Negative Direct Effect / Positive Induced Effects

I'm modeling a significant income shock in the NV statewide model that results in over $1B in household income and industry losses but also gets a result that shows a positive induced effect. I'm not entirely sure why. Here are the inputs I'm using: Brokers -3,555,790 Insurance Carriers -49,527,070 Physicians, dentists, & other health practitioners -138,779,913 Home health care services -13,849,256 Medical / diagnostic labs and outpatient and other ambulatory care services -12,461,714 Private hospitals -115,562,514 Capital -174,233,692 <10K 1044970 10K-15K 7,576,226 15K-25K 54,986,112 25K-35K 87,157,202 35K-50K 94,631,047 50K-75K -101,580,146 75K-100K -245,465,701 100K-150K -401,652,530 >150K -169,528,744