Raising the Minimum Wage

INTRODUCTION

The federal government sets a national minimum wage rate. Each state can, but doesn’t have to, set their own minimum wage rate. This article will show you how to analyze a change in the minimum wage for your state or the nation.

IMPLAN is an Input-Output model and one of the limitations is that I-O models are fixed price (and wage) models. So, we have to consider where the increased cost for industries will come from. The industry might cut employment, change their spending patterns, take less in profits, or even go out of business. Thus, if the data is available, consider performing a Net Analysis to analyze both the changes to industries as well as the increases to household spending. At the very least, be sure to mention that the increases in wage rates do not come without a change in how industries spend their money.

EXAMPLE USING HOUSEHOLD INCOME

Let’s say we want to look at raising the minimum wage in Nevada from $9/hour to $15/hour that will take place in 2021. We will use the state of Nevada, Data Year 2019, and Dollar Year 2021. Our research shows that this will affect approximately 260,000 employees in the state. We will make the assumption that the additional cost to the industries will be shifted from profit, which is treated as a leakage in I-O modeling. 

We can find the average hours worked per year by navigating to

> Region Details

     > Occupation Data

          > Region Occupation Summary

               > Filter for 2- major

Looking at Occupation Code 41 - Sales and Related Occupations, which includes retail workers, we see the 139,709 wage and salary employees work 218,826,381 hours. Dividing the two, we see that on average each employee works 1,566 hours per year. The addition of $6/hour in minimum wage would translate then to $9,396 for each of the 260,000 employees in the state for a total of  $2,442,960,000.

If we want to assume that all of these workers are at the lower end of the earners in the area, we can utilize a series of Household Income Events. We can see the distribution of Households in each bracket by navigating to 

> Region Details

     > Study Area Data

          > Area Demographics

In Nevada in 2019, the distribution of Households looks like this:

If we make the assumption that all of the increases will go to households earning less than $30,000, we can distribute the total as 42% going to Households LT15K ($1,026,043,200) and 58% going to Households 15-30K ($1,416,916,800) by normalizing the data for these two groups (10% / (10%+14%) = 42%).

In Nevada, we know that unemployment insurance is 2.95%, social security is 6.2% and Medicare is 1.45% on average across the state. So we can take our wage values and reduce them by the total payroll tax rate of 10.6% to get to a Household Income value we can enter in IMPLAN. Multiply the values for each Household Income level by what they will take home after payroll taxes (100% - 10.6% = 89.4%). Then we see Households LT15K getting an additional $917,282,621 and Households 15-30K taking home an additional $1,266,723,619; for a total of $2,184,006,240.

Given that we used a Household Income Event, we will only see Induced Effects in our Results. This method shows us that our $2,184,006,240 of Household Spending will support approximately 13,500 jobs, with $650M in Labor Income, $1.4B in Value Added, and $2.3B in Output.

EXAMPLE USING LABOR INCOME

We could also model the additional $6/hour using a Labor Income Event if we don’t want to distribute the earnings across household income groups.

First, we will use the FTE & Employee Compensation Conversion Table to take our $6 in wages and convert that to a fully loaded Employee Compensation value. Due to the nature of the work, we will assume that the current minimum wage jobs will most closely match the rates for retail industries. With this assumption, we see the $6 in wages translates to an Employee Compensation value of $7.10.

Now we can multiply the $7.10/hour value x the 260,000 affected employees x the average hours worked per year of 1,566 (found in example above). This gives us 7.10 x 260,000 x 1,566 = $2,890,836,000. This is the value we will run through the Labor Income Event.

Given that we used a Labor Income Event, we will again only see Induced Effects in our Results. This method shows us that our $2,890,836,000 of Labor Income will support approximately 12,600 jobs, with $623M in Labor Income, $1.3B in Value Added, and $2.2B in Output.

COMPARING THE TWO METHODS

In our first example using Household Income Events, we assumed that all of those earning the additional income would be those earning less than $30K per year. Folks at this end of the earning spectrum spend all of their income and therefore none is left for savings. They also are most likely to pay the least in personal taxes.

Even though we modeled $706,829,760 more through the Labor Income Events, the Household Income Events give us the larger impact. The Labor Income Event spreads the earnings across each of the nine Household Income groups and so it includes folks that save money and likely pay higher personal taxes.

Both methods are correct. The determination of which one you should use in your analysis depends on not only the data you have available, but also on the assumptions you want to make in your analysis. The article Understanding Types of Income has further information on the key differences between Household and Labor Income Events.

RELATED ARTICLES

FTE & Employee Compensation Conversion Table

Net Analysis: Considering Both Sides of an Impact

Understanding Types of Income

Written April 21, 2021