Mayor Rick Kriseman is trying to win over voters in his re-election bid with a pitch about falling poverty in a lower income neighborhood. Kriseman, who faces former Mayor Rick Baker, says significant strides have been made since he took office, particularly in south St. Petersburg. "We are seeing poverty being reduced in south St. Pete at a number larger than the national average, the state average, county or in Tampa, Jacksonville," Kriseman said after signing his qualifying papers inside City Hall. This area of town is significant to Kriseman, who helped with efforts related to the 2020 plan, a five-year plan to reduce poverty by 30 percent in a portion of south St. Petersburg. In order for this claim to be true, the poverty rate in south St. Petersburg would have to be dropping faster than in those other areas he named. But Kriseman's evidence for the claim is imprecise. Instead of providing data for that area, Kriseman pinpointed the poverty rate of the African-American population in the entire city of St. Petersburg between two years. Kriseman's staff relied on a 2013 Pinellas County report that highlighted south St. Petersburg as one of five at-risk communities, areas marked by "high concentrations of poverty and a small return to the taxbase." According to the report, approximately 48 percent of the population in this zone live in poverty. The demographic breakdown of those who live in poverty is as follows: 63 percent are African-American, 27 percent are white, 5 percent are Hispanic, and 5 percent are of another race. City staff said given that the majority of residents in that zone are African-American, addressing poverty among African-Americans is a priority for the mayor. So to support his claim, Kriseman compared the poverty rate for the entire St. Petersburg black population (which should go without saying, does not entirely live in south St. Pete) against the rates of other big cities, the state and the country. Even with that allowance, the numbers don't hold up. Kriseman's team sent PolitiFact Florida two sets of data from the Census Bureau's 2014 and 2015 American Community Survey for the African-American population. (The 2016 numbers will be released in September.) According to that data, the black poverty rate in St. Petersburg decreased by 8.5 percentage points between 2014 and 2015, from 34.9 percent to 26.4 percent. The other areas Kriseman mentioned — Pinellas County, Jacksonville, Florida, the United States, and Tampa — showed a smaller reduction in the black poverty rate, between a drop of 4.8 percentage points to a gain of .2 point, respectively. Most importantly, the data neglects the large margin of error in census data. Based on our calculations, the difference in the number of African-Americans in poverty — for St. Petersburg as well as the other locations — was statistically unchanged from 2014 to 2015. In other words, there's no way to know for sure if poverty went up or down, because the change in the poverty rate is within the margin of error. In 2015, the margin of error for data from the Census Bureau was plus or minus 25 percent, or 4,131 people. For 2014, the margin of error was plus or minus 17 percent, or 3,459 people. Those margins are significant. The Census Bureau data shows that about 17,000 African-Americans were in poverty in 2015. With the margin of error, the estimated range is between 12,566 and 20,828. The range of African-Americans in poverty in 2014 is between 16,596 and 23,514. Taken together, the estimates show that the number of African-Americans in poverty could have held steady, or even increased from 17,000 to 20,000. We just don't know with a margin of error this wide. Joshua Wilde, an assistant professor of economics at the University of South Florida, said Kriseman was using the best data available despite its imperfections. While he couldn't say this for sure, he said there was a strong chance that poverty fell in south St. Petersburg among African-Americans between 2014-15. He calculated the p-value — a statistical indicator based on the confidence interval — to determine the chance poverty fell based on numbers provided by the Census Bureau. "Once you peel back the layers of the onion on the numbers, it turns out the best numbers we have are just highly uncertain," he said. Still, Melissa Radey, a Florida State University professor who specializes in poverty research and data, said it would be better to have more years when looking at trends in poverty. "Making statements about trends is not appropriate with only two years of data," Radey said. Kriseman stuck by his conclusion. "To dismiss the year over year Census report about the number of people in poverty as statistically insignificant overgeneralizes the complexities of real-time poverty eradication," he said. For a statement that contains an element of truth but ignores critical facts that would give the reader a different impression, we rate this claim Mostly False. Times staff writer Nathaniel Lash and senior news researcher Caryn Baird contributed to this report. Read more rulings at PolitiFact.com/florida.