Dooley, J.
Defendant Bernard Carter appeals his conviction by jury of aggravated sexual assault, in violation of 13 V.S.A § 3253(a), raising three claims of error: (1) the trial court impermissibly excluded evidence that substantiated and explained his reasoning for leaving Vermont after he was charged with the crime; (2) the court erroneously allowed one of the State’s witnesses to corroborate the victim’s account of events through prior consistent statements; and (3) his Fifth Amendment rights under the United States Constitution were violated when the trial court allowed a state police officer to testify that defendant refused to discuss the charges with him. We conclude that defendant’s first two claims of error are valid, but because these errors were harmless beyond a reasonable doubt, we affirm.
I.
Defendant and the victim had separate apartments in the same complex in Newport, Vermont and were formerly romantically involved. During the evening of October 24,1992, defendant entered the victim’s apartment and accused her of giving him a sexually transmitted disease. Later that night, he entered the apartment again, accompanied by his nephew, and refused to leave. According to the victim, defendant held a knife to her throat and threatened to forcibly rape her and to kill her. Thereafter, according to the victim, he forcibly sexually assaulted her. According to defendant, who also testified, he never threatened the victim, and they had consensual intercourse.
Defendant was charged with aggravated sexual assault, unlawful mischief, and petty larceny. He was eventually apprehended in Arizona in February 1993, and stood trial in May 1994. Although he
was acquitted of the unlawful mischief and petty larceny charges,
he was convicted of aggravated sexual assault. The allegation on which he was found guilty specified that he sexually assaulted the victim, that at the time of the assault, he threatened to cause serious bodily harm to her, and that she reasonably believed he had the present ability to carry out the threat.
II.
The first issue on appeal relates to the State’s evidence that defendant fled the state to avoid prosecution. The victim gave a statement to the police on October 26, 1992, and they went to defendant’s apartment where he was seen running from the premises. Along with his nephew and girlfriend, he traveled from state to state thereafter until he was finally apprehended in Arizona. During his flight, defendant threatened to kill the victim and threatened to kill both his nephew and his girlfriend if they testified against him.
The State called on both the nephew and the girlfriend to show defendant’s flight. The issue of an alternative explanation for the flight, consistent with innocence, first surfaced with the nephew, who wanted to testify that he fled in response to an inaccurate newspaper story. He offered a copy of the story, which said that both he and defendant had committed the crime, that they had used a knife to rape the victim, and that the maximum punishment faced was life imprisonment. The court refused to allow the newspaper story to be introduced and also prohibited the nephew from describing it.
The issue then arose with respect to defendant, who made a claim similar to that of the nephew. When defense counsel asked defendant what he had heard about why the police were looking for him, the trial judge intervened and eventually ruled defendant could not testify to “any information gained from newspapers or the like” or make any statements about defendant’s belief about the maximum punishment for the offense. Defendant did testify that he understood the police were looking for him for a knife-point rape, that he did not commit such a rape, and that he fled out of fear. Oh cross-examination, he
testified that he learned about the charge from the newspaper, which had a front-page article with his picture and the statement that he and his nephew were wanted for a knife-point rape.
We have held that the State may introduce evidence of flight by a criminal defendant to show consciousness of guilt. At the same time, we have questioned the probative value of such evidence. See
State v. Pelican,
160 Vt. 536, 542, 632 A.2d 24, 28 (1993);
State v. Giroux,
151 Vt. 361, 366, 561 A.2d 403, 406 (1989);
State v. Unwin,
139 Vt. 186, 193, 424 A.2d 251, 255 (1980). Recently, in
State v. Perrillo,
162 Vt. 566, 570, 649 A.2d 1031, 1034 (1994), we ruled that if evidence of flight is admitted “the defendant should in fairness be afforded the opportunity to explain why his absence was consistent with his innocence.” We reversed the conviction in
Perrillo
because defendant was prohibited from showing his reason for leaving the state. He testified that, although he had received a plea bargain offer that required no jail time, he left the state to earn the money to hire a private lawyer, not to escape. He wanted to testify to the proposed plea agreement to show that he did not fear prosecution and he turned it down because he wanted to, establish his innocence. We held that the exclusion of the plea bargain evidence denied defendant the opportunity to refute the inference that his “flight” was motivated by consciousness of guilt.
Id.
at 570, 649 A.2d at 1034.
We agree with defendant that the court’s exclusionary ruling was error under
Perrillo.
Apparently, the court was concerned about the jury being exposed to improper pretrial publicity and knowing the -maximum punishment for the crime. To the extent the information would prejudice defendant, however, he waived any objection by offering the newspaper headline and article. To the extent it would prejudice the State, the prosecution necessarily accepted that prejudice by offering the flight evidence. The alternative, if any, was to exclude the evidence of flight, not to prevent defendant from offering a plausible explanation for his actions. It is inconsistent with our holding in
Perrillo
for the State to insist that evidence of flight be admitted and then to rely on other considerations, such as those contained in V.R.E. 403, to prevent defendant from providing a full explanation of his actions.
III.
Defendant next claims that the trial court erred by permitting the victim’s sister’s testimony as evidence of a prior consistent statement by the victim. The victim disclosed the sexual assault to her sister,
who was also her roommate, during the day following the assault. When asked how the victim described the attack, the sister responded, “She told me that he forced her to have sex with him. She also said that he made her have sex with him.” When asked to describe other details of what happened, the sister testified, “She told me that he had her pushed against the wall, against the stove, had a knife to her throat. Spit in her face. . . . You know, name calling, threatening her and her kids.” Defendant objected to this testimony because it was hearsay and was not authorized by the rule on prior consistent statements, V.R.E. 801(d)(1)(B).
Hearsay is defined as an out-of-court statement offered in evidence to prove the truth of the matter asserted. V.R.E. 801(c). Hearsay has been traditionally rejected as substantive evidence because the declarant was not under oath and was not subject to cross-examination at the time the statement was made. See generally 5 Wigmore on Evidence §§ 1362, 1367 (Chadbourn rev. 1974). Testimony regarding an out-of-court statement that is used not to prove the truth of the matter asserted, but is used instead to rebut an allegation of recent fabrication or improper motive or influence, is not hearsay. V.R.E. 801(d)(1)(B). When both the source of the statement is before the jury for scrutiny and the person who heard the prior consistent statement is in court to testify, the statement is as reliable as any other testimonial evidence, and is not regarded as hearsay. See
State v. Roy,
140 Vt. 219, 227, 436 A.2d 1090, 1093 (1981); McCormick on Evidence § 251, at 604 (2d ed. 1972); Reporter’s Notes, V.R.E. 801.
The dispute over the admission of the prior consistent statement here involves timing. In
State v. Roy,
decided before the promulgation of the Vermont Rules of Evidence, we adopted the content of Federal Rule of Evidence 801(d)(1)(B), which is identical to V.R.E. 801(d)(1)(B). We held that the rule has three requirements:
(1) the prior consistent statement corroborates the witness’s in-court testimony; (2) the party offering the prior consistent statement establishes that the statement is being offered to rebut an express or implied charge against the witness of recent fabrication or improper influence or motive; and (3) the statement is shown to have been made prior to the time that the supposed motive to falsify arose.
140 Vt. at 227, 436 A.2d at 1094. Defendant argues that the third element of
Roy
is not met because the motive to falsify was present from the beginning.
The State answers by relying on
State v. Robinson,
158 Vt. 286, 298, 611 A.2d 852, 858 (1992), in which a similar argument was made with respect to a prior consistent statement made by a victim of child sexual abuse to a police officer. Emphasizing that the trial court has discretion in determining whether to admit a prior consistent statement and that the child’s motive to fabricate as claimed by the defendant was “highly speculative,” we held it within the trial court’s discretion to admit the evidence. The State argues that timing is unimportant here because defendant claimed that the victim embellished the story over time, and as a result, admission was within the trial judge’s discretion.
The different results of Roy and
Robinson
reflect reliance on decisions from different circuit courts on an issue on which they were split. Thus, the Second Circuit, the decision of which is relied upon in
Roy,
has required that the prior consistent statement be made before the declarant had a motive to fabricate the story. See 4 J. Weinstein & M. Berger, Weinstein’s Evidence ¶ 801(d)(l)(B)[01], at 801-196 (1995) (citing cases). The Ninth Circuit, the decision of which is relied upon in
Robinson,
has not so required. See
id.
at 801-197, 198. The intra-circuit disagreement has now been resolved by the Supreme Court in
Tome v. United States,
513 U.S. 150, 115 S. Ct. 696 (1995), in favor of the Second Circuit approach. The holding of
Tome
is that Rule 801(d)(1)(B) “permits the introduction of a declarant’s consistent out-of-court statements to rebut a charge of recent fabrication or improper influence or motive only when those statements were made before the charged recent fabrication or improper influence or motive.”
Id.
at 167, 115 S. Ct. at 705. In reaching its decision, the Court acknowledged that in some cases “it may be difficult to ascertain when a particular fabrication, influence, or motive arose,” but concluded that the timing could be determined.
Id.
at 165-66,115 S. Ct. at 705. The main reasoning of the Court was that “[a] consistent statement that predates the motive is a square rebuttal of the charge that the testimony was contrived as a consequence of that motive. By contrast, prior consistent statements carry little rebuttal force when most other types of impeachment are involved.”
Id.
at 158,115 S. Ct. at 701. Based on the wording of the rule, we agree with the analysis in
Tome.
Even if there is a temporal requirement, the State argues that the trial court could find it was met because defense counsel argued to the jury that the victim’s story changed over time. The defense theory, which was developed through cross-examination of the victim, was that her initial disclosures to the sister were incomplete and did not include the allegation that defendant had held a knife to her or forced her to have sex. The sister’s testimony, to which defendant objected, however, included the full disclosures the victim eventually made. Thus, the prior consistent statement came after any claimed “recent fabrication.” There was no claim that the victim’s motive changed after the statements were made to the sister; in fact, the defense claimed her motive to lie existed from the beginning. Thus, we cannot hold that the timing element of Rule 801(d)(1)(B) was met. Admission of the prior consistent statement through the testimony of the sister was error.
IV.
Defendant’s third claim is that introduction of his failure to deny the crime to the police violated his self-incrimination right under the United States Constitution. The issue arose with respect to the State’s rebuttal witness, a police officer who testified that defendant called him a few days after the alleged assault. During the telephone conversation, the officer asked defendant to tell his side of the story and defendant refused. Although the trial court did let in this evidence, it instructed the jury that a citizen was not obligated to speak to a law enforcement officer and that claiming the right against self-incrimination should not be considered against the person.
In this case, defendant was not under arrest when he refused to answer the question. The United States Supreme Court has held that silence that precedes an arrest, or comes after an arrest but before
Miranda
warnings, can be introduced and commented upon to impeach defendant’s testimony at trial. See
Fletcher v. Weir,
455 U.S. 603, 606-07 (1982);
Jenkins v. Anderson,
447 U.S. 231, 239 (1980); see generally
State v. Percy,
149 Vt. 623, 628, 548 A.2d 408, 410 (1988). Accordingly, there was no violation of defendant’s federal self-incrimination right in allowing this impeachment evidence.
V.
Although we have concluded that the trial court erred by excluding rebuttal evidence to the State’s evidence of defendant’s flight and by
admitting the victim’s prior consistent statements, the State argues that these errors are not grounds for reversal of defendant’s conviction because they were harmless. This argument requires us to examine the standard for determining whether nonconstitutional errors are harmless in criminal cases, an issue that we have avoided in recent years.
The basic premise that errors must have some relation to the outcome of a criminal case, or some independent right of the defendant, before appellate intervention is warranted is stated in both the criminal and evidence rules. Both require that a substantial right of the party be affected by the error. See V.R.Cr.E 52(a); V.R.E. 103(a). This brief and cryptic phrasing is, however, insufficient to establish a standard for appellate review. It is the more modern equivalent of our often-stated proposition that where a defendant cannot show prejudice from an error, it is harmless. See, e.g.,
State v. Oakes,
129 Vt. 241, 255-56, 276 A.2d 18, 27 (1971).
Because “[evidence tending to inculpate the defendant always carries with it some prejudice,”
State v. Kelley,
163 Vt. 325, 329, 664 A.2d 708, 711 (1995), this statement establishes only that some relationship between the error and the outcome, or a substantial right of defendant, is necessary, without defining the needed relationship.
Our early decisions were clearer on the necessary relationship. The issue is explored in some detail in
State v. Meader,
54 Vt. 126, 130-32 (1881), where the State put in evidence with the understanding it would later be tied to the defendant but failed to make that tie. The State argued, in essence, that the evidence was harmless, but the Court rejected that argument using the following standard: “The question in all cases, is not whether the court, if trying the case, would disregard the obnoxious evidence, but whether the court is
assured
that the jury has done so.”
Id.
at 132.
Up until the late 1960s, it was rare for this Court to affirm a criminal conviction by holding that an error was harmless. Generally,
it occurred in cases where there was no possibility that the error could have affected the conviction. See, e.g.,
State v. Storrs,
105 Vt. 180, 186, 163 A. 560, 563 (1933) (error harmless where evidence excluded on hearsay ground is later testified to by person who made statement);
State v. Roby,
83 Vt. 121, 125, 74 A. 638, 639 (1909) (irrelevant evidence, improperly admitted, “was so entirely without significance” that error was harmless to defendant). Some of the more recent cases purport to announce harmless error standards, but they are not consistent and are difficult to reconcile. Compare
State v. Hunt,
150 Vt. 483, 490, 555 A.2d 369, 374 (1988) (change of venue, if improper, was harmless error, even though it could have affected defendant’s sentence, because his right to complete, fair and adequate trial was not jeopardized), and
State v. Jacques,
150 Vt. 508, 511, 554 A.2d 655, 656 (1987) (improper question was harmless error because it did not go to heart of close case, but instead there was overwhelming evidence of guilt), with
State v. Catsam,
148 Vt. 366, 372, 534 A.2d 184, 188 (1987) (improper expert testimony was not harmless because it could not be said beyond reasonable doubt that jury would have convicted defendant absent the error), and
State v. Wright,
154 Vt. 512, 519-20, 581 A.2d 720, 725 (1989) (error is harmless only if it is clear beyond reasonable doubt that jury would have convicted irrespective of the error; applied to error in charge to jury on lesser-included offense). Most recently, in
State v. Curavoo,
156 Vt. 72, 77, 587 A.2d 963, 966 (1991), and cases that follow it, we have noted the inconsistencies in our precedents and held the error involved was harmless under any standard. See
State v. Streich,
163 Vt. 331, 346-47, 658 A.2d 38, 49 (1995);
State v. Weller,
162 Vt. 79, 84 n.*, 649 A.2d 839, 842 n.*(1994). In
Curavoo,
we stated that our “harmless-error rule appears [to be] in need of clarification.” 156 Vt. at 76, 587 A.2d at 966.
Effectively, the issue of when a nonconstitutional error can be found harmless is an open one, on which we can start afresh. We start with an examination of federal law because it provides the two most widely accepted alternatives. The first, announced in
Chapman v. California,
386 U.S. 18, 24 (1967), applies to constitutional errors and provides that such errors may be found harmless only if the appellate court can state “a belief that it was harmless beyond a reasonable doubt.” It was intended to be a restatement of the holding of
Fahy v. Connecticut,
375 U.S. 85, 86-87 (1963), that a constitutional error in admitting evidence could not be harmless if “there is a reasonable possibility that the evidence complained of might have contributed to the conviction.” It was also intended to adopt the common-law rule
that the burden was placed on the beneficiary of the error to prove it was not harmful.
Chapman,
386 U.S. at 24.
The alternative is applicable to nonconstitutional errors and was first announced in
Kotteakos v. United States,
328 U.S. 750 (1946). The holding is in the following discussion:
If, when all is said and done, the conviction is sure that the error did not influence the jury, or had but very slight effect, the verdict and the judgment should stand .... But if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error, it is impossible to conclude that substantial rights were not affected. The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence. If so, or if one is left in grave doubt, the conviction cannot stand.
Id.
at 764-65 (citation omitted). The inquiry under
Kotteakos
is whether the error substantially influenced the outcome or there is grave doubt that the outcome is free of substantial influence. See
Bank of Nova Scotia v. United States,
487 U.S. 250, 256 (1988).
Most states have adopted either the
Kotteakos
or
Chapman
standards, or variations on these standards. In the New England states, for example, New Hampshire has adopted the
Chapman
test for nonconstitutional errors in criminal cases. See
State v. Vandebogart,
652 A.2d 671, 679 (N.H. 1994). Massachusetts has adopted the
Kotteakos
standard. See
Commonwealth v. Flebotte,
630 N.E.2d 265, 268 (Mass. 1994). Of the remaining states, the Rhode Island standard is similar to that in New Hampshire, see
State v. Dinagen,
639 A.2d 1353, 1357 (R.I. 1994) (nonconstitutional error, is harmless if ‘“it is not reasonably possible that such evidence would influence an average jury on the ultimate issue of guilt or innocence’ ”) (quoting
State v. Bowden,
439 A.2d 263, 269 (R.I. 1982)), and the Connecticut standard appears similar to that in Massachusetts. See
State v. Sauris,
631 A.2d 238, 252 (Conn. 1993) (harmless error if defendant cannot show “‘that it is more probable than not that the erroneous action of the court affected the result’”) (quoting
State v. Payne,
591 A.2d 1246, 1252 (Conn. 1991)). The Maine standard appears to fall somewhere in between. See
State v. Shortsleeves,
580
A.2d 145, 148 (Me. 1990) (error is harmless “if it is highly probable that it did not affect the judgment”).
For a number of reasons, we adopt the
Chapman
test for both constitutional and nonconstitutional errors. First, there is no logical reason to vary the standard based on whether the error has a constitutional or nonconstitutional source. As the Maryland Court of Appeals observed in adopting the
Chapman
test:
As we see it, there is no sound reason for drawing a distinction between the treatment of those errors which are of constitutional dimension and those other evidentiary, or procedural, errors which may have been committed during a trial. Although the Amendments to the United States Constitution are commonly considered a source of fair judicial procedure, other nonconstitutional evidentiary and procedural rules, signifying state policy with respect to judicial fairness, are often a defendant’s primary source of protection. An evidentiary or procedural error in a trial is bound, in some fashion, to affect the delicately balanced, decisional process. The abnegation of a particular rule upon which the defense intended to rely may often inflict more damage than initially apparent; a meritorious line of defense may be abandoned as a result; an important witness may not be called; strategies are often forsaken. The future course of the trial inevitably must be changed to accommodate the rulings made. It is the impact of the erroneous ruling upon the defendant’s trial and the effect it has upon the decisional process which is of primary concern, not whether the error is labelled as constitutional or nonconstitutional.
Dorsey v. State,
350 A.2d 665, 677 (Md. 1976). See
Commonwealth v. Story,
383 A.2d 155, 163 (Pa. 1978) (source of error is irrelevant in determining whether error is prejudicial to accused); Saltzburg,
The Harm of Harmless Error,
59 Va. L. Rev. 988, 1025 (1973) (“[I]t is erroneous to say that constitutional errors are more likely to have an injurious effect at trial than other errors.”). Indeed, the effect of
adopting different rules for constitutional and nonconstitutional errors is to require us to engage in an unnecessary inquiry into whether the error is constitutionally based. This inquiry is particularly required for evidentiary issues because they almost always have a constitutional dimension. See
State v. Cartee,
161 Vt. 73, 77, 632 A.2d 1108, 1111 (1993) (limits on cross-examination imposed by trial court under V.R.E. 403 analyzed as confrontation clause issue, and
Chapman
standard used to determine whether error was harmless);
State v. Lynds,
158 Vt. 37, 41-42, 605 A.2d 501, 503 (1991) (ruling that witness was unavailable so deposition could be admitted analyzed as confrontation clause issue to determine harmless error standard). We note in this regard that the New England states that have adopted the
Kotteakos
standard for nonconstitutional errors have required the
Chapman
standard for state constitutional errors, further blurring the distinction. See
State v. Oquendo,
613 A.2d 1300, 1314 (Conn. 1992);
Commonwealth v. Rios,
588 N.E.2d 6, 9 (Mass. 1992).
Either of the errors present in this case is arguably constitutional. We are reluctant to analyze them as such in light of our strong preference to resolve issues on nonconstitutional grounds where possible.
State v. Curtis,
157 Vt. 275, 277, 597 A.2d 770, 772 (1991).
Second, the
Chapman
standard is most consistent with the standard of proof in criminal trials and thus accords the greatest deference to the jury as trier of fact. As Professor Saltzburg points out, it makes “little sense to adopt the . . . standard, which is designed to prevent criminal convictions if there is even a reasonable doubt in the minds of the jurors as to the guilt of the person charged, and then on appeal to emasculate that evidentiary standard by allowing a conviction to stand when the trial court has violated evidentiary rules which might have influenced the jury by creating the requisite doubt.” Saltzburg,
supra,
at 992.
Third, the
Chapman
standard is in wide use, and we have used it frequently. The Supreme Court has developed factors to consider for its implementation. See
Delaware v. Van Arsdall,
475 U.S. 673, 684 (1986);
Lynds,
158 Vt. at 42-44, 605 A.2d at 503-04 (applying
Van Arsdall
factors). Thus, we believe application of the
Chapman
standard will lead to the most predictable results in an area where appellate analysis is often criticized as result-oriented. See, e.g., Allen,
A Serendipitous Trek Through the Advance-Sheet Jungle: Criminal Justice in the Courts of Review,
70 Iowa L. Rev. 311, 332 (1985).
Finally, we agree with the oft-repeated counsel that we must approach harmless error arguments cautiously. See
State v.
Randolph,
462 A.2d 1011, 1019 (Conn. 1983);
Shortsleeves,
580 A.2d at 148. The
Chapman
standard is the most consistent with careful and prudent use of harmless error.
Using the
Chapman
standard, we have no difficulty in concluding that the errors were harmless under that standard. Although the exclusion of the newspaper article was wrong, defendant testified on cross-examination that he read a newspaper that had his picture and stated, “Two sought for knife-point rape.” He testified that he fled in response to this story. Thus, the only impact upon defendant was that he was unable to show the actual newspaper story and its statement that the crime carried a maximum punishment of life imprisonment.
Defendant emphasizes that the State offered extensive evidence of his flight, that the jury acquitted him of the lesser charges and believed only one of the State’s sexual assault theories,
and that his sentence is long. None of these points, however, go to the significance of the evidence actually excluded. The excluded evidence could have improved the persuasiveness of defendant’s explanation for his flight only marginally. It did not address at all the State’s evidence that defendant fled from his apartment to avoid arrest before he read any newspaper articles. We can say beyond a reasonable doubt that introduction of the newspaper article would not have changed the outcome of defendant’s trial.
We have a similar reaction to the prior-consistent-statement evidence. Defense counsel’s opening statement to the jury stated that the victim told her sister she had been sexually assaulted by defendant. Without objection, the State elicited this information from the victim. In cross-examination, defense counsel asked in detail about the numerous conversations the victim had with her sister to show that she did not tell the sister the full story until the last conversation. Thus, the testimony of the sister simply confirmed what defense counsel had elicited from the victim, and we can say that admission of the sister’s testimony that the victim did relate the events consistent with her testimony at trial was harmless beyond a reasonable doubt.
Again, defendant argues that the overall case was close so that individual evidentiary items were important. Our focus on the evidence in issue should not be taken as acceptance of this position. Although defendant testified that he did not sexually assault the
victim, the State offered extensive corroboration of the victim’s version. Her physical condition and demeanor, defendant’s threats to his nephew and girlfriend as well as other statements he made, the defendant’s destruction of the victim’s shirt on which she testified he ejaculated after assaulting her, defendant’s failure to deny the crime and his flight to avoid arrest even before he knew the charge, all supported the State’s case. Perhaps the strongest corroboration was defendant’s own explanation of events. He asked the jury to believe that the victim consented to sex immediately after a fight in which he spat upon her, pushed her, and called her offensive names. According to defendant, the victim’s only concern about having sex with him was whether he wore a condom. Moreover, he testified that despite his anger because he thought he had contracted herpes from the victim, he nonetheless decided to have unprotected sex with her.
We conclude that any errors were harmless beyond a reasonable doubt.
Affirmed.