The U.S. Supreme Court issued in March a split vote on a decision that, if passed, would have prevented public sector unions from requiring non-union government workers to pay union fees. This split represents a major victory for organized labor groups, which have been squeezed by changes in the labor force and anti-union legislation for decades. Just 45 years ago, nearly one in four workers were union members. Today, barely one in 10 are in a union.

The strength of organized labor in the United States depends largely on political and economic forces. Because these factors differ across the country, unions hold considerable power in some states and are virtually nonexistent in others. In New York, nearly 25% of state private and public sector workers are unionized, while in South Carolina, just 2.1% of workers are in unions.

There are many reasons for the differences in unionization between states. In an interview with 24/7 Wall St., David Macpherson, E.M. Stephens professor of economics and department chair at Trinity University, explained that politics is likely one of the biggest factors affecting the regional strength of organized labor. Over half of the states in the country now have right-to-work laws, with West Virginia’s statute passing this February. All of the states with the lowest shares of union membership have right to work laws, while Michigan is the only state with high membership which does.

Click here to see the states with the strongest unions

Click here to see the states with the weakest unions

“The reason unionization rates are low in right-to-work law states is because those are the states that don’t like unions and so they pass laws that are not favorable to unions.” Macpherson said.

Union membership is also at least partially determined by the industrial composition of a state. Macpherson explained that public sector jobs and manufacturing jobs tend to be more unionized, with transportation workers in particular very likely to be union members. Michigan’s substantial manufacturing sector, for example, is likely driving up unionization rates in the state.

The overall decline in union membership has been due to a massive decline in private sector organized labor. At the same time, public sector membership has actually increased. As a result, public sector workers account for nearly half of union membership nationwide, even as they only account for about 15% of the nation’s total workforce.

When asked about the long-term decline of unions, Macpherson explained there was no single reason, but that a number of likely factors have combined to reduce membership. “One of the reasons is the shifts in the labor force. It used to be about one in three workers were in the [typically heavily unionized] manufacturing sector. It’s not that way anymore.”

Unions came into being in this country to protect the rights and safety of workers, to negotiate with employers on behalf of employees, and to ensure fair pay and benefits. In time, they also supported political candidates that had their interests in mind. Macpherson explained that over the years, some of the roles unions have typically fulfilled have been replaced — often by the government, which has been enacting laws to protect workers. “For example, OSHA gives workers job safety protection that unions used to provide.” This change in roles has also contributed to the general decline of unions over the years. Today, labor laws and workplace restrictions provide many of the protections that unions were once able to fulfill.

Based on figures published by Unionstats.com, an online union membership and coverage database, 24/7 Wall St. identified the states with the highest and lowest union membership as a percentage of total employment. The database, which analyzes Bureau of Labor Statistics’ (BLS) Current Population Survey, provides labor force and union membership figures in both the public and private sectors, including manufacturing and construction. Additionally, 24/7 Wall St. reviewed annual average unemployment rates for each state from the BLS, as well as income and poverty data from the 2014 American Community Survey, produced by the U.S. Census Bureau. Rates of change were calculated over a 10-year period from 2005 through 2014.