Skip Navigation
Expert Brief

Thwarting Amendment 4

New analysis shows that Florida legislation will disproportionately affect African Americans.

Published: May 9, 2019

Summary: Just months after Flor­idi­ans approved an amend­ment to their state’s consti­tu­tion to restore voting rights to 1.4 million people with felony convic­tions, the Legis­lature has passed a bill that weak­ens the impact of the ballot meas­ure by limit­ing who is eligible to vote. Amend­ment 4 was approved by 65 percent of voters in Novem­ber and enjoyed broad bipar­tisan support. And it was work­ing. The analysis below demon­strates that:

  • Nearly 100 times more formerly incar­cer­ated Flor­idi­ans registered in the first three months of 2019 than in previ­ous odd years.
  • Of the formerly incar­cer­ated Flor­idi­ans who registered to vote between Janu­ary and March of this year, more than 44 percent iden­ti­fied them­selves in their voter regis­tra­tion forms as Black, whereas Black voters comprise 13 percent of Flor­id­a’s over­all voter popu­la­tion.
  • The aver­age income of the formerly incar­cer­ated Flor­idi­ans who registered to vote between Janu­ary and March is nearly $15,000 below that of the aver­age Flor­ida voter.

There can be no mistak­ing the racial and class implic­a­tions of this regress­ive new legis­la­tion.


Flor­id­a’s long­stand­ing policy of perman­ent disen­fran­chise­ment had enorm­ous implic­a­tions for racial and economic justice in one of the largest states in the coun­try. More than four in ten indi­vidu­als released from prison over the past few years were Black, and people who live in minor­ity and low-income communit­ies are vastly overrep­res­en­ted among those caught up in the crim­inal justice system. Amend­ment 4 prom­ised to address this.

But on May 3, lawmakers passed a bill requir­ing Flor­idi­ans to pay back all fees, fines, and resti­tu­tion imposed as part of a sentence for a felony convic­tion – even those that a judge has conver­ted to civil oblig­a­tions – before they can register to vote. For some, these outstand­ing balances run to tens of thou­sands of dollars, thus making it virtu­ally impossible for these indi­vidu­als to register.


By look­ing at voter regis­tra­tion data since Amend­ment 4 went into effect, we can see that Black and low-income return­ing citizens are dramat­ic­ally impacted by this new legis­la­tion. Using data from Flor­id­a’s Depart­ment of Correc­tions and the Board of Elec­tions, we are able to identify Flor­idi­ans who have come home from prison and registered to vote. To be clear, this analysis includes only a slice of the popu­la­tion enfran­chised by Amend­ment 4: those indi­vidu­als released from prison since 1992. Our analysis does not include indi­vidu­als who were sentenced to felony proba­tion (because Flor­ida does not main­tain statewide data for this popu­la­tion), though many Flor­idi­ans sentenced to proba­tion likely also registered to vote during the first quarter of 2019. Nonethe­less, this data provides an early indic­a­tion as to the racial and income-level impact that Flor­id­a’s new legis­la­tion will have.

Before this year, Flor­idi­ans convicted of felon­ies had to receive indi­vidual pardons from the govern­ment in order to register to vote. Very few of these pardons were gran­ted. Between Janu­ary 2011 and April 2018, just 3,000 indi­vidu­als received them. Unsur­pris­ingly, very few formerly incar­cer­ated Flor­idi­ans registered to vote between 2010 and 2018 — on aver­age, fewer than 250 per year during that nine-year period.

Amend­ment 4 went into effect on Janu­ary 8, 2019, re-enfran­chising most Flor­idi­ans who had completed all the terms of their sentence. Many re-entry groups in Flor­ida took advant­age of the policy change and imme­di­ately launched voter regis­tra­tion drives. As the chart below makes clear, these drives were tremend­ously success­ful: In Janu­ary, Febru­ary, and March – months in which voter regis­tra­tions are usually very low – more than 2,000 formerly incar­cer­ated Flor­idi­ans registered to vote. That’s roughly ten times the annual aver­age number of regis­tra­tions for return­ing citizens in the preced­ing nine years and 99 times the aver­age for the first three months in 2017 and 2015 (recent years without a general elec­tion).

Using inform­a­tion in the state voter file, we know not just how many indi­vidu­als registered but also some of their demo­graph­ics. Of the formerly incar­cer­ated Flor­idi­ans who registered to vote between Janu­ary and March, more than 44 percent self-iden­ti­fied as Black. By contrast, just 13 percent of all Flor­idi­ans registered to vote are Black.

Using census data, we can also estim­ate the socioeco­nomic char­ac­ter­ist­ics of the elect­or­ate.[1] We found that formerly incar­cer­ated Flor­idi­ans who registered to vote in the first quarter of 2019 tended to be much lower income, have less college educa­tion, and come from neigh­bor­hoods with higher unem­ploy­ment than the rest of the state’s voters. In fact, the aver­age income for a formerly incar­cer­ated, newly registered Flor­idian is nearly $15,000 below the statewide aver­age.

The impacts of the crim­inal justice system fall heav­ily on lower-income and minor­ity communit­ies. Felony disen­fran­chise­ment policies, by exten­sion, dispro­por­tion­ally strip these communit­ies of their polit­ical power.

(Image: Drew Angerer/Getty)

[1] We used census block group data as a proxy for indi­vidual-level data, a common method in the social sciences when indi­vidual-level data is unavail­able. In this analysis, we assume that each regis­trant’s income is equal to the median income of their census block group. It is import­ant to note that this likely under­states the income discrep­ancy between recently re-enfran­chised indi­vidu­als and the rest of the elect­or­ate. Indi­vidu­als who have been to prison likely have lower incomes than those of their neigh­bors, which means that median block group numbers system­at­ic­ally over­es­tim­ate income for return­ing citizens.