 2013 column in the SF Examiner entitled "Equity just one of SFUSD's key goals", San Francisco schools' Superintendent Richard Carranza stated that "almost half the children enrolled in our schools live in poverty." This raised some eyebrows among parents, because it is pretty widely known that SFUSD's free and reduced price lunch rate, the official government measure of student poverty, is in reality over 61% and has been at or close to that level for years.

It's clear that poverty is at the root of many problems students have at school, everything from low income students starting school with smaller vocabularies than their wealthier peers, to low income students being more likely to suffer from malnutrition, asthma, painful tooth decay, and other poverty-related disorders that affect the ability to focus and learn.

Less clear is where to set the boundary between "in poverty" and "not in poverty", and whether those living just across the line from poverty are free of its ills.

The federal government uses several different methods to measure poverty in the general population, including "poverty thresholds" and "poverty guidelines." The "thresholds" are the original version of the federal poverty measure, are updated annually by the Census Bureau, and are used for statistical purposes, such as estimating the number of Americans in poverty. The "guidelines" are issued annually by the Department of Health and Human Services, and are used for administrative purposes, such as determining qualification for federal assistance programs, like SNAP (formerly called food stamps), certain parts of Medicaid, and government paid school meals.

The original poverty measure (thresholds) was developed in the early 1960s, used a family's before-tax income, and represented three times the cost of the family's food budget based on US Department of Agriculture food plans. At the time, the typical family spent one third of their income on food. This is still the method used by the Census Bureau for calculating the "official" poverty threshold, and official estimates of the number of Americans living in poverty.

Decades later, there was growing concern that the original 1960s measure did not reflect government policies like payroll taxes which reduce disposable income, nor food stamps or other feeding assistance which increase it. Also missing were the expenses of holding a job, such as transportation and childcare; the rising cost of health care; the effects of social changes such as child support payments or unmarried couples cohabitating; and geographic differences (ie - higher cost of living in some areas.)

A National Academy of Sciences panel suggested in its 1995 report Measuring Poverty: A New Approach, that a new measure should be developed that would more closely approximate the actual amount of money available to a family, which would reflect the changes brought about by these government policies and social changes. Years later, these new experimental thresholds, which are considered to be "a work in progress", are called the Supplemental Poverty Measure (SPM). By 2012, the SPM was refined to reflect higher rates in states like California with higher living costs, or less generous government support, rather than one uniform rate across the country.

How do these three federal poverty measures - the "thresholds", the "guidelines", and the SPM - compare? Let's look at how they measured poverty in 2011for a hypothetical San Francisco family of 4 comprised of two adults, a 4 year old, and an 8 year old. We'll call them the Lee family, and assume that both adults worked a 40 hour week at SF's minimum wage, which in 2011 was $9.92/hour, giving them income of $41,267 that year.

For 2011, the federal poverty threshold for this family was $22,811, and would have been the same no matter where in the country they resided. The Lee family was not in poverty in 2011 based on this measure.

That same year, the federal poverty guideline for the Lee family was $22,350 (for California and the other 47 contiguous states; higher for Alaska and Hawaii.) The family was also not in poverty in 2011 according to this measure.

The SPM for this family (assuming they are renters) for 2011 was $25,222, no matter where they lived. Again, not in poverty by this measure.

But the federal government isn't alone in trying to quantify poverty. In the mid 1990's, another measure, called The Self-Sufficiency Standard, was developed by Dr. Diana Pearce, then Director of the Women and Poverty Project at Wider Opportunities for Women. According to the Center for Women's Welfare, the Standard "defines the amount of income necessary to meet basic needs, including taxes, without public subsidies (e.g., public housing, food stamps, Medicaid or child care) and without private/informal assistance (e.g., free babysitting by a relative or friend, food provided by churches or local food banks, or shared housing)."

In this way, the Standard is more like the SPM and addresses many of the same issues. However, while the SPM remains more of a blunt instrument, identifying the state as a whole as having a higher rate, the California Self-Sufficiency Standard is able to calculate a poverty level for various household configurations by county, even adjusting for different age groups for the children.

As measured by the California Self-Sufficiency Standard, our SF family (2 adults, a preschooler, and a school age child) would have needed income of $76,352 in 2011 to meet their basic living expenses without relying on government or private support. That's more than three times the income which both the federal poverty threshold and the poverty guidelines, and even the SPM, use to quantify poverty for this family. Using this measure, the Lee family income of $41,267 was well below the poverty line in 2011.

Confused yet? But wait - there's more.

Government measures of poverty in schools are different than the measures used for the general population. There are two ways in which the government designates student as
"socioeconomically disadvantaged", a fancy term for "poor." The more obscure way relies on data about parent education level reported during standardized testing by children as young as 7. The Lee family's 8 year old had no idea what his parents' education level is, so he just guessed when he bubbled in his response on the student answer document; needless to say, this data is unreliable and seldom used for anything.

By far the most commonly used measure of poverty in schools is eligibility for government paid (free or reduced price) school meals. Districts collect information to calculate school and district wide percentages for "free and reduced" each fall. These figures are used, among other purposes, for determining eligibility for Elementary and Secondary Education Act (ESEA) programs, including Title 1, which provides extra federal money to schools with high percentages of low income students.

The cutoff for free school meals is income of 130% of the poverty guideline, or $29,055 for a family of 4 in 2011-12; for reduced price meals, it is 185% of the poverty guideline, or $41,348 for a family of 4 that year. With their $41,267 income, the Lee family just barely qualified for reduced price school meals, and the children counted as "socioeconomically disadvantaged."

In 2011-12, there were 26,766 SFUSD students qualified for free lunch, and 5555 qualified for reduced price lunch. A total of 32,321 students, or about 61.1% of SFUSD's official enrollment for that year, came from families with income at or below a level comparable to $41,348 for a family of 4 with 2 kids, at a time when the self-sufficiency standard indicated that income for that family would need to be $76,352 (about $35,000 higher) just to make ends meet.

Are students suddenly freed from conditions associated with poverty - food insecurity, inadequate medical care, substandard housing - just because their parents earn a few dollars more than a theoretical "poverty line"? Unlikely. Does that 61% figure for how many SFUSD students are "socioeconomically disadvantaged" really include every student impacted by poverty? Probably not. So how many SFUSD students were living in poverty that year? All we know for sure is that it was at least 61%; it might have been more, but it certainly wasn't less.

Dana Woldow has been a school food advocate since 2002 and shares what she has learned at Follow her on Twitter @nestwife.