Defining and Measuring Poverty
Before moving on to look at policy development, it is important to establish some basic metrics to judge the administration’s anti-poverty efforts. One crude measure would simply be to ask the following: Were Americans living in poverty raised out of it by the administration’s efforts? This, however, would be a deeply flawed approach. First, it assumes an uncomplicated definition of poverty when this is in fact a problematic concept.12 Second, even when there is a satisfactory working definition of poverty it is too simple to compare the poverty rate at two fixed points in time as this ignores the impact of broad economic forces. And very clearly the Obama administration inherited a worsening economic situation that was inevitably going to lead to an increase in the poverty rate, even if existing anti-poverty policy instruments were fully maintained. Third, looking only at the numbers of people above and below the headline poverty rate, however that is defined, misses changes among those with household incomes consistently below the poverty line.13 Even if people remain poor, the depth and intensity of that poverty can fluctuate according to policy paths either chosen or neglected, and any study of anti-poverty strategy needs to reflect on the circumstances of those in so-called deep poverty.
The most established data set for counting the number of poor in the USA is the Official Poverty Measure (OPM). This takes into account the pre-tax cash income for a household, adjusts for family size and then using a constant uniform national measure assigns (non) poverty status for all members of that household. When the OPM’s poverty thresholds were developed in the early 1960s this made some sense since few low-income households paid taxes and few received non-cash benefits. Decades later, however, the OPM was widely seen as having less utility. Rebecca Blank, who served on President Clinton’s Council of Economic Advisors and was then the Deputy Secretary at the Department of Commerce in the Obama administration, said of the OPM: “It is not too strong a statement to say that ... the poverty thresholds are nonsensical numbers.”14
Blank’s concerns were manifold, but of particular concern for assessing the effectiveness of government programs in alleviating poverty was the fact that the OPM did not count potentially significant in-kind benefits and services when calculating household income. Hence, the OPM likely overestimated poverty rates if the key indicator is household income by not including, for example, the value of Food Stamps and the Earned Income Tax Credit (EITC). In contrast, the OPM did not factor in any taxes paid or necessary work expenses, possibly resulting in the undercounting of the number of people who should be designated as living in poverty. In an effort to provide a second measure of poverty to the OPM, in November 2011 the Census Bureau released poverty data according to a new measure called the Supplementary Poverty Measure (SPM). Largely based on recommendations from the National Academy of Sciences, the SPM includes in-kind benefits and services as income, but also factors in any taxes paid, necessary work expenses and medical costs when assessing disposable income, while taking into account regional variation in the cost of living.15
This discussion of how to count poverty is not a matter of abstract debate or something to entertain statistical wonks while normal people get on with their lives. The question of whether particular anti-poverty strategies are perceived to be working varies significantly according to which poverty measure is used, which, in turn, can have an impact on whether certain policy tools are deemed worthy of expansion or even maintenance. These considerations were given a sharper focus as the calendar ticked into Obama’s sixth year in office as January 2014 marked the 50th anniversary of President Lyndon Johnson’s declaration of War on Poverty. In an official statement commemorating that landmark, the Obama administration reflected on the achievements of this effort: “if we hadn’t declared ‘unconditional War on Poverty in America,’ millions more Americans would be living in poverty today.”16 Yet the OPM suggests that after an initial reduction in poverty there was stagnation, with little further downward shift in the poverty population. In 1965, the OPM registered 17.3 percent of the population as living in poverty. This reached a low of 11.1 percent in 1973. The number then fluctuated, standing at 13.0 percent in 1980, 11.3 percent in 2000 before creeping back up again to 13.2 percent in the year of Obama’s election to the White House.17 These figures would suggest that the various monies and programs designed to reduce poverty were largely redundant from the mid- to- late 1970s onward.
This narrative was not without political consequence as it reinforced conservative ideas that welfare state spending was not achieving its primary objectives. As Blank reflects, this was a “reasonable conclusion” according to the official data. “Although this was a period of rapid growth in public spending on the poor, its effects were invisible because we had no official statistic that reflected how this public spending improved the resources or the lives of low-income families.”18 The SPM, however, which does take into account government spending in the form of Food Stamps, housing subsidies and the EITC has a different trajectory to the OPM, even though it actually shows higher overall poverty rates. The Census Bureau, in fact, only provides the SPM measure from 2009, but researchers at the National Bureau of Economic Research devised models to track longer-term effects of including a wider range of government programs, and household expenses, than is included in the OPM data. This research found that measuring poverty using the SPM and calculating a counterfactual rate of poverty-absent government programs shows the importance of those programs in reducing poverty:
The OPM shows the overall poverty rates to be nearly the same in 1967 and 2012—at 14 and 15 percent respectively. But our counterfactual estimates using SPM show that without government programs, poverty would have risen from 25 percent to 31 percent, while with government benefits poverty has fallen from 19 percent to 16 percent. Thus government programs today are cutting poverty nearly in half (from 31 percent to 16 percent) while in 1967 they cut poverty by only a quarter (from 25 percent to 19 percent).19
The authors particularly stressed the role of “tax credits and food and nutrition programs,” which are not captured by the OPM. This is particularly important in the context of the Obama presidency, which saw more expansion of these types of programs than new direct cash payments.