Indiana Law Review APPELLATE REVIEW OF CIRCUMSTANTIAL EVIDENCE IN INDIANA CRIMINAL CASES The United States Supreme Court has expressly held that the due process clause of the fourteenth amendment protects an accused against conviction except upon proof beyond a rea- sonable doubt. ^ Similarly, under Indiana statutory law, a de- fendant must be acquitted when there is a reasonable doubt of his guilt. ^ This Note will review the tests that the Su- preme Court of Indiana and the Indiana Court of Appeals^ have evolved for insuring convictions only upon proof of guilt beyond a reasonable doubt in cases in which the evidence against the accused is circumstantial in nature. The reasonable doubt standard^ has been described as "a prime instrument for reducing the risk of convictions resting 'In re Winship, 297 U.S. 385, 364 (1970). ^IND. Code §35-1-36-1 (1971). ^Jurisdiction to hear criminal appeals was only recently extended to the Indiana Court of Appeals under the Judicial Amendment to the Indiana Constitution adopted November 2, 1970. Ind. Const, art. 7, § 6. ^Attempts have been made to define the reasonable doubt standard both quantitatively and qualitatively. The definition most often appearing in Indiana decisions is: The rule of law defining proof beyond a reasonable doubt is well settled. It requires the trier of the facts to be so convinced by the evidence that as a prudent man he would feel safe to act upon such conviction in matters of the highest concern and im- portance to his own nearest, dearest and most important interests in circumstances where there was no compulsion or coercion to act at all. Easton v. State, 248 Ind. 338, 344-45, 288 N.E.2d 6, 11 (1967). Accord, Greer v. State, 252 Ind. 20, 30, 245 N.E.2d 158, 163-64 (1969) ; Smith v. State, 243 Ind. 74, 79, 181 N.E.2d 520, 522 (1962); Baker v. State, 236 Ind. 55, 61, 138 N.E.2d 641, 644 (1956). For articles discussing reasonable doubt as a mathematical standard, see Kaplan, Decision Theory and the Factfinding Process, 20 Stan. L. Rev. 1065 (1968); Liddle, Mathematical and Statistical Probability as a Test of Circumstantial Evidence, 19 Case W. Res. L. Rev. 254 (1968) ; Note, Evidence: Admission of Mathematical Probability Statistics Held Erroneous for Want of Demonstration of Validity, 1967 Duke L.J. 665. For discussions of the difficulty of defining the reasonable doubt standard, see C. McCormick, Handbook of the Law op Evidence §341, at 799-800 (2d ed. 1972) [hereinafter cited as McCormick]; 1 H. Underbill, Criminal Evidence § 25, at 43 (6th ed. 1973) [hereinafter cited as Under- bill] ; 9 J. Wigmore, Evidence §2497, at 317 (3d ed. 1940) [hereinafter cited as Wigmore]. 883 INDIANA LAW REVIEW [Vol. 7:883 on factual error/*^ The requirement of such a high standard of proof in criminal cases stems from a fundamental belief that it is far worse to convict an innocent man of a crime than to let a guilty man go free/ A few Indiana cases have limited the ap- plication of the reasonable doubt standard to the trial level/ However, the most frequently cited appellate standard of review is that there must be substantial evidence of probative value to establish every material element of the crime beyond a rea- sonable doubt/ It is apparent from the clandestine nature of criminal ac- tivity that convictions frequently will be wholly dependent upon circumstantial evidence/ In Indiana, a verdict of guilty may be ^In re Winship, 397 U.S. 358, 363 (1970). ^McCoRMiCK § 341, at 798. See also Fletcher, Two Kinds of Legal Rules: A Comparative Study of Burden-of-Persuasion Practices in Criminal Cases, 77 Yale L.J. 880 (1968). Professor Fletcher, in discussing the concern of Western societies that only the guilty be punished, related: This concern prompted Hale to remark in the late 1600*s that five guilty men should be acquitted before one innocent man is convicted. This kind of ratio, expressing toleration for acquitting the guilty has become a stock figure of common law rhetoric; Blackstone raised the ratio to ten to one, and others of libertarian sentiment have favored twenty to one. Like the presumption of innocence, these ratios express a commitment to the dignity of the individual. Id, at 881. ^Coach v. State, 250 Ind. 226, 227, 235 N.E.2d 493, 494 (1968); Green- wait V. State, 246 Ind. 608, 615, 209 N.E.2d 254, 257 (1965) ; Robinson v. State, 243 Ind. 192, 197, 184 N.E.2d 16, 18 (1962). ^Hendrickson v. State, 295 N.E.2d 810, 811 (Ind. 1973) ; McFarland v. State, 295 N.E.2d 809 (Ind. 1973). ^Although the meaning of the term circumstantial evidence is generally well understood. Professor McCormick's illustration of the differing effects of circumstantial and direct evidence upon the thought processes of the trier of fact is helpful: The characterization of evidence as "direct" or "circumstantial" points to the kind of inference which is sought to be drawn from the evidence to the truth of the proposition for which it is offered. If a witness testifies that he saw A stab B with a knife, and this testimony is offered to prove the stabbing, the inference sought is merely from the fact that the witness made the statement, and the assumption that the witnesses are worthy of belief, to the truth of the asserted fact. This is direct evidence. When, however, the evidence is offered also for some further proposition based upon some inference other than merely the inference from assertion to the truth of the fact asserted, then the evidence is circumstantial evidence of this further fact-to-be-inferred. Thus in the case men- tioned if the stabbing were proved and the culprit in doubt, testimony 1974] CIRCUMSTANTIAL EVIDENCE 885 based on circumstantial evidence alone/ ° but the reasonable doubt standard applies regardless of whether the evidence presented at trial is circumstantial or direct/^ In the event that the con- viction is entirely dependent upon circumstantial evidence, a variation of the reasonable doubt standard has been recognized. It has been held that such evidence must exclude every reason- able hypothesis of innocence before it will be sufficient to prove an accused's guilt beyond a reasonable doubt. Although this spe- cial standard to be applied in circumstantial evidence cases is a viable doctrine in Indiana today, it has had a difficult history and is sporadically applied. At the present time, the Indiana courts apply conflicting tests in reviewing the sufficiency of circumstan- tial evidence in criminal cases. The origin, application, conflict, and adequacy of these tests are the subject of this Note. I. Reasonable Hypothesis of Innocence Test The reasonable hypothesis of innocence test requires that cir- cumstantial evidence must not merely be consistent with the hy- pothesis of innocence to be sufficient to sustain a conviction.'^ The first reported case in Indiana mentioning this test is Sumner v. State.^^ In Sumner, the supreme court approved a jury instruction to the effect that circumstantial evidence, to be sufficient, should exclude every supposition inconsistent with guilt. '^ The first reported Indiana case to use the reasonable hypothesis test as a tool in aid of appellate review of sufficiency of the evidence is Schitsler v. State.' ^ In that case, a grocer was struck on the back of the head while locking up his store for the evening. He later died from the blow, and Schusler was con- that A fled from the scene, offered to show his probable guilt, would be direct evidence of the flight but circumstantial evidence of his murderous act. . . . McCoRMiCK § 185, at 435. See generally A. Burrill, A Treatise on the Nature, Principles and Rules of Circumstantial Evidence, Especially THAT OF THE PRESUMPTIVE KiND, IN CRIMINAL CASES 4-8 (2d ed. 1859) ; Underhill § 15, at 23; Wigmore § 25, at 398. ^°Gregory v. State, 286 N.E.2d 666, 668 (Ind. 1972) ; Johnson v. State, 284 N.E.2d 517, 519 (Ind. 1972). ^'Sutherlin v. State, 148 Ind. 695, 698, 48 N.E. 246, 247 (1897); Guyton v. State, 299 N.E,2d 233, 237 (Ind. Ct. App. 1973) ; Ind. Code § 35-1-36-1 (1971). '^Falk V. State, 182 Ind. 317, 321, 106 N.E. 354, 355 (1914). ^^5 Blackf. 579 (Ind. 1841). ^^/d. at 581. ^^29 Ind. 394 (1868). 886 INDIANA LAW REVIEW [Vol. 7:883 victed of first degree murder. The defendant had previously made threats against the deceased grocer, and his footprints, as well as the footprints of others, were found in the alley next to the store. The supreme court, in reversing the conviction, stated that " [t] sustain such a conviction, the facts proved must be sus- ceptible of explanation upon no reasonable hypothesis consistent with the innocence of the person charged.'"* Until 1901, the supreme court continued to reaffirm the reasonable hypothesis of innocence test as a proper test to be applied both at the trial level' ^ and by the supreme court itself in reviewing the suffi- ciency of the evidence.'® In 1901, the supreme court decided the case of Lee v. State"" on facts similar to Schusler but with a surprisingly opposite re- sult. In Lee, a grocer was struck on the back of the head as he was locking his store for the night and robbed of his watch and cash while unconscious. This grocer survived the blow, and Lee was convicted of assault and battery with intent to rob. Lee had been in the victim's grocery a short time before the attack and had asked the time. He was also later observed at closing time near the store carrying a cane. The supreme court affirmed Lee's conviction on the ground that it could only review the suffi- ciency of the evidence when there was an absence of evidence to support a material fact necessary to prove guilt. The court stated % that "[t]he fact that the evidence upon which the judgment of the trial court rests may be said to be weak or unsatisfactory is not available on appeal."^° In Lee, the supreme court expressly overruled an earlier case, Hamilton v. State,^' which applied the reasonable hypothesis of innocence test, and established the con- flicting inferences rule. The court held that when the circum- stantial evidence in a case raises two conflicting inferences, one tending to prove the guilt of the accused and the other favorable to his innocence, it is not within the province of the appellate court to determine which inference should have been accepted by the jury." For a period of sixty-seven years following the '''Id. at 395. '^Wantland v. State, 145 Ind. 38, 40, 43 N.E. 931, 932 (1896); Stout V. State, 90 Ind. 1, 12 (1883). ^sHamilton v. State, 143 Ind. 276, 279, 41 N.E. 588, 589 (1895); Cavender v. State, 126 Ind. 47, 48, 25 N.E. 875 (1890). 'n56 Ind. 541, 60 N.E. 299 (1901). 2°/d at 546, 60 N.E. at 301. 2^142 Ind. 276, 41 N.E. 588 (1895). 2=156 Ind. at 546, 60 N.E. at 301. 1974] CIRCUMSTANTIAL EVIDENCE 887 decision in Lee, the undisputed rule in Indiana was that the rea- sonable hypothesis of innocence test was exclusively for the guidance of juries and trial courts, not for appellate courts re- viewing the sufficiency of the evidence." In 1968, in Manlove v. State,^"^ the Supreme Court of Indiana carefully reviewed the general rules governing appellate review of sufficiency of the evidence and expressly readopted the rea- sonable hypothesis of innocence test for the appellate level. Judge Hunter, writing the opinion for the court, reasoned that when the evidence is wholly circumstantial and fails to exclude every reasonable hypothesis of innocence, such evidence is not suffi- ciently persuasive to allow a reasonable man to find the accused guilty beyond a reasonable doubt.^^ The court specifically dis- approved language in Christen v. State, "^^ which declared that the reasonable hypothesis of innocence test was solely for the guidance of the trier of fact. In reversing Manlove's second degree murder conviction, the court restated a rule often cited in suf- ficiency of the evidence cases: "A verdict based merely upon ^^Dennison v. State, 230 Ind. 353, 357, 103 N.E.2d 443, 444 (1952); Christen v. State, 228 Ind. 30, 38, 89 N.E.2d 445, 448 (1950) ; White v. State, 226 Ind. 309, 311, 79 N.E.2d 771, 772 (1948); Scharillo v. State, 207 Ind. 22, 24, 191 N.E. 76, 77 (1934) ; Wrassman v. State, 191 Ind. 399, 402, 132 N.E. 673 (1921). During the period from 1901 to 1968, the Supreme Court of Indiana did seemingly apply the reasonable hypothesis of innocence test at the appellate level in four cases: Hardesty v. State, 249 Ind. 518, 231 N.E.2d 510 (1967); Slater v. State, 224 Ind. 627, 70 N.E.2d 425 (1947); Osbon V. State, 213 Ind. 413, 13 N.E.2d 233 (1938) ; Falk v. State, 182 Ind. 317, 106 N.E. 354 (1914). The Osbon and Falk decisions relied on Cavender v. State, 126 Ind. 47, 25 N.E. 875 (1890), which was disapproved in Wrassman V. State, 191 Ind. 399, 132 N.E. 673 (1921). Hardesty relied on White v. State, 226 Ind. 300, 79 N.E.2d 771 (1948), which held that the reasonable hypothesis of innocence test was for the guidance of the trier of fact; and Slater relied on Osbon. Although Lee v. State, 156 Ind. 541, 60 N.E. 299 (1901), precluded the use of the reasonable hypothesis of innocence test at the appellate level, the Indiana Supreme Court continued to approve this special test as a jury instruction. Krauss v. State, 225 Ind. 195, 200, 73 N.E.2d 676, 678 (1947) ; Robinson v. State, 188 Ind. 467, 471, 124 N.E. 489, 490 (1919); Dunn v. State, 166 Ind. 694, 702, 78 N.E. 198, 200 (1906). =^250 Ind. 70, 232 N.E.2d 874 (1968). =^/d. at 79, 232 N.E.2d at 879. 2^228 Ind. 30, 38, 89 N.E.2d 445, 448 (1950). Note that Christen had disapproved of Osbon v. State, 213 Ind. 413, 13 N.E.2d 233 (1938); Falk V. State, 182 Ind. 317, 106 N.E. 354 (1914); Hamilton v. State, 142 Ind. 276, 41 N.E. 588 (1895) ; and Cavender v. State, 126 Ind. 47, 25 N.E. 875 (1890). See notes 18, 23 supra. INDIANA LAW REVIEW [VoL 7:883 suspicion, opportunity, probability, conjecture, speculation and unreasonable inferences of guilt . . . cannot be upheld."^^ Although the supreme court unequivocably readopted the reasonable hyp- othesis of innocence test in Manlove, this revitalization of the old rule was somewhat weakened by the court's opinion on petition for rehearing.^® On rehearing, the court denied an assertion by the Attorney General that the court had altered its standard for reviewing the sufficiency of the evidence and reaffirmed the substantial evidence of probative value test.^' Since Manlove, the reasonable hypothesis of innocence test has been followed, ignored, and distinguished. An appeal based on the grounds that there is a reasonable hypothesis of innocence rendering the conviction invalid as based on insufficient evidence may or may not be successful. Of the eighteen cases handed down since 1968 applying the reasonable hypothesis of innocence test, twelve were reversed^^ and six were affirmed.^' Of the twelve that were reversed, seven reversals were based on the ground that the evidence supported only a suspicion of guilt. This fre- quent coupling of the reasonable hypothesis of innocence test with the mere suspicion rule has been used in one decision of the Indiana Court of Appeals as a means of limiting the appli- cation of the reasonable hypothesis test. In Glover v. State,"^^ the court recognized the reasonable hypothesis of innocence test as a test for reviewing circumstantial evidence but limited its application to fact situations involving evidence which provided only a suspicion of guilt or opportunity to commit the crime. "250 Ind. at 84, 232 N.E.2d at 882. 28250 Ind. 85, 235 N.E.2d 62 (1968). 2'The probative value test is discussed in Part III infra. 3°Dunn V. State, 293 N.E.2d 32 (Ind. 1973); Banks v. State, 257 Ind. 530, 276 N.E.2d 155 (1971); Reynolds v. State, 254 Ind, 478, 260 N.E.2d 793 (1970); Seats v. State, 254 Ind. 457, 260 N.E.2d 796 (1970); Sharp v. State, 254 Ind. 435, 260 N.E.2d 593 (1970); Crawford v. State, 251 Ind. 437, 241 N.E.2d 795 (1968) ; Miller v. State, 250 Ind. 338, 236 N.E.2d 173 (1968) ; Carpenter v. State, 307 N.E.2d 109 (Ind. Ct. App. 1974) ; Anderson V. State, 295 N.E.2d 832 (Ind. Ct. App. 1973) ; Martin v. State, 300 N.E.2d 128. (Ind. Ct. App. 1973) ; Wilson v. State, 304 N.E.2d 824 (Ind. Ct. App, 1973); Bradley v. State, 287 N.E.2d 759 (Ind Ct. App. 1972). ^'Gregory v. State, 286 N.E.2d 666 (Ind. 1972); Martin v. State, 250 Ind. 433, 235 N.E.2d 64 (1968); Myers v. State, 251 Ind. 126, 239 N.E.2d 605 (1968) ; Tyler v. State, 292 N.E.2d 630 (Ind. Ct. App. 1973) ; Rogers V. State, 290 N.E.2d 135 (Ind. Ct. App. 1972) ; Williams v. State, 288 N.E.2d 580 (Ind. Ct. App. 1972). "300 N.E.2d 902 (Ind. Ct. App. 1973). 1974] CIRCUMSTANTIAL EVIDENCE 889 Another decision of the court of appeals, Guyton v. State,^^ ex- pressly rejected the view that different standards exist for ap- pellate review of circumstantial and direct evidence. To substan- tiate this view, the court in Guyton cited White v, State,"^^ a 1948 case expressing the view that the reasonable hypothesis rule is exclusively for the guidance of the trier of fact. Although the reasonable hypothesis of innocence test still appears infre- quently in Indiana circumstantial evidence cases, ^^ the most com- mon treatment has been to ignore it and apply either the rea- sonable inference test or the substantial evidence test.'^ The reasonable hypothesis of innocence test has also encount- ered rejection in other jurisdictions. Although this special test for reviewing sufficiency of circumstantial evidence is still ap- plied by many state courts, ^^ it has been rejected by most federal courts. In 1954, the Supreme Court of the United States in Hol- land V. United States^^ held that when a jury is properly instruc- ted on the reasonable doubt standard, an instruction on excluding every reasonable hypothesis of innocence is "confusing and in- correct." In dictum, the Court stated the view that circumstantial evidence is intrinsically no different than direct evidence. Both types of evidence may point to a wholly incorrect result and it is for the jury to weigh the chances that the evidence correctly points to guilt against the possibility of an inaccurate or am- biguous inference.^' Although the Holland case involved only a ^^299 N.E.2d 233 (Ind. Ct. App. 1973). ^^226 Ind. 309, 311, 79 N.E.2d 771, 772 (1948). ^^For the most recent application of the reasonable hypothesis of in- nocence test, see Carpenter v. State, 307 N.E.2d 109 (Ind. Ct. App. 1974). ^*^The substantial evidence test is discussed in Part III infra. ^^Underhill § 17, at 29 n.6. Accord, State v. Fortes, 293 A.2d 506 (R.I. 1972). Contra, State v. Harvill, 106 Ariz. 386, 476 P.2d 841 (1970); People V. Bennett, 515 P.2d 466 (Colo. 1973). ^»348 U.S. 121 (1954). ""^Id. at 140. The Court stated that: Circumstantial evidence in this respect is intrinsically no dif- ferent from testimonial evidence. Admittedly, circumstantial evidence may in some cases point to a wholly incorrect result. Yet this is equally true of testimonal evidence. In both instances, a jury is asked to weigh the chances that the evidence correctly points to guilt against the possibility of inaccuracy or ambiguous inference. In both, the jury must use its experience with people and events in weighing the probabilities. If the jury is convinced beyond a reason- able doubt, we can require no more. 890 INDIANA LAW REVIEW [Vol. 7:883 question of the proper jury instruction, eight of the eleven United States courts of appeals, many of them relying on the above dic- tum from Holland, have expressly rejected the application of a special standard of review for circumstantial evidence/^ The fate of the special test in the Fifth and Eighth Circuits may not yet be settled,"*' and only in the Seventh Circuit is there a possibility that the rule is still applicable at the appellate level /^ II. Reasonable Inference Test From 1901 to 1967, the period that the reasonable hypo- thesis of innocence test was held exclusively for guidance at the trial level, the most frequently cited rule for reviewing suf- ficiency of circumstantial evidence was what will be referred to as the reasonable inference test/^ This test requires the re- ^°United States v. Taylor, 482 F.2d 1376, 1377 (4th Cir. 1973) ; United States V. Currier, 454 F.2d 835, 838 (1st Cir. 1972) ; United States v. Fench, 470 F.2d 1234, 1242 (D.C. Cir. 1972), cert, denied, 410 U.S. 909 (1973); United States v. Hamilton, 457 F.2d 95, 98 (3d Cir. 1972) ; United States v. Henry, 468 F.2d 892, 894 (10th Cir. 1972) ; United States v. Ordones, 469 F.2d 70, 71 (9th Cir. 1972) ; United States v. Glasser, 443 F.2d 994, 1006-07 (2d Cir. 1970), cert, denied, 404 U.S. 854 (1971); United States v. Stroble, 431 F.2d 1273, 1276 (6th Cir. 1970). Also, the reasonable hypothesis of innocence test is clearly no longer a proper jury instruction in a federal criminal trial. United States v. Bartlett, 449 F.2d 700, 705 (8th Cir. 1971), cert, denied, 405 U.S. 932 (1972) ; United States v. Hansbrough, 450 F.2d 328 (5th Cir. 1971) ; United States V. Martine, 442 F.2d 1022, 1023 (10th Cir. 1971) ; United States v. Siragusa, 450 F.2d 592, 596 (2d Cir. 1971); cert, denied, 405 U.S. 974 (1972); United States V. Stubin, 446 F.2d 457 (3d Cir. 1971); Thompson v. United States, 405 F.2d 1106, 1108 (D.C. Cir. 1968); United States v. Whiting, 311 F.2d 191, 198 (4th Cir. 1962). ^'United States v. Irby, 480 F.2d 1101, 1103 (8th Cir. 1973); United States V. Sutherland, 463 F.2d 641, 643 (5th Cir. 1972), cert, denied, 409 U.S. 1178 (1973). But see United States v. Jones, 418 F.2d 818, 826-27 (8th Cir. 1969); United States v. Johnson, 469 F.2d 973, 796 (5th Cir. 1972). See also Note, Circumstantial Evidence in the Federal Courts—The Fifth Circuit and the Holland Case, 14 U. Fla. L. Rev. 89 (1961). ^^Doyle V. United States, 318 F.2d 419, 425 (7th Cir. 1963) ; United States V. McCarthy, 196 F.2d 616, 619 (7th Cir. 1952). ^^Maydwell v. State, 248 Ind. 270, 226 N.E.2d 332 (1967) ; Breedlove v. State, 235 Ind. 429, 134 N.E.2d 226 (1956) ; Shutt v. State, 233 Ind. 169, 117 N.E.2d 892 (1954) ; Petillo v. State, 228 Ind. 97, 89 N.E.2d 623 (1950) ; Stice V. State, 228 Ind. 144, 89 N.E.2d 914 (1950) ; Kestler v. State, 227 Ind. 274, 85 N.E.2d 76 (1949) ; McAdams v. State, 226 Ind. 403, 81 N.E.2d 671 (1948) ; White V. State, 226 Ind. 309, 79 N.E.2d 771 (1948); Mandich v. State, 224 Ind. 209, 66 N.E.2d 69 (1946); Scharillo v. State, 207 Ind. 22, 191 N.E. 76 (1934); Wood v. State, 207 Ind. 235, 192 N.E. 257 (1934); Howard v. State, 193 Ind. 599, 141 N.E. 341 (1923). 1974] CIRCUMSTANTIAL EVIDENCE 891 viewing court to examine the circumstantial evidence, not for the purpose of finding whether or not it is adequate to overcome every reasonable hypothesis of innocence, but to determine if an inference of guilt may be reasonably drawn from the evidence to support the finding of the trier of fact/^ The cases employing this test generally do not set forth the criteria the reviewing court is to use in determining if the inference of guilt is rea- sonable. For example, in Kestler v, State,"^^ the Supreme Court of Indiana admitted that there was no evidence to support de- fendant's conviction of the first degree murder of his wife but cursorily found the circumstantial evidence sufficient to author- ize the jury to find the defendant guilty of murder in the second degree. The evidence reviewed by the court disclosed that on the night of the shooting, defendant and his brother had been out to dinner and had been drinking. Upon returning home, the defendant put his arm around his wife and they went into the bedroom while the defendant's brother went downstairs to see his nieces. When he was returning upstairs, he heard his brother say, "Mable watch that gun, its loaded," and then there as a shot. The defendant claimed the shooting was accidental and that the gun went off when his wife was handing him the gun. There was no evidence in the record that defendant and his wife had argued that evening. The brother testified, however, that before the shot he could see his brother through the open bedroom door and that he could not see his brother's wife. Also, the gun had two safety devices and a trigger. Admitting that the evidence raised con- flicting inferences of guilt and innocence, the court relied upon the conflicting inferences rule and held that there was evidence from which the jury could have inferred that the defendant pur- posely and maliciously killed his wife. In reversing convictions when applying the reasonable in- ference test, the supreme court has stated that an inference can- not be based upon evidence which is uncertain or speculative.^^ For example, in Howard v, State,^^ the defendant and his com- panions accepted a ride with a stranger from Cincinnati. An of- ficer stopped the car, found whiskey in the back seat, and arrest- ed everyone for possession of liquor with intent to sell. The court, on appeal, reversed defendant's conviction and held that a con- ^^Dowty V. state, 203 Ind. 228, 235, 179 N.E. 720, 723 (1932). ^^227 Ind. 274, 85 N.E.2d 76 (dissenting: opinions). ^^Shutt V. State, 233 Ind. 169, 174, 117 N.E.2d 892, 894 (1954). ^^193 Ind. 599, 141 N.E. 341 (1923). 892 INDIANA LAW REVIEW [Vol. 7:883 viction based on circumstantial evidence alone cannot rest upon "improbable, imaginary and unnatural inferences.""*® Even though the reasonable inference rule is often directly contrary to the reasonable hypothesis of innocence test in terms of result, it is frequently applied today, often in combination with the substantial evidence rule. Since the Manlove decision in 1968, of the tv^enty cases applying the reasonable inference test,^' only three resulted in a reversal on the ground that the evidence proved only a suspicion of guilt.^° Four of the seventeen affirmed convictions engendered dissents,^ ^ and only eight of the seven- teen affirmed convictions would clearly have been affirmed under the reasonable hypothesis of innocence test." In Vaughn v. ^^Id, at 604, 141 N.E. at 343. It should be noted that in Indiana inferences based upon other inferences are not necessarily considered speculative or uncertain. The Indiana cases recognize an exception to the rule that one inference cannot be based upon another when the first inference is based on firmly established circumstan- tial evidence and becomes, in the mind of the trier of fact, a proven fact. Walker v. State, 250 Ind. 649, 653, 238 N.E.2d 466, 468-69 (1968) ; Brown v. State, 219 Ind. 21, 23-24, 36 N.E.2d 759, 160 (1941); Hinshaw v. State, 147 Ind. 334, 363, 47 N.E. 157, 166 (1897). ^'Boys V. State, 304 N.E.2d 789 (Ind. 1973); McAfee v. State, 291 N.E.2d 554 (Ind. 1973) ; Gunn v. State, 281 N.E.2d 484 (Ind. 1972) ; High- tower V. State, 256 Ind. 344, 269 N.E.2d 10 (1971); Vaughn v. State, 255 Ind. 678, 266 N.E.2d 219 (1971); Stuck v. State, 255 Ind. 350, 264 N.E.2d 611 (1970); Glover v. State, 253 Ind. 536, 255 N.E.2d 657 (1970); Kin- dred v. State, 254 Ind. 105, 257 N.E.2d 818 (1970); Kindred v. State, 254 Ind. 73, 257 N.E.2d 667 (1970); Hobbs v. State, 253 Ind. 195, 252 N.E.2d 498 (1969); Farno v. State, 308 N.E.2d 725 (Ind. Ct. App. 1974); Klebs V. State, 305 N.E.2d 789 (Ind. Ct. App. 1974) ; Cornman v. State, 294 N.E.2d 812 (Ind. Ct. App. 1973) ; Glover v. State, 300 N.E.2d 902 (Ind. Ct. App. 1973); Greeley v. State, 301 N.E.2d 853 (Ind. Ct. App. 1973); Guyton V. State, 299 N.E.2d 233 (Ind. Ct. App. 1973) ; Haynes v. State, 293 N.E.2d 204 (Ind. Ct. App. 1973) ; Sarten v. State, 303 N.E.2d 300 (Ind. Ct. App. 1973) ; Walker v. State, 293 N.E.2d 35 (Ind. Ct. App. 1973) ; Shank v. State, 289 N.E.2d 315 (Ind. Ct. App. 1972). ^°Glover v. State, 253 Ind. 536, 255 N.E.2d 657 (1970) ; Kindred v. State, 254 Ind. 105, 257 N.E,2d 667 (1970); Greeley v. State, 301 N.E.2d 853 (Ind. Ct. App. 1973). ^^McAfee v. State, 291 N.E.2d 554 (Ind. 1973); Vaughn v. State, 255 Ind. 678, 266 N.E.2d 219 (1971) ; Kindred v. State, 254 Ind. 73, 257 N.E.2d 667 (1970); Hobbs v. State, 253 Ind. 195, 252 N.E.2d 498 (1969). "Boys v. State, 304 N.E.2d 789 (Ind. 1973) ; Hightower v. State, 254 Ind. 344, 269 N.E.2d 10 (1971); Stuck v. State, 255 Ind. 350, 264 N.E.2d 611 (1970) ; Klebs v. State, 305 N.E.2d 781 (Ind. Ct. App. 1974) ; Cornman V. State, 294 N.E.2d 812 (Ind. Ct. App. 1973) ; Haynes v. State, 293 N.E.2d 1974] CIRCUMSTANTIAL EVIDENCE 893 State,^^ one of the nine cases in which application of the rea- sonable hypothesis of innocence test would probably have required reversal, the supreme court upheld a conviction of theft. The only evidence linking defendant to the theft of two saddles was his possession of the saddles two days after the theft. Even disregarding Vaughn's explanation that he bought the saddles from a stranger in a bar, the reasonable hypothesis of innocence test would seem to have required reversal of Vaughn's conviction. III. Substantial Evidence Test The substantial evidence test, in addition to being applied by the Indiana appellate courts in reviewing the sufficiency of direct evidence,^"^ is frequently applied in circumstantial evidence cases. The test appears in many variations, but it generally re- quires that the reviewing court look only to the evidence most favorable to the State and the reasonable inferences to be drawn therefrom to determine if there is substantial evidence of pro- bative value sufficient to establish every material element of the crime. In conjunction with this test, the court generally states the often repeated rule that the reviewing court will not weigh the evidence nor determine the credibility of the witnesses.^^ As demonstrated by Schooler v. State,^^ the appellate courts rarely define what is considered substantial evidence. In Schooler, the defendant was convicted of shoplifting several items from a men's clothing store. Several shirts and a man's suit were missing soon after defendant and her companion left the store. A parking lot attendant observed the defendant and her companion come from the men's store and go to their car. They then left and did not come back for several hours. In the meantime, the police observed the car and noticed a brown paper sack in the car con- taining what looked like a man's suit and some shirts. The police arrested defendant and her companion when they returned to the car. At that time, both defendant and her companion were carrying straw bags which contained nothing except a head scarf. 204 (Ind. Ct. App. 1973) ; Sarten v. State, 303 N.E.2d 300 (Ind. Ct. App. 1973); Walker v. State, 293 N.E.2d 35 (Ind. Ct. App. 1973). "255 Ind. 678, 266 N.E.2d 219 (1971). ^^Dickens v. State, 295 N.E.2d 613, 617 (Ind. 1973); McFarland v. State, 295 N.E.2d 809 (Ind. 1973) ; Jones v. State, 293 N.E.2d 545, 550 (Ind. Ct. App. 1973). "Ledcke v. State, 296 N.E.2d 412 (Ind. 1973). ^*247 Ind. 624, 218 N.E.2d 135 (1966) (Jackson, J., dissenting). 894 INDIANA LAW REVIEW [Vol. 7:883 The bag in the car was found to contain the merchandise from the men's store. In affirming defendant's conviction, the court stated that "a conviction can rest entirely upon circumstantial evidence if there is substantial evidence of probative value to support an inference of guilt."^^ Cases discussing the meaning of substantial evidence have described it as more than "seeming and imaginary."^® In Easton V. State,^^ the supreme court explained that the test of substantial evidence v^as whether the trier of fact could reasonably find the fact at issue beyond a reasonable doubt. With substantial evidence so defined, the reasonable inference test and substantial evidence test are obviously closely related, and they are often cited in combination.^° The conflicting inference rule is also fre- quently cited in cases applying the substantial evidence test.*' This rule, however, is completely incompatible with the reasonable hypothesis of innocence test since it requires the court on appeal ^Ud. at 627, 218 N.E.2d at 136. ^^Powell V. State, 250 Ind. 663, 665, 237 N.E.2d 95, 96 (1968) ; Finch V. State, 249 Ind. 122, 126, 231 N.E.2d 45, 47 (1967). ^'248 Ind. 338, 228 N.E.2d 6 (1967). See also Baker v. State, 236 Ind. 55, 62, 138 N.E.2d 641, 645, quoting from State v. Gregory, 339 Mo. 133, 143 96 S.W.2d 47, 52 (1936), as follows: Now since the test of substantial evidence is whether a jury reason- ably could find the issue thereon, the result must depend in some measure upon the degree of persuasion required. In a criminal case liberty and sometimes life are involved, and there cannot be a con- viction except upon a finding of guilt beyond a reasonable doubt. Necessarily, therefore, it becomes the duty of an appellate court as a matter of law to decide whether the evidence was sufficient to induce a belief of the defendant's guilt beyond a reasonable doubt in the minds of jurors of average reason and intelligence. And in resolving that question the court undoubtedly can pass on the credi- bility of the testimony to the extent of determining whether it was substantial in the sense above explained. In no other way can the rights of the defendant be protected. It would be an incongruous situation if the court were compelled to let a conviction stand as being supported by evidence warranting a verdict of guilt beyond a reasonable doubt, when for any reason made manifest on the record the court is convinced the evidence reasonably could not support a conviction. 6°Phillips V. State, 295 N.E.2d 592, 593-94 (Ind. 1973) ; Tibbs v. State, 255 Ind. 309, 311-12, 263 N.E.2d 728, 730 (1970); Christen v. State, 228 Ind. 30, 39, 89 N.E.2d 445, 448 (1950) ; Klebs v. State, 305 N.E.2d 781, 784 (Ind. Ct. App. 1974). ^^ Young V. State, 257 Ind. 173, 177, 273 N.E.2d 285, 287 (1971); Tait V. State, 244 Ind. 35, 44, 188 N.E.2d 537, 542 (1963) ; Byrd v. State, 243 Ind. 452, 458, 185 N.E.2d 422, 425 (1962). 1974] CIRCUMSTANTIAL EVIDENCE 895 to accept the inference of guilt even though there is an equally compelling inference of innocence. As in the case of the reasonable inference test, the sub- stantial evidence test is less likely to result in reversal than the reasonable hypothesis of innocence test. Of eleven circumstantial evidence cases using the substantial evidence test since 1968/^ five resulted in reversals on the ground that the evidence sup- ported only a suspicion of guilt." Of the remaining six affirmed convictions, at least one, Ledcke v. State,^^ v^ould probably have resulted in reversal had the reasonable hypothesis of innocence test been applied. In Ledcke, the defendant wsls convicted of pos- session of marijuana. The only evidence sustaining a finding of constructive possession w^as Ledcke's presence in an apartment where marijuana v^as being manufactured. Ledcke v^as arrested while trying to leave the apartment by the back door during a raid. Ledcke was not a tenant of the apartment nor was mari- juana found on his person. As Justice DeBruler pointed out in his dissent, the defendant could have been visiting the apartment for a "multitude of other and completely lawful activities."*^ Application of the reasonable hypothesis of innocence test cer- tainly would have required a reversal since reasonable hypotheses of innocence did abound. IV. Appellate Review and the Reasonable Doubt Standard From the foregoing discussion, it can be seen that a convicted criminal defendant in Indiana appealing his conviction on the grounds of insufficient circumstantial evidence cannot know with certainty which test of review the Indiana appellate courts will apply. With a view towards insuring convictions only upon evi- ^=^Baker v. State, 298 N.E.2d 445 (Ind. 1973); Ledcke v. State, 296 N.E.2d 412 (Ind. 1973) ; Tom v. State, 302 N.E.2d 494 (Ind. 1973) ; Taylor v. State, 284 N.E.2d 775 (1972) ; Young v. State, 257 Ind. 173, 273 N.E.2d 285 (1971); Tibbs v. State, 255 Ind. 309, 263 N.E.2d 728 (1970); Vuncannon v. State, 254 Ind. 206, 258 N.E.2d 639 (1970) ; Ellis v. State, 252 Ind. 472; 250 N.E.2d 364 (1969); Listen v. State, 252 Ind. 513, 250 N.E.2d 739 (1969); Durham v. State, 250 Ind. 555, 238 N.E.2d 9 (1968); Pfeifer v. State, 283 N.E.2d 567 (Ind. Ct. App. 1972). "Tom V. State, 302 N.E.2d 494 (Ind. 1973); Tibbs v. State, 255 Ind. 309, 263 N.E.2d 728 (1970) (reversing in part) ; Vuncannon v. State, 254 Ind. 206, 258 N.E.2d 639 (1970) ; Liston v. State, 252 Ind. 513, 250 N.E.2d 739 (1969) (reversing in part); Durham v. State, 250 Ind. 555, 238 N.E.2d 9 (1968). *^296 N.E.2d 412 (Ind. 1973). *^/rf. at 422. 896 INDIANA LAW REVIEW [Vol. 7:883 dence sufficient beyond a reasonable doubt, the reasonable hyp- othesis of innocence test would seem to be the better test to achieve this purpose. Courts and legal writers in rejecting the reasonable hypothesis of innocence test frequently assert that a special standard of review for circumstantial evidence is unnec- essary. It is contended that since both direct and circumstantial evidence can lead to an incorrect result, the substantial evidence test is sufficient for reviewing both types of evidence.^^ It is not to be refuted that direct evidence, which is dependant upon truthful testimony, can lead to erroneous convictions. However, direct evidence is entirely dependant upon credibility of the wit- nesses, and the appellate courts have valid reasons for refusing to review this type of evidence. The reviewing court must accept the trier of fact's apparent assessment of testimonial evidence since it is obviously difficult to ascertain truthfulness without an opportunity to observe the witnesses.^^ To illustrate, if an eyewitness to a crime positively identifies the defendant as the perpetrator and the jury returns a verdict of guilty, the jury obviously believes the witness. The court on review should reverse the conviction only if the record discloses conclusive evidence that the witness was lying—for example, if the defendant was con- clusively shown to have been in prison at the time of the alleged crime. However, in a circumstantial evidence case, the testimonial evidence will not directly establish guilt. The jury must accept or reject the offered testimonial evidence and draw an inference ^^Holland v. United States, 348 U.S. 121, 140 (1954); WiGMORE §26, at 401; Note, Sufficiency of Circumstantial Evidence in a Criminal Case^ 55 CoLUM. L. Rev. 549 (1955); cf. Underhill §16, at 24; 2 C. Wright, Federal Practice and Procedure §467 (1969). *^The Supreme Court of Indiana in Cox v. State, 49 Ind. 568, 571 (1875), described the importance of allowing the trier of fact to judge the credibility of the witnesses as follows: In considering the weight of testimony by this court, it must be re- membered that such evidence comes before us merely in written words; while the court and jury trying the cause have it from the living voice, with whatever peculiar accent, emphasis, or intonation it may have; and that they see the witness, his countenance, looks, expression of face, manner, readiness, or reluctance, and the many nameless indices of truth or falsehood, which it is impossible to put in words. A statement of facts in words, though testified to by dif- ferent witnesses, of various degrees of credibility, comes to us with the same weight, while to the court and jury their weight would be, in some instances, the full import of the words, and in others scarcely worth consideration. Hence it is that the credibility of a witness is a question solely for the jury, they being the triers of fact; and the presumption in this court must be that they understood their duty, ?ind performed it. 1974] CIRCUMSTANTIAL EVIDENCE 897 of guilt from the evidence accepted as credible. It is submitted that the reviewing court is equally capable of determining the inference to be drawn from the evidence most favorable to the State and that if such evidence supports a reasonable hypothesis of innocence, the inference of guilt is not reasonable in light of the standard of proof of guilt beyond a reasonable doubt.''^ A distinction should be made between relitigating the facts on appeal and weighing the sufficiency of the inference of guilt drawn from those facts. The policy behind the rule that courts of review do not weigh the evidence lies in the guarantee of a jury trial under the Indiana Constitution.^' Regardless of how the appellate courts express the test for reviewing sufficiency of circumstantial evidence, however, it is apparent that some weighing of the evidence must take place. Were the courts of review to refuse to weigh the evidence in criminal appeals based on sufficiency of the evidence, convictions could only be overturned upon a showing of no evidence to support the verdict. This would be a return to the scintilla rule^° and a clear rejection of the appellate courts' role in insuring convictions only upon proof beyond a reasonable doubt. It is conceded that each of the tests reviewed in this Note depends upon an elusive standard. It would be impossible for the courts to promulgate a comprehensive definition of substantial evidence, reasonable inference, or reasonable hypothesis of in- nocence. Every criminal fact situation reviewed in a sufficiency of the evidence appeal is different. However, the use of the rea- sonable hypothesis of innocence test rather than the other two tests results in a shift in the emphasis of the court's review. By using the reasonable hypothesis of innocence test, the reviewing court determines the reasonableness of a conviction in light of all the reasonable inferences to be drawn from the evidence most favorable to the State. By using the reasonable inference test or the substantial evidence test, the court's review is limited to only those inferences supporting the verdict. It is submitted that the latter type of review cannot insure a conviction upon proof *«Manlove v. State, 250 Ind. 70, 79, 232 N.E.2d 874, 879 (1968). *9Ind. Const, art. 1, § 19. ^°Although the scintilla rule has been rejected in Indiana in the case of Easton v. State, 248 Ind. 338, 343, 228 N.E.2d 6, 10 (1967), earlier cases allowed reversal of a conviction only upon a showing of no evidence to sup- port the verdict. Krstovich v. State, 186 Ind. 556, 557, 117 N.E. 209 (1917) ; Lee V. State, 156 Ind. 541, 546, 60 N.E. 299, 301 (1901); McCarty v. State*, 127 Ind. 223, 225, 26 N.E. 665, 666 (1891). 898 INDIANA LAW REVIEW [Vol. 7:883 beyond a reasonable doubt in a case entirely dependent upon cir- cumstantial evidence. The seemingly high reversal rate resulting from the use of the reasonable hypothesis of innocence test at the appellate level is not unexpected. Plea bargaining and resultant guilty pleas ac- count for the disposition of many criminal charges based on substantial evidence.^ ^ Criminal cases that go to trial are of basic- ally three types: cases in which the defendant is innocent, cases in which the defendant is guilty but the evidence is so weak that he elects not to plea bargain, and cases in which the evidence against the defendant is so strong that the State refuses to plea bargain. On appeal, convictions based on weak evidence, regardless of the defendant's guilt or innocence, would be reversed if the appellate court applied the reasonable hypothesis of innocence test. Such convictions might not be reversed if the court applied the reasonable inference test or the substantial evidence test. Convictions based on substantial evidence would be affirmed re- gardless of the test applied. Until society determines that it would be better served by convicting innocent m.en to insure that all criminals are punished, the reasonable hypothesis of innocence test is the best test for insuring convictions upon proof beyond a reasonable doubt. ! V. Conclusion The appellate court's first consideration in reviewing a suf- ficiency of the evidence question should be to determine whether the record discloses any evidence to support the conviction. When no evidence is disclosed on review, none of the tests discussed above need be applied, and the reviewing court must reverse as a matter of law. In such a case, to affirm would be a denial of due process. ^^ A finding of no evidence may arise when there is a complete absence of a vital fact, when the court is barred by rules of law or of evidence from giving weight to the only evidence ''^Indiana county prosecutors estimated the number of felonies plea- bargained in their counties during 1973 in a prosecutors' survey compiled by the Criminal Justice Planning Agency. There are ninety-two counties in In- diana and eight-one county prosecutors representing eighty-five counties reported. One reporting county, Sullivan, gave no estimate and three counties reported the exact number of felonies plea-bargained: Clinton County—99%, Hamilton County—53.5%, and Putnam County—88%. Of the remaining seventy-seven prosecutor estimates, the average number of cases plea bar- gained was 63%, the median was 65%, and the mode was 50%. 72Garner v. Louisiana, 368 U.S. 157, 163 (1961); Thompson v. City of Louisville, 362 U.S. 199, 206 (1960). 19743 CIRCUMSTANTIAL EVIDENCE 899 offered to prove a vital fact, and v^hen the evidence establishes conclusively the opposite of the vital fact/^ When sufficiency of curcumstantial evidence is at issue, the court should accept the testimonial evidence most favorable to the State, but it should not limit its consideration to only those inferences therefrom supporting the conclusion of the trier of fact. In reviewing all of the reasonable inferences to be drawn from the evidence most favorable to the State, if the evidence discloses a reasonable hypothesis of innocence, the conviction should be reversed. The conflicting inferences rule should be limited to apply only to conflicting direct evidence. If the cir- cumstantial evidence supports a reasonable inference of innocence as well as guilt, the conviction should be reversed as unsupported by proof beyond a reasonable doubt. Linda A. Hammel ''Xalvert, *Wo Evide7ice" and ''Insufficient Evidence" Points of Error, 38 Texas L. Rev. 361, 362 (1960).