top of page

RESPONSE TO JEFFREY JOHNSON’S BOOK REVIEW OF MY BOOK “THE EXTENT OF THE ATONEMENT”

Response to Jeffrey Johnson’s Review of David Allen, The Extent of the Atonement: A Historical and Critical Review (Nashville: B&H Academic, 2016)

David L. Allen

Dean, School of Preaching

Southwestern Baptist Theological Seminary

Let me first express my appreciation to Jeffrey Johnson for his review of my book The Extent of the Atonement (hereafter EA) in the most recent edition of the Founders Journal (Jeffrey Johnson, “Book Review: The Extent of the Atonement,” Founders Journal 109 [Summer, 2017]: 39–51). Link here: http://founders.org/reviews/the-extent-of-the-atonement/. I am especially grateful when Calvinists read and interact with it. Additionally, Johnson said a number of kind things about the book, for which I am also grateful. Interestingly, on the very day I read his review, I received in the mail his book He Died for Me: Limited Atonement and the Universal Gospel, (hereafter HDFM), which I had ordered. Having read his book, I have a better understanding of where he is coming from in his review of my book. I will occasionally refer to HDFM to clarify what Johnson is saying in his review of EA.


As a Calvinist, Johnson has attempted to connect how we think about the extent of the atonement and its impact on the nature of the gospel offer to all people, including the non-elect. He is aware of some of the problems of a strictly limited atonement in this area. Though the early chapters of the book are confusing as to what Johnson himself actually believes about the extent of the atonement, he finally makes clear he affirms universal atonement understood as an actual satisfaction for the sins of all people and which is objectively sufficient (HDFM, 129–43). This fact is obscured in his review (hereafter “Review”) because he never clearly asserts his views on the extent of the atonement and because he states he “disagrees with Allen’s view on universal atonement” (“Review,” 41). The fact is, he and I are in agreement that Christ’s death did indeed satisfy for the sins of all people. Johnson correctly notes there is both biblical and theological support for universal sufficiency in the sense that Christ died for the sins of all people. He cites key texts such as John 3:16, 1 Timothy 2:4–6, Hebrews 2:9–10, and 1 John 2:2. He correctly notes that “world” in these texts cannot be limited to the elect. He also correctly states that universal sufficiency is a “necessary inference” from the Scriptural commands to preach the gospel to all. Johnson clearly argues for a universal atonement under the terminology of “universal sufficiency” in HDFM, 129–43.


Nevertheless, there are a number of problems with HDFM which bleed over into his review of EA. The first problem is terminological. Terminology and definitions are always important for correct understanding. Johnson makes use of several terms, not all of which he defines. He appears to use the term “limited atonement” not in its correct theological meaning of a limited imputation of sins to Christ for the elect only, but for the belief that Christ’s death, either subjectively or objectively, secures its own application in the case of the elect. He also labels this idea “limited efficacy.” While the latter is usually a corollary of limited atonement, it is not what distinguishes limited atonement.


“High Calvinism” for Johnson appears to refer to anyone who adheres to the modern five points of TULIP and who explicitly affirms only a hypothetical sufficiency in the atonement for the non-elect, due to their (at least in the case of Owen) quantitative conception of the imputation of sin to Christ. “Moderate Calvinism” for Johnson appears to refer to those who adhere to all five points of the TULIP as well, but argue the death of Christ is sufficient for all people because of a qualitative conception of the imputation of sin to Christ, regardless of whether they think Christ suffered for the sins of all people, or bore the legal punishment of any who are non-elect. “Low Calvinism” refers to one who rejects “limited atonement,” which for Johnson means one rejects the notion of a self-applying atonement. On this view, Christ suffered for the sins of all people and Christ’s death is sufficient for all people. Johnson describes these as “4-point Calvinists” (HDFM, 15–25).


At the outset, Johnson is using the terminology in idiosyncratic ways. He applies his definitions to these terms and then reads my book through the lens of his own definitions, not through the definitions I set out at the beginning. Limited atonement is traditionally understood to assert Christ died only for the sins of the elect. No Calvinist who affirms that Christ did die for the sins of the non-elect can be described as affirming limited atonement (unless they use “atonement” to refer to the application, as in Dabney’s case) or as being a “High Calvinist.” For Johnson, “limited atonement” may mean that Christ only died for the sins of the elect, but it may also mean that Christ died, in some sense, qualitatively, for the non-elect. In either case, what is primary to limited atonement for Johnson is the notion that the atonement secures its own application. Anyone who believes that, regardless of what they believe about the actual extent of Christ’s sin-bearing in the satisfaction, is an adherent to limited atonement as far as Johnson is concerned. This is terminological disaster.

When I refer to “high Calvinists,” I mean those who affirm a strictly limited atonement understood to mean a limited satisfaction for the sins of the elect alone, such that Christ died only for the sins of the elect. By the phrase “moderate Calvinists,” I mean those who affirm Christ died for the sins of all people, or suffered in the stead of all humanity. When Johnson uses “moderate Calvinist,” his focus is primarily on those who say they affirm a universal sufficiency in the atonement. Thus, this category for him includes those Calvinists who are adherents of limited vicarious satisfaction and those who are not. This is also terminological disaster and leads to confusion.


The biggest problem with Johnson’s review is he has missed the main point and core issue of EA. The question concerning the extent of the atonement is “for whose sins did Jesus die?” Asked another way, did Christ obtain a satisfaction for the sins of all people or of some people only? Were the sins of all people imputed to Christ, or only the sins of the elect? All discussions about the nature of the sufficiency of the atonement or the efficacy of the atonement are dependent on the starting point of this question: for whose sins did Jesus die?


There are only two possible answers to this question. Either Christ died for the sins of all people, or he died only for the sins of the elect. Limited atonement (or Definite Atonement or Particular Redemption, if you prefer) theologically means Christ only died for the sins of the elect. For clarity’s sake, when I use the term “limited atonement,” I am referring to the viewpoint that Christ’s death satisfied only for the sins of the elect. When I use the term “unlimited atonement,” I am referring to the view that Christ’s death satisfied for the sins of all people. The question of the intent of Christ in making the atonement, the application of the atonement, the nature of the sufficiency of the atonement, and the efficacy of the atonement, can only be answered in light of the foundational question: for whose sins did Jesus die. Unless this is the starting point of the discussion, confusion will reign.


The question of whether the atonement is “self-applying,” which is a big part of Johnson’s critique, is only a secondary issue in EA. Most of the seven “weaknesses” Johnson asserts of EA deal with secondary issues I am not directly addressing and some of these he has either misunderstood from the standpoint of the history of the discussion, or he has misunderstood what I am saying, or both.


Johnson discusses issues of sufficiency and efficacy without once clarifying the core issue: for whose sins did Jesus die? His inattention leads to confusion in accessing what a particular Calvinist believed about this question based on what he said concerning the sufficiency or efficacy of the atonement. For example, John Davenant, signatory of Dort, believed the atonement to be sufficient in its nature to save all people, and he believed in unlimited atonement. Others at Dort believed in a strictly limited atonement, and at the same time affirmed the atonement was sufficient in its nature to save all people.


Johnson is aware of the various ways of understanding the sufficiency of the atonement, but he appears to be fuzzy on exactly what “universal sufficiency” means and then often fails to connect this with whether a particular person affirmed a particular form of sufficiency from the platform of a strictly limited atonement or a universal atonement. Furthermore, what one believes about the sufficiency of the atonement impacts what one believes about the nature of the offer of the gospel to all people, which is a key point in HDFM.


For example, in his chapter on “Universal Sufficiency” (HDFM, 43–57), Johnson cites numerous Calvinists whom he lists as affirming limited atonement and universal sufficiency: the Gibson brothers (editors of the book From Heaven He Came and Sought Her), Raymond Blacketer, John Calvin, Jerome Zanchi, Jacob Kimedoncius, Zacharias Ursinus, William Ames, The Canons of Dort, Thomas Boston, Jonathan Edwards, Andrew Fuller, John Brown, Charles Hodge, and D. A. Carson. The problem with this listing is that many of these men did not affirm limited atonement and rather held to a universal atonement. The only limitarians on Johnson’s list are the Gibson brothers, Blacketer, Ames, D. A. Carson, and possibly Boston. All others, excepting Boston, held to unlimited atonement.


This illustrates Johnson’s confusion. He assumes that since all these men, whom he believes are “moderate Calvinists” in his sense (or those who still maintain a self-applying atonement), say they affirm what they call the “sufficiency” of the atonement, then limited atonement is compatible with a universal sufficiency and thus the well-meant gospel offer is alleged to be consistently grounded in this sufficiency.


Both Davenant and the British Delegation at Dort (who believed in an unlimited atonement), along with other delegates who affirmed a strictly limited atonement, believed the gospel offer was grounded in its sufficiency. Some interpreted this sufficiency expressed at Dort as hypothetical for the non-elect (which Martinius specifically went on record objecting to in thesis 10, in Acta Synodi Nationalis [1620], part 2, p. 104; see EA, 155, especially 773n24), and others interpreted it as actual for all and they believed in a limited efficacy of the atonement such that it was designed or intended by God to be applied only to the elect. Also, those who held to a strictly limited atonement, and some who held to unlimited atonement (e.g. Davenant), believed there is an efficacy inherent in the atonement or in the extent of Christ’s impetration itself.


Johnson’s mistake is to equate “limited atonement” with an inherent limited efficacy (HDFM, 28) in Christ’s satisfaction. He further compounds this mistake when he asserts that all five-point Calvinists agree with a limited saving efficacy in the atonement itself. That is true for the strict variety. But it is also true that all Calvinists who reject a limited satisfaction (who may also be described as five-pointers by Dort’s standards) believe in a limited efficacy as well in terms of a special intent to apply the “merits” of Christ to the elect alone. This equivocation in the sense of “limited efficacy” is a major problem in HDFM and likewise in Johnson’s “Review” of my book.


Furthermore, just because a given Reformed theologian has said or claimed that the death of Christ is sufficient for all, it does not necessarily mean they affirm an “extrinsic sufficiency” for all. Anyone can inadvertently hold conflicting ideas without realizing it. Johnson often fails to recognize this. For example, he asserts: “Moreover, a case could be made that many, if not the majority, of those holding to limited atonement have upheld the universal sufficiency of the atonement” (HDFM, 45). While that is true in most cases verbally, it usually amounts to nothing more than a mere affirmation of infinite intrinsic value in the atonement.


What Johnson fails to recognize is many Calvinists who affirm universal sufficiency really mean nothing more than a conditional sufficiency, such that, if all were to come to Christ, it would have turned out that he had also died for them. It is valuable enough in itself to have been a price or a vicarious satisfaction for the sin of the non-elect, but it actually was not. Johnson does not see the illogic, or at least he does not mention it, of affirming a “limited atonement” in that sense and any kind of sufficiency that is anything other than hypothetical by entailment. This is illustrated in Johnson’s final sentence in the chapter “Universal Sufficiency”: “By holding to the Lombardian formula [Christ died sufficiently for all, but efficiently only for the elect] they were able to embrace a limited atonement and preach the gospel as a universal and sincere offer to all” (HDFM, 55). By “they” in the context and in the rest of his writing he is even including those at Dort and some later theologians who believe Christ only suffered for the sins of the elect. But this claim is precisely what the moderate Calvinists at Dort who rejected a strictly limited atonement (in the sense of a limited vicarious satisfaction for sin) argued against, and can be seen in the later writings of people like Andrew Fuller, Robert Dabney, and Charles Hodge.


As a consequence of confusion in this area, Johnson’s taxonomy of positions and people who held them is often flawed. For example, he speaks of Calvin, Jonathan Edwards, Andrew Fuller, and Charles Hodge as asserting “limited atonement” (HDFM, 45–54; “Review,” 42–43). This is false, both in the sense of a limited satisfaction and in terms of arguing for an inherent efficacy in the satisfaction itself. All four affirmed a universal satisfaction for sin (unlimited atonement), as I demonstrate in EA. Johnson rightly points out the way these theologians differ with Beza, Owen, and other high Calvinists on the nature of the sufficiency of the atonement, but he apparently failed to see how their view of sufficiency is grounded in their belief in a universal satisfaction for sins in the atonement—a position completely rejected by Beza, Turretin, Owen, etc. Moreover, Johnson rightly discerns that Calvin, Edwards, Fuller and Hodge assert the offer of the gospel is grounded in the universal sufficiency of the atonement, but fails to discern that this is due to the fact of their commitment to a universal satisfaction or atonement, not from a belief in a limited satisfaction or atonement.


Johnson also has misunderstood the Lombardian formula and its original intent and usage. The Lombardian formula—Christ died sufficiently for all, but efficaciously only for the elect—in the first part originally meant Christ actually died for the sins of all people (universal sufficiency) and this part was later reinterpreted or revised by Beza, Owen, and Turretin to mean Christ death would have been sufficient for all had God so intended it to be. This is only a hypothetical sufficiency, not a real or universal sufficiency. Johnson is aware of these distinctions, and employs them in his book. But he seriously errs when he states the Lombardian formula was originally a statement affirming limited atonement: “In short, unlike Beza’s alteration, the original Lombardian distinction (universal sufficiency and limited efficacy) affirmed the doctrine of limited atonement while also providing warrant for the universal offer of the gospel” (HDFM, 115). The original Lombardian formula affirmed universal atonement, in the sense of universal satisfaction for sins, and was later altered by Beza and company to support the new position of “limited atonement” in the sense of a limited satisfaction for sin. One might add that there is nothing in the second part of the formula as stated by Lombard that affirms an inherent efficacy in the satisfaction itself; rather, the efficacy, as taught by the medieval Augustinians, was grounded in God’s will or special intent to apply.


Moreover, Johnson fails to recognize that many of the people he labels as affirming an actual sufficiency and not a hypothetical sufficiency (“moderate Calvinists” in his taxonomy), actually held to an unlimited or universal satisfaction, not a limited atonement. Johnson thinks the debate is over the nature of the sufficiency among moderate Calvinists and high Calvinists, but he wrongly thinks that many or most of those in both groups affirmed a “limited atonement” since he uses this label to refer to an inherent efficacy in the atonement itself. This is an egregious historical error and involves further equivocation in the sense of “limited atonement.” In Johnson’s taxonomy in an appendix, he labels those who actually hold to an unlimited atonement as either moderate or “low Calvinists,” depending on what he thinks their thoughts are on the inherent efficacy idea. The following names are examples: John Calvin, John Davenant, Edmund Calamy, the later Andrew Fuller, Charles Hodge, W. G. T. Shedd, Amyraut, John Preston, James Ussher, Lewis Sperry Chafer, and Norman Douty (HDFM, 189). He does not seem to realize that all of these Hypothetical Universalists, or moderate Calvinists, held to unlimited atonement.


Johnson correctly identifies Beza, Owen, and Perkins as affirming only a hypothetical sufficiency for the non-elect, such that it “could have been (hypothetically) sufficient for all the world, but only if God had intended it to be sufficient for them.” He then states that “many 5-point Calvinists, as Allen notes, argued for more than just hypothetical (intrinsic) sufficiency” (“Review,” 40). I did so, but my point in the book is that none could do so with consistency if they were the kind that held to a limited satisfaction. Johnson seems to think they can do so with consistency, if they merely say the atonement is sufficient for all (e.g. Berkhof, Boettner, Kuiper), and do not adopt explicitly hypothetical language when speaking of sufficiency: “For the majority of Calvinists, universal sufficiency is extrinsic; the atonement is actually sufficient for the salvation of the non-elect.” But this is the point of contention for all non-Calvinists like me, and actually the minority of Calvinists who affirm an unlimited atonement, because we assert it is not possible for the atonement to be actually sufficient for the non-elect on a limited satisfaction scheme.


Johnson appears to be confused about how, historically, sufficiency has been described by Calvinists who hold to limited atonement and Calvinists who affirm unlimited atonement in the sense in which I defined these terms in EA. All Calvinists who affirm limited atonement believe only in an intrinsic sufficiency by definition, in the case of the non-elect. Extrinsic sufficiency by definition means Christ actually atoned for the sins of the non-elect, and no Calvinist who asserts limited atonement (according to my definition) believes that. I never argued that any 5-point Calvinists (in the strict sense of the label) affirmed extrinsic sufficiency for that would be contradictory to their position. See my definitions of “Infinite or Universal Sufficiency,” “Limited Sufficiency,” “Intrinsic Sufficiency,” and “Extrinsic Sufficiency” in EA, xxii–xxiii.


Johnson quotes the Canons of Dort to make the point that Calvinists who affirmed limited atonement also affirmed the “sufficiency” of the atonement: “And, whereas many who are called by the gospel do not repent nor believe in Christ, but perish in unbelief, this is not owing to any defect or insufficiency in the sacrifice offered by Christ upon the cross, but is wholly to be imputed to themselves (2:6)” (HDFM, 103). Johnson does not seem to take note of the fact that the strict Calvinists at Dort interpreted the sufficiency to be only an intrinsic sufficiency in connection with the non-elect. The truly moderate Calvinists, such as Davenant and Martinius, interpreted it to be an actual extrinsic sufficiency such that Christ actually died for the sins of the non-elect as well. The key here is what meaning one invests in the term “sufficiency.”


Nevertheless, Johnson asserts: “Subsequently, not only does universal (extrinsic) sufficiency provide a warrant for the universal offer of the gospel, it brings greater judgment on those who reject the gospel (“Review,” 40–41). He then appeals to John Calvin who stated:

And indeed, our Lord Jesus was offered to all the world… Our Lord Jesus suffered for all and there is neither great nor small who is not inexcusable today, for we can obtain salvation in Him. Unbelievers who turn away from Him and who deprive themselves of Him by their malice are today doubly culpable. For how will they excuse their ingratitude in not receiving the blessing in which they could share by faith? And let us realize that if we come flocking to our Lord Jesus Christ, we shall not hinder one another and prevent Him being sufficient for each of us… Let us not fear to come to Him in great numbers, and each one of us bring his neighbours, seeing that He is sufficient to save us all (ibid, 41).

Johnson has misunderstood this quotation by Calvin as well. Johnson presumes Calvin affirms limited atonement (“Review,” 50). Calvin clearly affirms universal atonement here and in dozens of other quotations, as I make clear in EA. Johnson fails to include the previous sentences in Calvin’s paragraph which make it clear Calvin is affirming universal atonement! And there is nothing in this quote that speaks to inherent efficacy in the satisfaction itself (or “limited atonement”) in Johnson’s sense either. Here are the missing sentences which, along with the rest of Johnson’s quote above, can be found in EA, and pertain to Calvin’s idea of universal satisfaction.

That, then, is how our Lord Jesus bore the sins and iniquities of many. But in fact, the word “many” is often as good as equivalent to “all.” And indeed our Lord Jesus was offered to all the world. For it is not speaking of three or four when it says “God so loved the world, that he spared not His only Son.” But yet we must notice what the Evangelist adds in this passage: “That whatsoever believes in Him shall not perish but obtain eternal life” (EA, 60).

The quotation is from Calvin’s sermon on Isaiah 53. In the same book of sermons on Isaiah 53 as well as his Commentary on Isaiah, Calvin clearly states the “many” means “all,” and he defines the “all” as the entire human race. In fact, Calvin says: “none is excepted” because “the prophet includes all . . . .even to the last individual, . . . .without any exception” (Calvin, Sermons on Isaiah, 66, 70, 79, 81). The “greater judgment” of which Calvin speaks is due to the fact of the universal suffering for all/bearing the sins of all, plus their unbelief, or refusal of this grace.


Extrinsic sufficiency (Christ died for the sins of all people) does indeed provide a universal warrant for the gospel offer. That’s what Calvin believed, along with all who are rightly called “moderate Calvinists” historically (See John Humfrey’s 17th century use of this label in EA, 228n324; R. Godfrey, J. Beeke, J. Moore, and many other contemporary Reformed historians use this descriptor in the same way for the “middle-way” men, as they were described in the 17th century). But there is no sufficiency in the death of Christ that is extrinsic for the non-elect in what I have called the “limited atonement/satisfaction” scheme. Johnson appears to think there is, or at least he has not made himself explicit or clear enough on the sense of sufficiency he wishes to advocate: one involving universal sin-bearing or not.


Johnson suggests that Amyraut did not believe there was a special design in the death of Christ for the elect (“Review,” 41). What is his primary source for this? Whether Amyraut saw inherent efficacy in the death itself is unknown, so far as I can tell, but not likely. What is clearly the case, however, is that Amyraut believed in “limited efficacy” in Christ in the sense of God’s special intent to apply His death to the elect alone (see EA, 166, and the reference to John Quick’s Synodicon for documentation of this point).


What was it “within the death of Christ itself” that makes it inherently efficacious only for the elect? Johnson does not say. Moderates can affirm that Christ died for all people, but not equally for all in terms of intent to apply. The inequality is usually located in the divine intention, not something inherent in the satisfaction itself. As Charles Hodge, and many other Calvinists who affirm an unlimited atonement affirm, there is nothing in the atonement itself that secures its own application. That is the work of the Holy Spirit in tandem with the design and intent of the Father and Christ. It is false to say, “The atonement, by special design, secured its own application for the elect alone.” All that seems necessary to say, given a Calvinistic perspective on election, is that “Christ, in making his all-sufficient satisfaction, had a special design or purpose for it to be applied efficaciously by the Holy Spirit to the elect alone through his grant of faith.”  Or, as Tony Lane said describing Calvin and others in his school of thought, “…our [i.e. the elect’s] salvation is made certain, not merely possible, by the combined work of Father, Son ‘and’ Holy Spirit (i.e., not by the cross alone, taken in isolation)” (“The Quest for the Historical Calvin,” EQ 55.2 [1983]: 100). Calvin himself said, “…as long as Christ remains outside of us, and we are separated from him, all that he has suffered and done for the salvation of the human race remains useless and of no value for us” (Institutes, 3.1.1.).


Johnson notes that the limitation is in the design of the atonement by the Father and the application of the atonement by the Holy Spirit. He then asserts the atonement has “objective efficacy,” by which he means there is an inherent efficacy or inequality within the atonement itself. (“Review,” 41). Johnson needs to explain to us how it is that the atonement has “objective efficacy,” since no one is saved by it until they believe according to the New Testament. Texts in the NT used to prove the purchase of faith or an inherent efficacy in Christ’s death also fail to convince (see EA, 757–58; esp. 757n396).


Johnson states: “For this reason, the hypothetical universalism of Preston and Davenant remains within the boundaries of the Canons of Dort, while the hypothetical universalism of 4-point Calvinists remains outside of those boundaries” (“Review,” 41). This statement is fraught with problems of ambiguity. First, it is nonsensical. All Hypothetical Universalists are “four-point Calvinists,” according to modern expression, in that they believed Christ died for the sins of all people. Second, Preston and Davenant, as many other Calvinist historian/theologians have demonstrated, were Calvinists who affirmed a universal satisfaction/atonement. Third, the Synod of Dort did not codify limited atonement over against universal atonement. The final Canons were written with sufficient ambiguity to allow all Calvinists to sign, whether they believed in a limited or universal satisfaction/atonement, and regardless of how they defined the sufficiency of the atonement.


Johnson labors under the misconception that I think Andrew Fuller, and Charles Hodge, for example, denied limited efficacy (“Review,” 44). In fact, I clearly state the opposite in EA, in the sense of a special intent to apply. What I do state is that they denied limited atonement in the sense of a limited satisfaction for sin. That is different from denying limited efficacy. Fuller and Hodge likewise believed in universal sufficiency because they believed Christ actually died for the sins of all people. Johnson offers no counterfactual evidence against my claim for Fuller, and the quotations he cites from Edwards and Hodge don’t state they affirmed limited atonement by my definition at all (“Review,” 42–43). Moreover, as I document in EA, Fuller explicitly rejected the “purchase of faith” idea (EA, 757n396) and Hodge grounds the efficacy in God’s purpose in the Covenant of Redemption, not in the satisfaction itself.


Johnson appears confused on the issue of sufficiency in connection with the Lombardian formula. The first half of the formula supports more than just bare universal sufficiency; it affirms a universal satisfaction for sin as well, which is why the atonement is universally sufficient. As I demonstrate in EA, the Lombardian formula was revised by Beza and later limitarians to be a hypothetical sufficiency only because they denied Christ substituted for the sins of all. Johnson also has conflated the notions of limited atonement and limited efficacy. He wrongly equates the two. The first has to do with the actual extent of Christ’s sin-bearing. The latter concerns the limited intent to apply the atonement to the elect only. Contrary to Johnson, there is nothing in the Lombardian formula itself about inherent power in the atonement “to bring about its own application” (“Review,” 45).


Johnson speaks of the “limited efficacy” of the atonement, by which he means “the atonement secured its own application.” The same idea is found throughout HDFM. He doesn’t seem aware that someone like Baxter or Amyraut can also believe in “limited efficacy,” but not in the same sense as Johnson is using the phrase. They do not think there is in the death of Christ something that secures its own application. Rather, they see the electing will of God as that which results in the grant of faith to some people in the effectual calling of the Spirit, so that the death of Christ is effectually applied to all those appointed to eternal life. By “limited efficacy,” Baxter did not argue that Christ’s death purchased faith (it is unclear where Amyraut stood on this issue), our new nature, etc. Johnson just assumes that’s what “limited efficacy” means, and so in HDFM he categorizes those who disagree with the idea that the atonement secured its own application as “low Calvinists” (HDFM, 23, 189). In point of fact, many on his list of “moderate Calvinists” also disagreed with the idea that the atonement secured its own application.


All orthodox Christians, whether Calvinists, Arminians, or otherwise, limit the efficacy of the atonement only to those who believe, though they argue about the ultimate decisive cause (see my chapter on “The Atonement: Limited or Universal,” in Whosoever Will, 65). This is totally different from limiting the actual extent of the atonement in the sense of “for whose sins did Christ die,” which is the proper state of the question on extent.


From the quotations he cites, Johnson also apparently thinks Boston, Edwards and Charles Hodge held to limited atonement (“Review,” 42–43). This is not true of Edwards and Hodge, as I have defined “limited atonement,” though I would concede Boston for the sake of his argument. But if Boston thought “Christ died only in the room and stead of his elect,” in terms of a limited legal satisfaction (limited atonement) then Johnson errs by saying Boston rightly “held both sides of the Lombardian formula” (“Review,” 42) in the original sense.


The Edwards quote clearly demonstrates he held to a form of universal atonement or redemption. Notice carefully in the Edwards quote the word “design.” Johnson follows the quote with his explanation using the word “purpose” (“Review,” 42). Edwards, like all Calvinists, affirms that the design or purpose of the atonement is to save only the elect. But Edwards, like all Calvinists who affirm unlimited atonement (the four-point Calvinist moniker is imprecise so I resist using it), affirms Christ died for the sins of all people with respect to the extent of the atonement. The second Edwards quote does not prove he held to limited atonement as I have defined it either. In fact, every orthodox Christian could affirm what Edwards says here. For numerous quotations proving Edwards held to unlimited atonement, see EA, 268–77. And when Edwards spoke of “what purchase he [Christ] should get,” it is clear Edwards is referring to people who shall be effectually redeemed, not to things, as if faith is a commodity, being purchased in his satisfaction. Neither quotation shows that Edwards believed Christ purchased faith in His satisfaction itself, as Johnson seems to think. He just assumes Edwards’s “something particular in the design of his death” connotes the idea of Christ’s death securing its own application (ibid.). That may be elsewhere in Edwards’s writings, but the two quotes he lists do not support his case even on that point, which is tangential to my argument concerning universal satisfaction or redemption in Edwards’s theology.


With respect to the Hodge quote, I clearly affirm that Hodge believed in a limited efficacy of the atonement in terms of its planned application to the elect only in his idea of the “Covenant of Redemption,” which involves a special intent to apply. However, Hodge did in fact affirm an unlimited atonement with respect to the extent of Christ’s satisfaction. For proof of Hodge’s support of unlimited satisfaction, see EA, 330–32. Look carefully at the Hodge quote Johnson provides. Hodge says Christ did not die “equally” for all mankind (“Review,” 43), but his reference is to God’s purpose to save in the “Covenant of Redemption.” No Calvinist denies that sense of “limited efficacy.” Hodge is affirming a dual intent in Christ in making a satisfaction for sins. God purposed that Christ should satisfy for the sins of all people (a universal extent), and He purposes to save the elect alone. Hodge explicitly rejects the double payment argument, and so, undermining any pecuniary casualty in Christ’s death, rejects the idea that the satisfaction secures its own application.


Johnson’s final sentence: “So, it seems strange to remove these men from the limited atonement camp when their positions are safely within the orthodoxy of the Canons of Dort” (“Review,” 43) is therefore a non-sequitur: the men I name are both within the orthodoxy of the Canons of Dort, and they held to unlimited or universal satisfaction. If Johnson agrees that they affirm a universal sufficiency in the sense of a satisfaction for the sins of all men, then as I defined “limited atonement,” he would also have to “remove these men from the limited atonement camp.” As to whether Dort mandates the belief in a satisfaction that in itself secures its own application—or Johnson’s sense of “limited atonement”—through a direct sense of the purchase of faith, which would then exclude Baxter’s indirect sense of that idea, is also debatable. But it is yet again tangential to my arguments about universal satisfaction in the theology of Edwards, Hodge, et al.

Johnson asserts: “Allen fails to acknowledge that the main distinctive of particular redemption is the doctrine of limited efficacy” (ibid.). Johnson errs here, and again engages in ambiguity and equivocation by exchanging the label “limited atonement” with “particular redemption.” That is not the main distinctive of limited atonement, as properly defined. That distinctive has been: for whose sins did Christ make satisfaction? For all men? Or only for the elect? This is the crucial point Johnson is continually missing and doesn’t address clearly in HDFM or his “Review.” To conflate limited atonement with limited efficacy merely begs the question concerning the atonement’s extent.


Johnson continues the error of equating and thus conflating limited atonement with limited efficacy. The issue is the extent of the satisfaction: for whose sins did Christ die? Johnson leaves the impression that the intent, extent, and application of the atonement are all coextensive, so that God’s effectual intent to save also limits the extent of the atonement’s inherent efficacy, which then results in its limited application. Johnson doesn’t merely see particularism in Christ’s intent to save alone; he’s arguing for particularism in the satisfaction of Christ itself, so that “limited atonement” involves the latter concept, and “limited efficacy” involves more than the former.


Johnson states that I think the debate centers on “the extent of the atonement’s sufficiency” (ibid.). That is imprecise. The debate centers on the nature of the atonement’s sufficiency: whether it is only extrinsically sufficient for the elect and only intrinsically sufficient for the non-elect (hypothetically sufficient in nature for all but not actually sufficient for the non-elect since Christ did not die for the sins of all), or whether it is extrinsically and universally sufficient for all (because Christ did actually die for the sins of all).


When Johnson states: “There are many advocates of limited atonement who do not believe in limited sufficiency as exposed by Beza, Owen, and Perkins. Beza, Owen, and Perkins, however, do not represent all limited atonement advocates” (ibid.), he continues to demonstrate a mistaken understanding of sufficiency, without taking into account the distinction of Davenant, and all Calvinists who affirm an unlimited satisfaction/atonement, who argue against a “bare” sufficiency for some, and argue for  an “ordained” sufficiency for all. All Calvinists who maintain a limited substitutionary satisfaction, or “limited atonement,” can only have a “bare” sufficiency in Christ for the non-elect, whether they speak in the explicitly hypothetical language of Beza and Owen, or whether they ambiguously say Christ’s death “is” sufficient for all (Berkhof, Boettner, Kuiper, et al.).

Johnson states: “Allen is right in pointing out that intrinsic (hypothetical) sufficiency, held by Beza, Perkins, and Owen, has been the minority position among 5-point Calvinists” (ibid.). Actually, hypothetical sufficiency is the only sufficiency that can be held in any consistent fashion by strict Calvinists, since they deny Christ died for the sins of all people.


Johnson states: “Limited efficacy is the one thing all advocates of limited atonement have in common” (“Review,” 44). The problem here is another case of ambiguity and equivocation; all evangelicals believe in limited efficacy, in a sense, in terms of application. They don’t believe in an unlimited efficacy for the salvation of all people in a literal sense. That would be universalism. That said, it must be noted that all Calvinists who affirm unlimited atonement/satisfaction also affirm limited efficacy, in terms of Christ’s special intent to save, which then results in the Holy Spirit’s “effectual call.” Where there is some disagreement is in the idea of Christ’s satisfaction itself securing its own application, or that sense of “limited efficacy,” and so there is debate among them on the double-payment argument, a literal/direct “purchase of faith,” and related arguments. Johnson errs again when he says “Beza, Perkins, Owen, Boston, Edwards, Fuller, and Hodge (with every other 5-point Calvinist) believed that the atonement secured its own application for the elect and for the elect alone” (ibid.). As already stated above, and note well, Fuller rejected the idea of “purchasing of repentance and faith, as well as other spiritual blessings” (italics mine) unless used in an “improper or figurative sense” (Andrew Fuller, “The Harmony of the Christian Religion Considered as an Evidence of its Divinity,” The Complete Works of Andrew Fuller, ed. by Andrew Gunton Fuller [Boston: Gould, Kendall and Lincoln, 1836], 1:164; See EA, 757n396), and that concept is historically connected with all those arguing for an intrinsic efficacy in the satisfaction itself. Likewise, Hodge grounded the securing of the application in the eternal Covenant of Redemption, and clearly rejected the double-payment argument, which has also been central to arguing for intrinsic efficacy in the satisfaction. It makes no sense, then, using Johnson’s taxonomy and categories, to list Fuller and Hodge as so-called “5-point Calvinists.”

Johnson states: “Allen is mistaken when he limits the extent to sufficiency alone. He is wrong when he says: ‘For all who affirm limited atonement, the atonement can only be sufficient for those for whom it is efficient.’ This is not true for the majority of 5-point Calvinists who have affirmed that actual (extrinsic) sufficiency extends to all universally” (“Review,” 44).


My point, along with all non-strict Calvinists, is that when the stricter sort of Calvinists assert the atonement “is” sufficient for all, while still maintaining a satisfaction for the sins of the elect alone, they always mean it is hypothetically sufficient for the non-elect by necessary entailment. Even when some claim the atonement is “sufficient” for all, they are doing so in a manner inconsistent with their belief in a limited satisfaction, namely, a limited imputation of the sin of the elect alone to Christ. Strict Calvinists, or High Calvinists as I have defined them, cannot explain how the atonement is in any way sufficient for those for whom it was not made, i.e., the non-elect. Johnson’s claim that most Calvinists can consistently claim the sufficiency is for all is a claim I and all less-strict Calvinists dispute. That is one of the major points of contention.


Johnson states: “Thus, to disprove limited atonement, as it is presented in the Canons of Dort, Allen has to do more than disprove the limited extent of its actual sufficiency. Allen has to do something more difficult, he has to disprove the limited extent of its inherent efficacy” (ibid.)

First, Dort did not canonize limited atonement over against unlimited atonement, as previously noted, as I have used these labels.


Second, Johnson thinks that in order for me to defeat limited atonement, I have to “disprove that Christ died effectually only for some” (ibid.). But that’s not really the issue. By “limited atonement” I mean “limited satisfaction,” and that is always conjoined with the claim that the efficacy is limited because the satisfaction itself is limited. The limited satisfaction theory says, “Christ only died for the sins of the elect, therefore it is only efficacious for them.”


Third, Johnson’s concept of limited efficacy is problematic, because it raises the question, could the satisfaction be conditionally applicable (or potentially) efficacious for all people? Johnson wants to answer “yes,” but according to his logic, the answer would seem to be “no.” Johnson has to face this problem, and it is crucial: if the atonement is conditionally applicable to all, then it cannot be said, as he wants to say, that the satisfaction contains its own self-applying or built-in efficacy for all for whom it was especially made.


Johnson states that according to Lombard, the cross accomplished universal (extrinsic) sufficiency for all, and “it ‘brought about salvation only for the predestined’ due to its ‘efficacy'” (“Review,” 45; italics mine). Actually, the cross accomplished propitiation and expiation for all people. As a result of this accomplishment it has “extrinsic” or “universal” sufficiency. The cross itself did not “bring about” salvation for anyone. That is the work of Christ himself working through the Holy Spirit according to Scripture. Johnson is stuck in a false framework that assumes the cross itself actually does something intrinsic to itself. That is the problem in his categories.

Johnson states: “The efficacy of the atonement is the atonement’s inherent power to bring about its own application” (ibid.). But this is precisely what Charles Hodge denied, along with many other Calvinists. Where is the scriptural support for this claim? Johnson completely ignores the point of universal sin-bearing in Calvin, the later Fuller, C. Hodge, and a host of other Calvinists I documented. This is the critical foundational point I made in EA. To state that the atonement effects its own application is to beg the question, and change the central thesis and focus of my book EA.


In the history section of EA, my aim was to demonstrate that moderate Calvinists, as I have defined that label, never denied Christ died with an effectual intent to save only some (the elect), or that sense of “limited efficacy.” They, along with all high (strict) Calvinists, affirm such. My point was to show historically that these men also held that Christ died for others (the non-elect) in the sense of bearing their sin, the key biblical and theological concept which those I have labeled the “limited atonement” folks have been trying to deny for 430 years.


Johnson wrongly concluded that I argued that Boston, Fuller, and Hodge “denied limited efficacy” (Review, 47). I argued no such thing, in terms of Christ’s intent to save as found in the limited grant of faith by the Spirit to the elect alone. But Johnson is correct that most five pointers (or the stricter sort), and I would grant him Boston, believed “that the atonement secured its own application by the special intention or design of God. That is, God caused the cross to be inherently efficacious because that is how He designed or intended for the cross to operate” (“Review,” 48). But my point in some sections of the book (EA, 757–58) is: where is the Scriptural evidence for this claim? Phil. 1:29 is certainly not a good proof-text, and that is the usual text brought forward to substantiate the claim.


The question is what does Johnson mean by actual sufficiency? What does that look like on the terms of limited imputation which is in some men he claims to be moderate? It can only have been a hypothetical sufficiency for the non-elect by entailment. There is enough inherent value in the satisfaction that had God elected (I’m assuming Calvinism’s notion of election here for the sake of the argument) otherwise, there would have been enough external sufficiency in the atonement for them or for more than were actually elected in this world.


Johnson buys into the error that faith is somehow purchased at the cross, a notion argued by Owen and rebutted by Richard Baxter along with numerous Calvinists who affirm an unlimited satisfaction or universal substitution, such as Fuller. Johnson says: “And, though the Bible does not directly say faith was procured by the death of Christ, it does teach that the new nature was procured by the death of Christ” (“Review,” 48). Johnson rightly acknowledges here that the “faith purchase” notion is foreign to the direct teaching of the New Testament, but then proceeds to substitute the concept of “new nature” for “faith” as that which is purchased by the cross. But, where does Scripture say that the new nature of a believer was somehow “purchased” by the death of Christ? It does not. Titus 2:14 says Christ gave himself for us in order to make us who believe holy, or zealous for good deeds. It doesn’t say his satisfaction bought a new nature. The same goes for Hebrews 13:12, which Johnson brings forth to establish his point. Granted, some less strict Calvinists (such as Davenant) speak of faith being merited or purchased by Christ in his impetration, but they are unclear on whether this is to be considered directly or indirectly, literally or non-literally. For a critique of the “faith purchase” notion from other Calvinists, see EA, 207–19.

Johnson says I have “misunderstood” the gift of faith (ibid.). I have not misunderstood it at all. I deny the claim that faith is a gift purchased in Christ’s satisfaction itself, as have numerous Calvinists.


Johnson states: “So, according to Allen, the question of (1.) intent is connected to the question of (3.) application, but not connected with the question of (2.) extent” (“Review,” 49). This statement is misleading because it obscures what those moderate Calvinists who affirm unlimited satisfaction/atonement mean by the term “intent.” First, God “intended” that Christ should suffer for the sins of all men. Second, God “intended” that the saving benefit of the atonement would be applied only to the elect. One must acknowledge and distinguish between these two notions of intent. The stricter Calvinists do indeed conflate the intent and extent of the atonement, and deny any intent on the part of God to provide an atonement for the sins of the non-elect. I do not “disjoin” the three, but as Johnson rightly noted, I do distinguish the three, but in order to engage in accurate history, or to properly describe the beliefs of theologians such as Davenant.

And, I might add, if Johnson or anyone else is going to engage in accurate historical description, they will have to distinguish among them as well.


Johnson states: “I agree that Calvin, Edwards, Boston, Fuller, Hodge, and many other Calvinists believed that the death of Christ has sufficiently opened the door of salvation for everyone, and that this universal (extrinsic) sufficiency makes the gospel a sincere and warranted offer to all” (“Review,” 50). Notice what is missing in this statement: there is no clear affirmation of whether Christ died for the sins of all people or only for the sins of the elect. If Johnson takes the latter view, then he is at odds with Calvin, Edwards, Fuller, and Hodge, for they all affirmed the former (excepting Boston). In their theology, the atonement opened the door of salvation for all people, because Christ died for the sins of all people: a universal, extrinsic sense of sufficiency for all. It is this universal, extrinsically sufficient atonement for all which grounds God’s sincerity in his indiscriminate gospel offers.


Johnson continues: “To disprove limited atonement, Allen has to prove that the atonement did not effectually purchase, redeem, and purify a particular people for God” (ibid.). This statement reveals Johnson’s confusion throughout his review of my book. He’s defining “limited atonement” in ways I have not. I have disproved “limited atonement” in the sense of a limited vicarious satisfaction, but Johnson is missing that point, and rather focusing on an atonement that secures its own application. Also, I believe the atonement is effectual for all who believe (the elect). I just disagree with Johnson’s view on the intent of the atonement and his notion that the atonement somehow secures its own application. He has misplaced and misstated the point at issue: did Christ die for the sins of all people or only for the sins of some people? All Calvinists who reject limited vicarious satisfaction affirm two things: A) Christ died for—bore the sins of—all people as to the sufficiency of the satisfaction, and B) Christ died for—or especially intended to save —only some people, or that sense of “limited efficacy.” Johnson wants to make B the issue of dispute, but with the additional component of a satisfaction that somehow secures its own application (B1). There is no dispute here among Calvinists on B, except in some cases with respect to Johnson’s added component (B1). The dispute is essentially over A. Johnson seems unclear on all this, and that is the problem with his review. To disprove limited atonement in the sense of a limited satisfaction, all one has to do is to demonstrate where Scripture affirms an unlimited atonement or a universal satisfaction. I believe I have done that in EA.


Here is the bottom line:

1) For whose sins was Christ punished?

2) If Christ bore only the sins of the elect, in what manner can it be said that Christ’s satisfaction is actually for all people, or even is sufficient for all people, beyond a mere hypothetical sufficiency?

3) If Christ bore only the sins of the elect, how can the indiscriminate offer of pardon be grounded as genuine or sincere on God’s part?


Johnson seems to think the main difference between Calvinists on the extent question has to do with the sufficiency issue: intrinsic for the non-elect; extrinsic for all, with extrinsic sufficiency for all meaning something less than Christ actually died for the sins of all people in the case of many theologians he enlists in this position (Berkhof, Boettner, Kuiper, et al). This is simply historically in error as to the discussions among Calvinists themselves, and it is biblically, theologically, and practically flawed.


As I argued in EA, if Christ bore the sins of all men, then:

1) You have a conditional universal efficacy which defeats the claim that the atonement can only be an unconditional limited efficacy.

2) The atonement, or satisfaction, cannot have inherent in itself its own application. It cannot be “self-applying.” An atonement that procures its own application is incompatible with an unlimited satisfaction for sins. It is also incompatible with any form of genuine sufficiency for all people. The certainty of application of the atonement resides in the Godhead, not the extent of the atonement itself.

3) To deny universal satisfaction for sins, i.e., universal atonement, precludes any form of universal extrinsic sufficiency. Something cannot be sufficient for someone (the non-elect) for whom it was not made. No atonement exists for the non-elect on the limited satisfaction scheme. Christ’s “organic unity” with humanity and the elect’s federal union with Christ cannot alone ground universal sufficiency as Johnson wants to claim (HDFM, 176–77). It is Christ’s organic unity with humanity coupled with the fact that he died for the sins of all humanity (or the vicarious legal union of all in his death) that grounds the universal offer of the gospel. This is what Scripture teaches.

4) Only on this assumption can the genuineness and sincerity of the gospel offer be established.

Here is the dilemma Johnson faces: How can a penal satisfaction not made for a given man’s sins be said to be sufficient for or applicable to that man? This is the dilemma all Calvinists face who wish to deny universal satisfaction for sins, i.e., universal atonement.


Johnson’s “Review” is an inconsistent jumble until he first clarifies the state of the question: for whose sins did Jesus die?, and then properly define the terms. Only then can the issues of “limited atonement,” sufficiency, efficacy, intent, etc. be addressed. My purpose in EA was to demonstrate the myriads of Calvinists in Reformed history who have rejected limited atonement in the sense of a limited vicarious satisfaction. Only secondarily did I address the issue, through their quotations, of whether the atonement secures its own application or not, or purchases faith and all other spiritual blessings.


But these serious problems aside, all is not lost. Johnson is on the right track, it seems. He is historically aware enough of the issues that he sees some of the problems with the strictly limited atonement position, at least in Owen’s expression of it, and the problems entailed by it for the sufficiency question and the well-meant gospel offer. If he can come to the point where he understands the key question is: for whose sins did Christ die?, and, (with Andrew Fuller) jettison the notion of the atonement itself purchasing faith and thus possessing an “inherent efficacy,” the rest will begin to fall into place.


It seems that Johnson wants to affirm an unlimited atonement (extent) in his concept of “organic union,” along with a Calvinistic sense of a designed limited application (intent) only to the elect. He states as much HDFM, 129–43, 171–88, which is sure to raise the ire of most Calvinists who assert a strictly limited atonement/satisfaction. Johnson can get the train back on track if he changes three words in two key sentences in the last chapter of HDFM. First, on p. 186: “How can the atonement be limited in its efficacy extent [where “extent” is understood as a satisfaction for sins] without restricting the content or the scope of the free offer of the gospel?” Second, on p. 188: “We need not worry if we are among God’s elect, for there has been a sufficient provision made for all who believe, . . . .” The atonement, or satisfaction, is a provision for all, not just a provision for all “who believe,” or even only for those who will eventually come to believe. That is what unlimited atonement (satisfaction) is all about. As I said in the final two sentences of EA: “There is a provision of forgiveness for all to whom the gospel comes. There is a provision of forgiveness for all who come to the gospel” (EA, 791).

bottom of page