The following is a table of mine that shows how Tucson compares to El Paso, San Antonio and Albuquerque in population, in the percent of the population in poverty, in per-capita income, in median household income, in the percent of those 25 years and older who have at least a bachelor's degree, and in the percent of the population that is Latino. 

Statistics are for the municipalities in question and not the larger metro areas and surrounding counties.  There might be different findings if the boundaries of the analysis were expanded outside the cities to wealthy suburbs.   

Included in the table are comparisons to Seattle; Mountain View, California; and Arlington County, Virginia.  The reasons for their inclusion will be explained later.

Sources:  U.S. Census Bureau, World Population Review, CityData.com

The percentages of Latinos are listed because, in general, the higher the percent of Latinos in a locale, the higher the poverty rate and the lower the income—for the simple reason that many Latinos are recent immigrants who arrived in the USA without much money, education or skills.   

Note:  Before someone becomes hysterical over the above remark, it should be noted that I used to live in the barrio of San Antonio and am the grandson of Italian immigrants, who, along with other poor and poorly educated Italian immigrants in the early 20th century, increased the poverty rate and decreased the per-capita and median household income in the locales where they settled, at least initially. Also, for the record, I'm a classical liberal (a k a libertarian) who didn't vote for either Trump or Clinton.  It's a shame that in these times of super-sensitivity about race and immigration, it's necessary to qualify an analysis that has no racial motive and takes no stand on the border wall.

Anyway, as can be seen, Tucson has the highest poverty rate of the sampled cities, has the second-lowest per-capita income, and the lowest median household income.  While it has a high percentage of Latinos (42.9%), it has a lower percentage than the other cities.

High school graduation rates and K-12 test scores are not included in the analysis, because it's almost a certainty that these are negatively impacted by the other demographics and not by levels of school spending—as is borne out across the country with few exceptions.

So what accounts for Tucson's poor showing?  I have no idea.  It's not as if the other cities are more heavily industrialized than Tucson, or have a wealth of natural resources, or are on a navigable river, or are port cities.  One would think that Tucson would have an advantage in being the home of a major state university, the University of Arizona, which has a medical school in addition to all the standard schools of a large university.

What's your theory for the poor showing?

Seattle, Mountain View and Arlington County are included for the following reasons:  Seattle is where such rich companies as Microsoft, Amazon, and Starbucks are headquartered; Mountain View is a proxy for Silicon Valley, where Google, Facebook and other rich tech companies are headquartered; and Arlington County in Northern Virginia is one of the uber-wealthy suburbs surrounding Washington, DC, which has become one of the wealthiest metropolises in the nation. (I'll resist making a political comment.)  Arlington is also where Amazon is establishing a second headquarters.

The demographics for these wealthy locales show what such cities as Tucson are facing in trying to attract corporate headquarters of tech and tech-retail companies.  Site-selection committees see the Tucson demographics and run away.  Actually, as Amazon shows, they run to locales that have similar demographics as their home bases. 

Now to close with a political statement:  The home bases of these companies have a much lower percentage of Latinos than Tucson, but the companies espouse diversity and few restrictions on emigration from south of the border.  They couldn't be hypocrites, could they?