You beat me to the punch. I was going to mention the antiquated nature of the multiplier, too.
But I'd point out that there's also reason to believe that the multiplier may be too high. The fact that the average American spends a smaller portion of their income on food than they did in the mid-20th century is a mark of increasing affluence. The average American simply spends more on discretionary purchases than in the past. (Part of this is due to a decreased level of national saving.) So whether or not the current poverty level under- or overestimates poverty really depends on which form of poverty one prefers to measure: relative, how poor you are compared to your countrymen, or absolute, do you have enough money to provide the basics- food/shelter/clothing.
As one who prefers to judge poverty in absolute terms, I tend to think that a multiplier of 3 is too high, as many people who fall below the poverty line in the area in which I live have enough disposable income for items that would have been deemed luxuries in my modest (but not poor) childhood. An absolute measure is also much more dependant on local costs. It seems that this is how India chooses to calculate its levels of poverty. On the other hand, if income inequality motivates your views of poverty, then the the multiplier definitely needs to be revised upwards.
